What to do With Wake-Up Stroke

Abstract

Wake-up stroke, defined as the situation where a patient awakens with stroke symptoms that were not present prior to falling asleep, represents roughly 1 in 5 acute ischemic strokes and remains a therapeutic dilemma. Patients with wake-up stroke were excluded from most ischemic stroke treatment trials and are often not eligible for acute reperfusion therapy in clinical practice, leading to poor outcomes. Studies of neuroimaging with standard noncontrast computed tomography (CT), magnetic resonance imaging (MRI), and multimodal perfusion-based CT and MRI suggest wake-up stroke may occur shortly before awakening and may assist in selecting patients for acute reperfusion therapies. Pilot studies of wake-up stroke treatment based on these neuroimaging features are promising but have limited generalizability. Ongoing randomized treatment trials using neuroimaging-based patient selection may identify a subset of patients with wake-up stroke that can safely benefit from acute reperfusion therapies.

Introduction

Definition

Acute stroke evaluation and management is fundamentally predicated on time from symptom onset.^1,2 Intravenous tissue plasminogen activator (tPA) remains the only Food and Drug Administration (FDA)-approved nonsurgical reperfusion therapy for acute stroke with evidence-based efficacy, and well-designed, adequately powered studies have consistently shown that efficacy is exquisitely time sensitive.^3–7 That being the case, knowledge of the exact time of symptom onset, or at least the time at which the patient was last known to be normal, is paramount.

Patients who go to sleep normal and awaken with stroke symptoms, a phenomenon known as “wake-up stroke,” present a management dilemma for acute stroke providers. Sometimes the period of sleep is short and a patient can still be eligible for tPA based on standard time-based criteria; however, when the time at which the patient was last known to be normal is the night prior to a morning presentation, which is often the case, the acute stroke provider is left without the key time-based data by which one typically makes safe therapeutic decisions for tPA candidacy. This makes for a diagnostic and therapeutic “gray area” in acute stroke practice.

Epidemiology

The wake-up stroke phenomenon is common. Numerous studies of various size and methodological strength through the years have given a sense of actual incidence of wake-up stroke as compared to other stroke presentations. These mostly stroke-registry–based studies range in estimation of wake-up stroke incidence from 8% in California⁸ to 33% in a region of France⁹ to nearly 39% in Ohio¹⁰ but most typically suggest somewhere in between 15% and 25%.^11–24 Clinical and radiographic characteristics distinguishing wake-up stroke from other modes of stroke onset have been sought, but results have been conflicting. Some older studies have suggested that wake-up strokes seem to be more severe at onset^14,19 and portend a worse outcome overall,^17,19 while others suggested there are no appreciable clinical or radiographic differences between wake-up and “while awake” strokes.^{16,22,25–27}

The best estimate of wake-up stroke prevalence comes from a retrospective population-based study of 1854 acute ischemic strokes in the Greater Cincinnati/Northern Kentucky region. In this representative biracial sample, 273 (14.3%) of acute strokes were wake-up strokes, resulting in an adjusted event rate of 26.0/100 000. No clinically significant differences in baseline characteristics were observed between wake-up and nonwake-up strokes; however, patients with wake-up stroke were older (72.3 vs 70.0 years, P = .01) and had higher baseline retrospectively calculated National Institutes of Health Stroke Scale (NIHSS) scores (4 vs 3, P = .004). Importantly, 98 (35%) patients were otherwise eligible for tPA if time was not a factor.²⁷

Overall, in spite of the methodological heterogeneity and different focus of published studies, the common theme is that wake-up stroke is not rare and the clinical features suggest that there is a place for therapeutic optimism; although, no definitive clinical or radiographic paradigm has yet been established to select wake-up stroke candidates for safe and efficacious reperfusion therapy (Table 1).

Table 1.

Wake-Up Stroke Characteristics by Study.

Authors	Study Design	Total Patients	Wake-Up Stroke #, %	Clinical Differences vs While-Awake Stroke	Outcome Differences vs While-Awake Stroke	Imaging
CASPR group8	Retrospective, prospectively collected data, US state registry	374	30 (8)	—	—	—
Michel et al9	Retrospective, prospectively collected data, hospital registry	1633	568 (33.1)	—	—	—
Tanimoto et al10	Retrospective, prospectively collected data, hospital registry	72	28 (38.9)	WUS: tended to be African American, younger, small vessel mechanism, less severe NIHSS, worse lipid profile	—	—
Marler et al11	Retrospective, prospectively collected data, hospital registry	1167	331 (28)	—	—	—
Ricci et al12	Retrospective, prospectively collected data, regional registry	375	68 (18.1)	—	—	—
Lago et al13	Retrospective, prospectively collected data, hospital registry	1223	309 (25.2)	—	—	—
Bornstein et al14	Retrospective, prospectively collected data, national registry	1671	311 (18.6)	WUS more severe	—	—
Chaturvedi et al15	Subanalysis of prospective RCT	1272	323 (25.4)	—	—	—
Serena et al16	Retrospective, prospectively collected data, national registry	1248	301 (24.1)	None	—	WUS: CT head normal in 39.4% of patients seen within 6 hours of symptom recognition (60% in stroke while awake)
Nadeau et al17	Retrospective, prospectively collected data, national registry	2585	349 (13.5)	WUS had higher BP and ischemic stroke subtype	WUS less likely to return home	—
Boode et al18	Retrospective, hospital registry	263	48 (18.3)	—	—	—
Jiménez-Conde et al19	Retrospective, prospectively collected data, hospital registry	813	127 (15.6)	WUS had more obesity, less AF, and higher initial stroke severity	WUS had worse 3-month outcome	—
Silva et al20	Prospective cohort study, hospital registry	676	131 (19.4)	None	None	Similar prevalence of CTP mismatch and arterial occlusion in WUS and known onset groups
Turin et al21	Retrospective, prospectively collected data, national registry	897	87 (9.7)	WUS more hypertension and increased initial severity	None	—
Fink et al22	Retrospective, prospectively collected data, hospital registry	364	100 (27)	None	—	Similar prevalence of MRI DWI/PWI mismatch
Moradiya et al23	Subanalysis of a prospective RCT	17 398	5152 (29.6)	WUS initially less severe	None	—
Koton et al24	Retrospective, prospectively collected data, national registry	4408	820 (18.6)	None	None	20%-40% prevalence of penumbra
Todo et al25	Retrospective, prospectively collected data, hospital registry	158	17 (10.8)	—	—	CT findings in WUS similar to patients within 3 hours of known symptom onset
Huisa et al26	Prospective cohort study, hospital registry	96	28 (29.6)	None	Trend toward favorable (0-1) 90 d mRS in WUS vs 4 hours from symptoms controls (73% vs 45%)	Favorable CT ASPECTS (8-10) similar in WUS and known 4 h from symptoms (89.3% vs 95.6%)
Mackey et al27	Population-based registry	1854	273 (14.7)	“Minor differences” in age and rNIHSS (WUS older, higher rNIHSS)	None	—
Roveri et al28	Retrospective, prospectively collected data, hospital registry	1531	190 (12.4)	None	Outcome better in controls (patients treated with tPA within 3 hours of symptoms)	Baseline ASPECTS similar in WUS and controls within 3 hours of symptoms and treated with tPA
Manawadu et al29	Retrospective, prospectively collected data, hospital registry	1836	193 (10.5)	—	Outcome better in thrombolyzed WUS vs nonthrombolyzed WUS	CT ASPECTS and CTP to select patients for IV tPA

Abbreviations: WUS, wake-up stroke; RCT, randomized controlled trial; CT, computed tomography; CTP, CT perfusion; MRI, magnetic resonance imaging; DWI, diffusion-weighted imaging; PWI, perfusion-weighted imaging; ASPECTS, Alberta Stroke Program Early CT Score; rNIHSS, retrospective National Institutes of Health Stroke Scale; BP, blood pressure; AF, atrial fibrillation.

Pathophysiologic Hypotheses

The wake-up stroke phenomenon is incompletely understood pathophysiologically. What seems clear, though, is that wake-up strokes are not actuarial quirks of evenly spread stroke risk through the course of a day but likely the result of circadian changes in coagulability, serum catecholamine levels, and autonomic tone. Much like cardiac events,³⁰ there is a preponderance of strokes of all subtypes in the morning as compared to evening onset.^11,31 Several homeostatic and structural factors may contribute to this phenomenon. Proposed factors include sleep-disordered breathing with or without patent foramen ovale,^32–34 overnight changes in autonomic tone affecting blood pressure with morning surges,^35,36 morning increases in platelet aggregation^37,38 relatively refractory to clopidogrel,³⁹ endothelial dysfunction,⁴⁰ blood viscosity,⁴¹ and fluctuating prothrombotic/fibrinolytic factor level balance.^42–44 The circadian blood pressure-related changes behind the morning “surges,” which essentially mirror stroke incidence through the course of a day, are a tempting therapeutic target. Given the preponderance of strokes of all types between 0600 and 1200,³¹ a treatment trial targeting morning blood pressure changes did not change the distribution of strokes through the course of a day.⁴⁵ Lending credence to the contribution of overnight paroxysms of atrial fibrillation,⁴⁶ a recent study demonstrated a significant association between wake-up stroke and a new diagnosis of atrial fibrillation.⁴⁷ Given the heterogeneity of wake-up stroke subtypes, it is likely that no one factor underlies wake-up stroke but some combination of the aforementioned and other yet undiscovered contributors (Table 2).

Table 2.

Proposed Pathophysiologic Mechanisms of Wake-Up Stroke.

Structural	Sleep-disordered breathing32 and PFO33 New-onset atrial fibrillation47 Endothelial dysfunction40
Homeostatic	Morning blood pressure “surge”31,35,36
Serological	Increased viscosity41 Increased platelet aggregation37–39 Increased procoagulant factors42–44

Abbreviation: PFO, patent foramen ovale.

Neuroimaging and Wake-Up Stroke

The key feature of wake-up stroke that makes it a therapeutic dilemma is the absence of distinct time of symptom onset, which limits the ability to establish eligibility for acute reperfusion therapies. Diagnostic neuroimaging in a patient with wake-up stroke thereby plays an even stronger role than usual in acute stroke evaluations. Several studies have been conducted to evaluate neuroimaging modalities as a surrogate marker of cerebral ischemia and the so-called “tissue clock” to supplant the absent time of onset that starts the strictly clinical “time clock.”

Several noncontrast computed tomography (CT)-based studies compared early ischemic changes on CT between wake-up stroke and stroke of known onset. Overall, there was no significant difference in early CT changes between wake-up stroke and stroke of known onset within 3 hours^25,28 or 6 hours.²⁶ These results suggest that the ischemic insult may occur shortly before or at the time of awakening in the absence of early ischemic change.

Perfusion and volume-based imaging with magnetic resonance (MR) or CT-based studies provide more granular physiologic data than noncontrast CT for acute stroke (see Table 3). These advanced neuroimaging techniques estimate the volume of brain tissue potentially at risk for progression to infarction (ie, ischemic penumbra) if recanalization does not occur. The volumetric difference between a surrogate for established infarction and penumbra, if present, is referred to as a “mismatch” and represents a rational biomarker for treatment selection (see Figure 1). Studies of advanced neuroimaging techniques have been conducted in patients with wake-up stroke. An MR-based study of diffusion-weighted imaging (DWI) and perfusion-weighted imaging (PWI), imaging surrogates of ischemic, and “at-risk” tissue, respectively, found that patients with wake-up stroke and stroke of known onset had similar DWI and PWI lesion volumes as well as a similarly high proportion of DWI-PWI mismatch.⁵² A similarly designed CT-based study of mismatch between cerebral blood volume (like DWI, the estimate of infarcted tissue) and cerebral blood flow (like PWI, the at-risk tissue) found no difference between the percentage of mismatch between patients with wake-up stroke and stroke of known onset within a therapeutic window.²⁰

Table 3.

Wake-Up Stroke Imaging Modalities.

Parenchymal imaging	CT	Noncontrast CT	Screen for early ischemic changes	Time to ischemic change: variable; with MCA occlusion as little as 1 hour48 but often 3+ hours
		Perfusion CT CBV—total volume of  blood in a given  volume of brain  (mL/100 g) CBF—total volume of  blood moving through a  given volume of brain  (mL/100 g) MTT—the average transit  time of blood through a  brain region TTP—time from contrast  arrival-to-peak-intensity  flow through a brain  region	Screen for infarct/penumbra mismatch (see Figure 1)	Time to ischemic change: variable, depends on patient hemodynamics and CBV threshold for “infarct,” but is typically abnormal before clear signs on noncontrast CT
	MRI	Routine MRI	Screen for infarct volume, DWI/FLAIR mismatch (see Figure 2)	Time to ischemic change: DWI—3 minutes49; T2WI—1-4 hours50; FLAIR—3-6 hours51
		Perfusion (PWI) MRI CBV, CBF, and  MTT as defined  previously are the  basic sequences  of a PWI protocol	Screen for infarct/penumbra mismatch
Vascular imaging	CT	CT angiography	Screen for large artery occlusion
	MRI	MR angiography
	Ultrasound	Complete neurosonology (carotid duplex ultrasonography plus transcranial Doppler)

Abbreviations: CT, computed tomography; CBV, cerebral blood volume; CBF, cerebral blood flow; MTT, mean transit time; TTP, time to peak; MRI, magnetic resonance imaging; DWI, diffusion-weighted imaging; PWI, perfusion-weighted imaging; FLAIR, fluid attenuated inversion recovery.

Figure 1.

Multimodal CT mismatch (or “penumbra”). Panel (A) is a cerebral blood volume (CBV) map. The dark area noted in the left frontal operculum suggests low contrast volume in the region and is considered a surrogate for infarcted tissue, or the “infarct core.” The other maps—cerebral blood flow (CBF) in panel (B), time to peak (TTP) in panel (C), and mean transit time (MTT) in panel (D)—are different measures of contrast movement through cerebral vasculature (see Table 3) and clearly involve much more of the left hemisphere than the CBV map. This discordance is referred to as a multimodal CT mismatch or “penumbra” and may represent tissue at risk of infarction but potentially salvageable by reperfusion therapy. Siemens SOMATOM, syngo perfusion software. CT indicates computed tomography.

Another interesting mismatch approach utilizes DWI and fluid attenuated inversion recovery (FLAIR) sequences of magnetic resonance imaging (MRI) to identify infarcted versus at-risk tissue. Both sequences detect cerebral water changes but in different time sequences (see Figure 2). The DWI sequence is very sensitive to early cerebral water changes; however, the abnormalities noted do not change once they appear after the first several minutes of ischemia, so exact timing of an ischemic injury cannot be made with DWI. The T2-based FLAIR sequence measures the accumulation of cerebral edema as the infarction process proceeds. Thus, in principle, the presence of a DWI lesion and absence of a matched FLAIR abnormality should represent a relatively early infarct. This idea has been studied in several single-center^51,53–56 pilots and a multicenter investigation.⁵⁷ Overall, the DWI-FLAIR mismatch was found to very accurately identify ischemic tissue beyond 3 to 6 hours and can identify ischemia within the 3- to 4.5-hour window with excellent specificity. The large multicenter study of 543 patients supported the findings of the single-center studies, with DWI-FLAIR mismatch identified in patients within 4.5 hours of stroke onset with a sensitivity of only 62% but a good specificity of 81%, although interrater agreement was less than ideal (κ = .569). A recent publication sought to increase interrater reliability of DWI-FLAIR mismatch identification by color-coding FLAIR intensity and did just that with a roughly 10% increase in positive predictive value for both observers (85%-95% in one and 72%-82% in the other) once color coding was introduced.⁵⁸ A recent small observational study noted a DWI-FLAIR mismatch in 44% of their patients with wake-up stroke.⁵² These findings supported the initiation of thrombolysis treatment trials based on the identification of a DWI-FLAIR mismatch, which are ongoing and will be discussed further in a subsequent section.

Figure 2.

The DWI/FLAIR mismatch. These 2 axial images of the brain at a level just above the lateral ventricles represent the so-called DWI/FLAIR mismatch that can be seen in the early hours after symptom onset when DWI (left) hyperintensity—which can arise in minutes from symptom onset—occurs in the absence of T2-based FLAIR (right) hyperintensity, which takes 3 to 6 hours to develop. DWI indicates diffusion-weighted imaging; FLAIR, fluid attenuated inversion recovery.

Emergent neuroimaging is of heightened importance in the setting of wake-up stroke, given the absence of a clear time of symptom onset. Imaging modalities as simple and rapid as noncontrast CT and as sophisticated as multimodal CT and MRI show promise in identifying patients with acute ischemic stroke who may benefit from systemic thrombolysis. Ongoing randomized treatment trials based on these neuroimaging findings have practice-changing implications.

Treatment Evidence

Currently, there is a lack of a high-level evidence base to support any acute treatment in the setting of wake-up stroke. In the absence of high-level data, there are now many small studies of off-label use of acute reperfusion therapies for wake-up stroke employing various clinical and neuroimaging criteria for inclusion.

The majority of studies conducted so far are small, single-center observational studies of off-label intravenous tPA with or without endovascular reperfusion therapies.^29,59–75 Largermulticenter studies,⁷⁶ organized phase II studies,⁷⁷ and randomized placebo-controlled treatment trials^78–80 are fewer in number. Of note, the only large, multicenter, randomized placebo-controlled trial studied a glycoprotein IIb/IIIa inhibitor (abciximab) as an adjunct to intravenous tPA and patients with wake-up stroke made up only a small subgroup of the cohort. The individual characteristics and levels of evidence are detailed in Table 4.

Table 4.

Wake-Up Stroke Treatment Studies and Level of Evidence.

Authors	Study Design	Clinical Inclusion Criteria	Imaging Inclusion Criteria	# Patients With Wake- Up Stroke (Treated)	# Control Patients	Treatment Type	Mean Door to Treatment in WUS	Mean/Median NIHSS of Treated Patients (Control)	Mean Age of Treated Patients (Control)	sICH in WUS, % (Control)	mRS 0-1, % (Control)	mRS 0-2, % (Control)	Level of Evidencea
Iosif et al60	Case report	“Admitted to hospital rapidly after waking up with stroke”	MRI DWI/PWI and DWI/FLAIR mismatch, MRA occlusion	2 (2)	0	IAT	–	15.5	45	50	At d/c: 100	–	Class IIb level C
Kuruvilla et al71	Case report	–	CT	2 (2)	–	IAT	–	16	33	0	–	–	Class IIa level C
Bracco et al74	Case report	–	Multimodal CT	1 (1)	–	IAT	–	12	74	0	–	–	Class IIb level C
Sung and Lee75	Retrospective review	–	–	10	–	IAT	168 min	19	–	0	–	90 d: 20	Class IIb level C
Stampfl et al69	Retrospective review	WUS, NIHSS ≥10	CTA occlusion, multimodal MRI mismatch (DWI/PWI)	19	–	IAT	–	11	73.7	21.1	–	10.5	Class IIb level C
Jung et al70	Retrospective review	NIHSS >4, symptoms <24 h but >6 h from last normal	Multimodal MRI mismatch (DWI/PWI)	55 (55)	804	IAT	–	16.8 (16.8)	61.9 (62.6)	3.7 (6)	90 d: 16.7 (23.3)	90 d: 37 (39.7)	Class IIa level C
Natarajan et al68	Retrospective review	7-23 h from last normal, WUS, mRS ≤1, NIHSS 5-22	CT ASPECTS, multimodal CT mismatch	30 (30)	–	IAT	–	13	72	10	–	20	Class IIb level C
Natarajan et al72	Retrospective review	WUS within 12 h of noticing symptoms, NIHSS >8	Multimodal CT mismatch, CTA occlusion	25 (25)	–	IAT	–	–	–	14.3b	–	42.9b	Class IIb level C
Barreto et al59 , c	Retrospective review	Major deficit from stroke, neurologically normal prior to stroke	No EIC in >1/3 of vascular territory	80 (46)	34/174	IV tPA, IAT, or both	2.4 h/1.2 h	16 (10.5/11)	62 (64/65)	4.3 (0/2.9)	At d/c: 14 (6/48)	At d/c: 28 (13/48)	Class IIb level C
Cho et al61	Retrospective review	Present within 6 hours of symptom recognition	Multimodal CT or MRI mismatch	26 (26)	223	IV tPA, IAT or both	154 min (90 min)	14.5 (13)	67.1 (65.8)	6.3 (5.8)	90 d: 37.5 (35)	90 d: 50 (49.3)	Class IIa level C
Breuer et al62	Retrospective review	Present within 6 hours of symptom recognition	Multimodal MRI mismatch	45 (10)	35	IV tPA	–	–	–	0 (0)	90 d: 30 (31)	90 d: 50 (60)	Class IIb level C
Kim et al63	Retrospective review	Present within 3 hours of symptom recognition	CT ASPECTS, multimodal MRI mismatch	26 (-)	49	IV tPA, IAT or both	–	13 (12)	67 (72)	10.3 (8.2)	90 d: 27.6 (4.1)	90 d: 44.8 (14.3)	Class IIa level C
Manawadu et al73	Retrospective case-control	WUS, last normal <12 hours but >4.5 hours	CT ASPECTS, multimodal CT mismatch	122 (68)	54	IV tPA	–	11.5 (9)	73.9 (70.6)	2.9 (0)	90 d: 16.2 (9.3)	90 d: 36.8 (25.9)	Class IIa level C
Aoki et al64	Prospective cohort	Unknown onset stroke, “last known normal not consistent with first found abnormal”	MRI DWI/FLAIR mismatch	4 (4)	–	IV tPA	1 hour	15.5	73.25	0	90 d: 25	90 d: 25	Class IIb level C
Ebinger et al54,65	Trial substudy, observational cohort	European guideline, “disregarding the contraindication of unknown time of onset”	Multimodal MRI mismatch	13 (13)	131	IV tPA	86 mind (60 min)	13d	81d	0 (3.1)	90 d: 29.4d (38.9)	90 d: 35.3d (49.6)	Class IIb level C
Manawadu et al66	Prospective case-control	WUS, last normal <12 hours but >4.5 hours vs controls within 4.5 hours symptom onset	CT ASPECTS	68 (68)	326	IV tPA	73 min (60 min)	12 (13)	73.9 (72.8)	2.9 (3.4)	90 d: 16 (24)	90 d: 37 (38)	Class IIa level C
Bai et al67	Prospective case—control	WUS and all patients with ischemic stroke within 12 h of symptom onset	Multimodal MRI mismatch (DWI/FLAIR)	68 (48)	172	IV tPA	–	–	–	2 (2)	90 d: 77 (76)	–	Class IIa level C
Kang et al76	Prospective multicenter observational	Last normal and symptom awareness times discordant, in emergency department within 6 hours of symptom awareness	Multimodal MRI mismatch (DWI/PWI and DWI/FLAIR)	63 (63)	156	IV tPA, IAT or both	155 min	14 (12)	67 (70)	3.2 (-)	90 d: 28.6 (-)	90 d: 46 (-)	Class IIa level B
Hill et al77	Prospective observational	Last normal <12 hours, WUS, “disabling stroke”	Arterial occlusion, ASPECTS >5	89 (20)	–	IV tPA	148 min (awakening to treatment)	13	75	0	45	–	Class IIa level C
Adams et al78,79	Subanalysis of RCT	WUS within 3 hours of symptom awareness	CT	43 (43; 22 tPA + abciximab, 21 tPA only)	758	IV tPA + abciximab or placebo	–	10 (8)	69.5 (68.9)	18.2 (4.8)	–	90 d: 9.3 (29.2)	Class III harm level B
Michel et al9,80	Pilot RCT	Supratentorial stroke, too late for standard tPA but not more than 24 hours or WUS	Multimodal CT	9 (4)	5	IV tPA	109.5 min (113 min)	17 (14.5)	69.5 (49)	0 (0)	-e	-e	Class IIa level C

Abbreviations: sICH, symptomatic intracranial hemorrhage; mRS, modified Rankin Scale; CT, computed tomography; MRI, magnetic resonance imaging; CTP, CT perfusion; DWI, diffusion-weighted imaging; PWI, perfusion-weighted imaging; FLAIR, fluid attenuated inversion recovery; IAT, intra-arterial therapy; IV tPA, intravenous tissue plasminogen activator; d/c, discharge; ASPECTS, Alberta Stroke Program Early CT Score; WUS, wake-up stroke; EIC, early ischemic change; NIHSS, National Institutes of Health Stroke Scale; d, day.

^aClassification schema from the American Heart Association/American Stroke Association.

^bAnterior circulation WUS only.

^cTwo control groups: control 1 = untreated WUS and control 2 = treated patients within 3 hours of known symptom onset (control 1/control 2).

^dInvestigators combined WUS with unknown onset strokes, so exact proportions of WUS alone is unknown.

^eMean mRS at 90 d was 1.5 in the treatment group, 3 in placebo.

Their inherently biased study designs, small numbers, and marked heterogeneity in inclusion criteria and reported results do not allow for much to be said about acute reperfusion therapy for wake-up stroke other than there seems to be a case for therapeutic optimism and large, randomized, placebo- and sham-controlled studies are justified. Ongoing trials include Efficacy and Safety of MRI-based thrombolysis in Wake Up Stroke (WAKE-UP),^81,82 THrombolysis for Acute Wake-up and Unclear-onset Strokes With Alteplase at 0.6 mg/kg Trial (THAWS),⁸³ Extending the Time for Thrombolysis in Emergency Neurological Deficits (EXTEND),⁸⁴ A Phase IIa Safety Stusdy of Intravenous Thrombolysis With Alteplase in MRI-Selected Patients (MR WITNESS),⁸⁵ Safety of Intravenous Thrombolysis for Wake Up Stroke,⁸⁶ Diffusion-Weighted Imaging or Computerized Tomography Perfusion Assessment With Clinical Mismatch in the Triage of Wake Up and Late Presenting Strokes Undergoing Neurointervention (DAWN),⁸⁷ Safety of Intravenous Thrombolytics in Stroke on Awakening (SAIL-ON),⁸⁸ and Wake Up Symptomatic Stroke in Acute Brain Ischemia (WASSABI; Table 5).⁸⁹

Table 5.

Ongoing Wake-Up Stroke Treatment Trials.

WAKE-UP81,82 THAWS83 EXTEND84 MR WITNESS85 Safety of Intravenous Thrombolysis for Wake Up Stroke86 DAWN87 SAIL-ON88 WASSABI89

Abbreviations: DAWN, Diffusion-Weighted Imaging or Computerized Tomography Perfusion Assessment With Clinical Mismatch in the Triage of Wake Up and Late Presenting Strokes Undergoing Neurointervention; EXTEND, Extending the Time for Thrombolysis in Emergency Neurological Deficits; MR WITNESS, A Phase IIa Safety Study of Intravenous Thrombolysis With Alteplase in MRI-Selected Patients; SAIL-ON, Safety of Intravenous Thrombolytics in Stroke on Awakening; WASSABI, Wake Up Symptomatic Stroke in Acute Brain Ischemia; THAWS, THrombolysis for Acute Wake-up and Unclear-onset Strokes With Alteplase at 0.6 mg/kg Trial.

What to do With Wake-Up Stroke

Evaluation

Based on the studies mentioned previously and our own clinical experience, we recommend a wake-up stroke evaluation proceeds as any other acute stroke assessment, with as much clinical detail as possible, an NIHSS examination and basic laboratory studies according to guidelines. This should be followed by multiparametric CT or MRI with noninvasive angiography or fast-track neurosonology (eg, transcranial Doppler and carotid duplex ultrasonography)^90–94 to screen for an infarction/at-risk mismatch and large arterial occlusion, respectively. Even if one is uncomfortable providing treatment recommendations based on the results given the lack of high-level supporting evidence, the diagnostics can serve to inform prognostication.

Treatment

Regarding treatment of wake-up stroke, the authors agree with the sentiment of the recent systematic review of treatment strategies in wake-up stroke⁹⁵ that routine treatment of wake-up stroke cannot be offered based on available evidence. Readers are encouraged to participate in clinical trials so that we as stroke providers may know how best to treat our patients with wake-up stroke (Table 6).

Table 6.

Wake-Up Stroke Recommendations.

Evaluation	Clinical history, NIHSS, CT head, laboratories per AHA/ASA guidelines96 Multiparametric CT or MRI (eg, CT or MR perfusion)
Treatment	No routine therapy can be offered based on available evidence Trial participation is encouraged

Abbreviations: NIHSS, National Institutes of Health Stroke Scale; CT, computed tomography; AHA/ASA, American Heart Association/American Stroke Association; MRI, magnetic resonance imaging.