Wake-up Stroke Within 3 Hours of Symptom Awareness: Imaging and Clinical Features Compared to Standard Recombinant Tissue Plasminogen Activator Treated Stroke Luisa Roveri, MD,* Sara La Gioia, MD,* Chiara Ghidinelli, MD,* Nicoletta Anzalone, MD,† Costantino De Filippis, MD,† and Giancarlo Comi, MD*

Background: Patients with wake-up stroke (WUS) are excluded from thrombolysis because of unknown time of symptom onset. Previous studies have reported similar stroke severity and early ischemic changes (EICs) in patients with WUS and stroke of known onset. These studies, however, included patients within a large timeframe to imaging or did not quantify EICs. The aim of our study was to quantify EICs of patients with WUS presenting within 3 hours of symptom recognition compared to standard 3-hours recombinant tissue plasminogen activator (rt-PA)–treated patients and assess the extent of ischemic lesion and functional independence at follow-up. Methods: Patients were selected from our prospectively collected stroke database. Baseline and follow-up computed tomographic scans were graded with Alberta Stroke Program Early Computed Tomography Score (ASPECTS). Clinical outcome measures were modified Rankin Scale score, mortality, and symptomatic intracerebral hemorrhage. Results: Demographic features, risk factors, stroke severity, and baseline ASPECTS were similar in both groups. WUS and rt-PA–treated patients had similar tissue outcome (median ASPECTS 7.0 vs 7.5; P 5 .202). Functional outcome was more favorable in rt-PA–treated patients (61.6% vs 43.1%; odds ratio [OR] 2.12; 95% confidence interval [CI] 1.05-4.28; P 5 .037). After adjusting for age, stroke severity, treatment, and EICs in less than one-third of middle cerebral artery territory, rt-PA and National Institutes of Health Stroke Scale scores remained the only significant predictors of outcome (OR 7.76; 95% CI 2.40-25.05; P 5 .001 and OR 0.74; 95% CI 0.67-0.82; P ,.001, respectively). Conclusions: Within 3 hours of symptom recognition, patients with WUS have EICs similar to rt-PA–treated patients. It is reasonable to expect that selected WUS patients might benefit from thrombolysis within 3 hours of symptom awareness. Key Words: ASPECTS—early ischemic changes— recombinant tissue plasminogen activator—thrombolysis—wake-up stroke. Ó 2013 by National Stroke Association

Even though stroke has a high prevalence, thrombolytic treatment is currently applied only to a small group of patients.1 Under the present criteria for thrombolysis, From the *Departments of Neurology; and †Neuroradiology, San Raffaele Scientific Institute, Milan, Italy. Received August 26, 2011; revision received October 6, 2011; accepted October 15, 2011. Address correspondence to Luisa Roveri, MD, Department of Neurology, Stroke Unit, Ospedale San Raffaele, via Olgettina 48, 20132 Milan, Italy. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2013 by National Stroke Association doi:10.1016/j.jstrokecerebrovasdis.2011.10.003

wake-up stroke (WUS) patients who awaken with the neurologic deficit are excluded from treatment because it is not possible to establish the real time of symptom onset.2 Because WUS represents about 25% of stroke events,3-7 extending the use of recombinant tissue plasminogen activator (rt-PA) to these patients might have a relevant positive impact on outcome and longterm health costs. Some instances of WUS may in fact occur near the time of awakening or patients may wake up at the time of stroke onset. In this subgroup of patients, thrombolysis may be of benefit provided that patients arrive rapidly to the emergency department. If reperfusion therapy is to be used in patients with WUS, it is

Journal of Stroke and Cerebrovascular Diseases, Vol. 22, No. 6 (August), 2013: pp 703-708

703

704

important to carefully select patients that will more likely benefit from treatment. This requires a better knowledge of the actual differences between WUS patients and standard rt-PA stroke patients. In the current literature, a number of studies that compare clinical and radiologic features of patients with WUS and patients with stroke of known onset have not found major differences in stroke severity or early ischemic changes (EICs).8-11 The principal limitations when comparing results of previous studies are different inclusion criteria in the WUS group, the large variability of time to imaging, heterogeneity of imaging methods, lack of imaging score to quantify EICs on baseline computed tomographic (CT) scan, and lack of radiologic outcome. Neuroimaging is a potentially powerful tool to identify patients with uncertain time of stroke onset that could possibly benefit from thrombolysis. Magnetic resonance imaging (MRI) and perfusion CT scan have been progressively introduced to study stroke patients. However, advanced imaging is not routinely available in most centers. In a routine clinical setting with a limited time window, CT scan is still the imaging of choice for acute stroke patients. The aim of our study was to assess whether baseline and follow-up clinical and imaging features of patients with WUS within 3 hours of symptom awareness, eligible for thrombolysis except for stroke onset modality, differ from patients with stroke onset within 3 hours treated with thrombolysis according to the current guidelines. In order to standardize the measurement of EICs, we used the Alberta Stroke Program Early Computed Tomographic Score (ASPECTS) that has established validity and reliability in real-time ratings.12

Methods We selected consecutive patients between February 2005 and July 2010 from our prospectively collected stroke database. The database contains data on demographic features, time of symptom onset, time of arrival to the emergency department, time to CT scan, stroke severity measured with the National Institutes of Health Stroke Scale (NIHSS) score, vascular risk factors, treatment, outcome at hospital discharge and at 3 months measured with modified Rankin Scale (mRS). We split patients with known onset stroke from patients with WUS. We then considered among patients with WUS those arriving at the emergency department within 3 hours of symptom onset and potentially eligible for thrombolysis except for stroke onset modality for inclusion in the WUS group. Patients arriving at the emergency department within 3 hours of symptom onset eligible for thrombolysis and treated accordingly were included in the standard rt-PA group. Although the current labelling of rt-PA is still restricted to the 3-hour time window, after the results of the European Cooperative Acute Stroke Study III (ECASS-3) trial showing that there is a benefit

L. ROVERI ET AL.

also in patients treated between 3 and 4.5 hours, the European Stroke Organization has recommended extending rt-PA treatment to 4.5 hours (http://www.esostroke. org). In compliance with this recommendation, rt-PA is administered at our institution up to 4.5 hours after stroke, allowing patients arriving within 3 hours of stroke onset to perform the basic workup for thrombolysis. Additional exclusion criteria for both groups of patients were: (1) pre-existing disabling neurologic deficit with mRS score $3; (2) missing baseline CT scan imaging for ASPECTS grading; and (3) stroke involving arterial territory other than middle cerebral artery (MCA), because ASPECTS does not address ischemic changes occurring in the anterior cerebral artery (ACA) and posterior circulation territory. The presence of EICs was assessed in all baseline CT scans and measured with ASPECTS. The grading system, based on the identification of 10 regions in the territory of the MCA across 2 standardized axial CT cuts, assigns 10 points to a normal CT scan. One point is subtracted for each region with EIC identified as parenchymal hypoattenuation, loss of gray–white boundary (which corresponds to hypoattenuation of the gray matter) or swelling.12 All axial cuts were used to examine individual regions. At our institution, the standard CT scan protocol for patients with suspected stroke uses 3-mm axial cuts, which allows for better visualization of all ASPECTS regions on more than one slice and improves differentiation of hypoattenuation as a result of volume averaging of adjacent sulci rather than real EIC. The extent of the final infarction was also evaluated with ASPECTS on followup imaging.13 Baseline CT scans were performed within 4 hours, and follow-up imaging was performed between 24 to 72 hours. All images were evaluated by an independent reader blinded to patient demographics, stroke onset modality, treatment, and clinical outcome. The only clinical information given to readers was the side on which patients experienced the symptoms. Baseline data for all patients included age, gender, time of symptom onset or recognition, NIHSS score, vascular risk factors, baseline ASPECTS grading, symptoms to CT scan, and symptoms to needle time for rt-PA–treated patients. Stroke etiology was assigned according to Trial of Org 10172 in Acute Stroke Treatment criteria.14 Clinical outcome was assessed by mRS and defined as favorable for scores 0 to 2 (functionally independent). Radiologic outcome included evolution of EICs and the presence of intracranial hemorrhage (ICH) according to Safe Implementation of Thrombolysis in StrokeMonitoring Study criteria which define symptomatic intracerebral haemorrhage as a parenchymatous haemorrhage type 2 (blood clots exceeding 30% of the infarct area with significant space occupying effect), at post-treatment scan combined with neurological deterioration leading to an increase of 4 points on the NIHSS.15 The study was approved by the institutional review board.

WAKE-UP STROKE WITHIN 3 HOURS OF SYMPTOMS AWARENESS

Statistical Analysis Categorical variables were presented as counts (percentage) and continuous variables as means (SD) or medians (interquartile range [IQR]). Continuous variables were analyzed using the Mann–Whitney or t tests. The Chi-square was used to analyze categorical variables. Patients with WUS were compared to standard rt-PA– treated patients with regard to patient demographics, risk factors, NHISS score, baseline ASPECTS, follow-up ASPECTS, and 90-day mRS score. Univariate logistic regression analysis was used to determine variables associated with outcome. Multivariate logistic regression analysis was calculated for parameters showing significant association or a trend in univariate analysis (P , .1) with outcome as the dependent variable. Missing data were not included in the statistical analysis. Statistical levels were 2-tailed. The a significance level was set at P , .05. All analyses were performed using SPSS software (version 13.0; SPSS Inc, Chicago, IL).

Results A total of 1531 ischemic stroke patients were admitted to our stroke center between February 2005 and July 2010. Of these, 125 patients consecutively arrived within 3 hours of symptom onset and were treated with standard of care rt-PA (8.2%). The exclusion of 26 patients with stroke in the ACA or posterior circulation, 11 patients with missing baseline CT scan images, and 2 patients with prestroke mRS scores $3 left 86 patients in the standard rt-PA group. A total of 190 consecutive WUS patients were

705

selected from our database (12.4%). After excluding 105 patients arriving beyond 3 hours of symptom recognition or with contraindication to thrombolysis other than stroke onset modality, 11 patients with ACA or posterior circulation stroke, 6 patients with missing baseline CT scan images, and 8 patients with prestroke mRS score $3, 60 patients remained in the WUS group who arrived within 3 hours of symptom awareness. Nine WUS patients received rt-PA according to the judgement of the attending neurologist and were therefore excluded from follow-up analysis of WUS group along with other 9 WUS patients with missing follow-up imaging. Demographic features, vascular risk factors, and stroke etiology of WUS patients and standard rt-PA–treated patients were similar. Stroke severity of WUS patients (NIHSS mean 10.50; median 7.5) was similar to standard rt-PA–treated patients (NIHSS mean 11.86; median 11; P 5 .110; Table 1). Baseline EICs were similar in both groups (median ASPECTS 9; P 5 .178). The distribution of EICs at baseline CT scan was also comparable between the 2 groups (P 5 .384; Table 2). The extent of ischemic lesions at follow-up imaging was equivalent in both groups (median ASPECTS 7.0 in WUS patients; 7.5 in rt-PA patients; P 5 .202). Follow-up ASPECTS differed from baseline ASPECTS in 76.2% of WUS patients and 73.3% of standard rt-PA–treated patients (P 5 .722). Functional independence (mRS score #2) was reached in 61.6% of standard rt-PA–treated patients compared to 43.1% of WUS patients (odds ratio [OR] 2.12; 95% confidence interval [CI] 1.05-4.28; P 5 .037). In univariate analysis, functional outcome

Table 1. Baseline clinical and imaging characteristics of patients with wake-up stroke and patients treated with standard recombinant tissue plasminogen activator

Age, y (mean 6 SD) Gender, male, n (%) NIHSS score, median (IQR) Smoking, n (%) Arterial hypertension, n (%) Coronary heart disease, n (%) Diabetes mellitus, n (%) Dyslipidaemia, n (%) Previous stroke history, n (%) TOAST classification, n (%) Cardioembolic Large artery Small vessel Unknown Time from symptoms to CT scan, min (mean 6 SD) Baseline ASPECTS (median IQR)

WUS (n 5 60)

Standard rt-PA (n 5 86)

P value

68.75 (11.37) 33 (55.0) 7.5 (4-18.5) 12 (20.7) 36 (60.0) 24 (40.0) 14 (23.3) 19 (31.7) 14 (23.3)

65.64 (12.42) 48 (55.8) 11.0 (7-17) 21 (25.9) 55 (64.0) 45 (52.3) 10 (11.6) 30 (34.9) 21 (24.4)

.125 .922 .110 .474 .628 .142 .060 .685 .880 .445

26 (43.3) 18 (30.0) 9 (15.0) 7 (11.7) 127.32 (52.68) 9 (7-10)

40 (46.5) 30 (34.9) 12 (14.0) 4 (4.7) 102.42 (38.85) 9 (8-10)

0.002 0.178

Abbreviations: ASPECTS, Alberta Stroke Program Early Computed Tomography Score; CT, computed tomography; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; rt-PA, recombinant tissue plasminogen activator; SD, standard deviation; TOAST, Trial of Org 10172 in Acute Stroke Treatment; WUS, wake-up stroke.

L. ROVERI ET AL.

706

Table 2. Distribution of ASPECTS grading at baseline computed tomographic scan

WUS (n 5 60) Standard rt-PA (n 5 86)

ASPECTS 5 10, n (%)

ASPECTS 5 8-9, n (%)

ASPECTS 5 0-7, n (%)

P value

26 (43.4) 42 (48.8)

17 (28.3) 28 (32.6)

17 (28.3) 16 (18.6)

.384 —

Abbreviations: ASPECTS, Alberta Stroke Program Early Computed Tomography Score; rt-PA, recombinant tissue plasminogen activator; WUS, wake-up stroke.

was significantly associated with treatment, NIHSS score, and EICs in less than one-third of MCA territory, equivalent to ASPECTS .7 (Table 3). After multivariate logistic regression analysis, the two parameters that significantly predicted the functional outcome were treatment (independently associated with favorable outcome [OR 7.76; 95% CI 2.40-25.05; P 5 .001]) and NIHSS score (inversely associated with favorable outcome (OR 0.74; 95% CI 0.67-0.82; P , .001; Table 4). The proportion of patients with normal baseline CT scan (ASPECTS 10) was similar in WUS (39.2%) and standard rt-PA–treated patients (48.8%; P 5 .274) as well as their rate of subsequent positive follow-up imaging (76.5% and 64.3%, respectively; P 5 .364). ICH occurred in 15.0% of standard rt-PA–treated stroke patients, 3 of whom were symptomatic (3.5%), and in 18.6% of WUS patients, all asymptomatic (P 5 .620). Mortality was 8.1% in standard rt-PA–treated patients and 11.8% in WUS patients (P 5 .484). In addition, we looked at the small group of 9 WUS patients treated with rt-PA. Clinical and imaging features were comparable to standard rt-PA and nontreated WUS except for younger age of rt-PA–treated WUS relative to nontreated WUS (Table 5). Symptoms to baseline imaging time was 98 minutes for rt-PA–treated WUS (95% CI 59.36136.64). Symptoms to needle time of rt-PA–treated WUS patients was 188.22 minutes (95% CI 143.30-221.15) compared to 162.08 minutes (95% CI 156.12-168.05) of standard rt-PA–treated patients (P 5 .061). Favorable outTable 3. Univariate analysis for identification of parameters associated with favorable outcome

Univariate variable

Favorable outcome (n 5 137), OR (95% CI)

P value (P , .1)

Thrombolysis NIHSS Age ASPECTS .7 Diabetes mellitus Coronary heart disease

2.12 (1.05-4.28) 0.79 (0.73-0.85) 0.98 (0.95-1.01) 2.57 (1.11-5.94) 1.12 (0.44-2.87) 0.93 (0.48-1.83)

.037 ,.001 .127 .024 .810 .841

Abbreviations: ASPECTS, Alberta Stroke Program Early Computed Tomography Score; CI, confidence interval; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio. Data are calculated as continuous or categorized variables (see Table 1). Significant values are shown in bold.

come was observed in 55.6% of rt-PA–treated WUS patients. No symptomatic ICH, 2 asymptomatic ICHs, and no deaths occurred in the rt-PA–treated WUS group.

Discussion In the present study WUS patients accounted for 12.4% of all ischemic stroke patients admitted to our stroke centre. Around a third of WUS could have been appropriate candidate for thrombolysis if stroke onset modality were not a factor. Very recently a population-based study found similar figures evaluating the proportion of WUS and how many patients would have been eligible for thrombolisys if time were not a factor.7 Our study was focused on the clinical and imaging features of WUS patients who arrived within 3 hours of symptom recognition compared to stroke patients within 3 hours of symptom onset treated with rt-PA. We did not find significant differences in clinical and demographic features between the two groups. In addition we evaluated baseline and follow-up CT scans for ischemic changes graded with ASPECTS. In WUS within 3 hours of symptom recognition, the degree of EICs was similar to standard rt-PA–treated stroke. In addition, the extent of final infarct graded with ASPECTS on follow-up CT-scan was comparable between WUS patients and standard rt-PA–treated patients. As expected, standard rt-PA–treated patients had a significantly higher rate of favorable outcome (18.5%). The similar degree of EICs in both groups may be relevant to the assumption that stroke onset may occur close to or upon awakening. Moreover, the observed similar pattern of evolution of Table 4. Multivariate regression analysis for outcome

Multivariate

Favorable outcome (n 5 137), OR (95% CI)

P value (P , .05)

Thrombolysis NIHSS Age ASPECTS .7

7.76 (2.40-25.05) 0.74 (0.67-0.82) 1.01 (0.97-1.05) 0.99 (0.29-3.35)

.001 ,.001 .762 .988

Abbreviations: ASPECTS, Alberta Stroke Program Early Computed Tomography Score; CI, confidence interval; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio. Variables that remained independently related to favorable clinical outcome (P , .05) are shown in bold.

WAKE-UP STROKE WITHIN 3 HOURS OF SYMPTOMS AWARENESS

707

Table 5. Comparison of clinical and neuroradiologic features of treatment groups

Age, y (mean 6 SD) NIHSS score, median (IQR) Time from symptoms to CT scan, min (mean 6 SD) Baseline ASPECTS, median (IQR) Time from symptoms to treatment, min (mean 6 SD) Favorable outcome, n (%) sICH (%) aICH (%) Mortality

Nontreated WUS (n 5 51)

P value

rt-PA–treated WUS (n 5 9)

P value

Standard rt-PA (n 5 86)

70.43 (10.28) 7 (4-18) 132.81 (51.78)

.005 .860 .068

59.22 (13.13) 9 (4-19) 98.00 (50.26)

.146 .492 .755

65.64 (12.42) 11.0 (7-17) 102.42 (38.85)

9 (7-10) —

.340 —

10 (7-10) 188.22 (50.64)

.691 .061

9 (8-10) 162.08 (27.49)

22 (43.1) 0 6 (11.8) 6 (11.8)

.490 — .596 .278

.722 .569 .792 .374

53 (61.6) 3 (3.5%) 16 (18.6) 7 (8.1%)

5 (55.6) 0 2 (22.2) 0

Abbreviations: aICH, asymptomatic intracranial hemorrhage; ASPECTS, Alberta Stroke Program Early Computed Tomography Score; CT, computed tomography; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; rt-PA, recombinant tissue plasminogen activator; SD, standard deviation; sICH, symptomatic intracranial haemorrhage; WUS, wake-up stroke.

EICs to ischemic lesions at follow-up CT scan found in our study might be taken as indirect evidence that WUS occurred close to awakening. Only one previous study compared 28 patients with WUS to 68 randomly selected strokes with known onset using ASPECTS to score EICs.8 The authors reported a significantly lower ASPECTS in patients with WUS who arrived at the hospital between 4:00 AM and 10:00 AM who underwent baseline CT scan within 15 hours from last seen normal. However, the difference disappeared when ASPECTS was dichotomized at values #7. The significance of EICs in thrombolytic treatment decision is a matter of debate, because there is no definitive evidence that thrombolysis is ineffective or harmful even when EICs are present.16-20 The results of our study revealed that rt-PA and NIHSS score were significant predictors of outcome. In agreement with previous reports, this finding might suggest that EICs do not appear to warn against rt-PA treatment. The increased likelihood of favorable response to rt-PA in patients with minor EICs deserves additional investigation with advanced imaging, because the selection of patients based only on the time of stroke onset might exclude the patients who could benefit from thrombolysis. Although a standard approach with advanced imaging has yet to be established, penumbra imaging with CT perfusion or MR perfusion has a major potential to improve patient selection as well as to address the debate regarding which reperfusion approach is more appropriate. In addition, assessing the state of the ischemic brain and collateral circulation may help to explain worse clinical outcome in patients with recanalization but not reperfusion. Refining selection criteria with advanced imaging may prevent to treat patients for whom recanalization is futile or perhaps even harmful.21 One of the imaging indicators of potential benefit from thrombolysis is the MRI mismatch pattern. One study reported a mismatch in 73% of patients with WUS presenting within 3 hours of symptoms aware-

ness.22 More recently, patients with WUS and known onset stroke patients were found to have similar ischemic lesion volumes and frequency of mismatch evaluated with CT perfusion up to 12 hours.10 Much less is known about tissue outcome after thrombolysis. This issue has been addressed by the analysis of CT measurements from the National Institutes of Neurological Disorders and Stroke rt-PA stroke trial23 that found only a trend toward a reduction in median lesion volume. The Echoplanar Imaging Thrombolytic Evaluation Trial24 reported that rt-PA was not associated with lower infarct growth but with increased reperfusion measured with MRI. We found similar tissue outcome in WUS and standard rt-PA–treated patients, although this was measured with ASPECTS. Even though functional outcome is the real goal of thrombolysis, additional investigations to better understand imaging biomarkers of tissue outcome are also needed. A recent study reported similar rates of favourable outcome in rt-PA–treated WUS patients compared to standard rt-PA–treated patients and a higher rate of favorable outcome compared to nontreated WUS patients. Considering the safety profile, rt-PA–treated WUS patients had a rate of sICH comparable to standard rt-PA–treated stroke patients.25 We found similar rates of favorable outcomes in rt-PA–treated WUS patients and standard rt-PA–treated patients and a 12.5% higher rate of favorable outcome in rt-PA–treated WUS compared to nontreated WUS, although not statistically significant. No sICH occurred in WUS patients. Although the nonrandomized design and the limited number of patients calls for caution in drawing conclusions, our results may indicate a positive signal that needs to be replicated in a larger randomized trial. In the literature, the only randomized trial treating WUS patients based on CT scan is the Abciximab in Emergency Stroke Treatment Trial-II (AbESTT-II),26 which enrolled a total of 43 patients

L. ROVERI ET AL.

708

with WUS before being stopped because of an increased rate of sICH. Aside from the different thrombolytic agent used, the disappointing results of AbESTT-II with WUS patients might possibly be related to the lower rate of normal baseline CT scan in the treatment arm of the WUS cohort; in addition, the amount of EICs was not quantified. Despite limitations such as nonrandomized comparisons, relatively small sample size, and first-level imaging, the findings of our study add a piece of evidence that in the 3-hours time window, patients with WUS and standard rt-PA–treated stroke patients are comparable. Although multimodal neuroimaging techniques can improve the selection of WUS patients compared with noncontrast CT scan and may became the method of choice, they need to be implemented in emergency conditions without taking extra time before thrombolysis. Most physicians who treat stroke currently depend on CT because it is more accessible. Our study reflects real clinical conditions in a high-volume stroke center. In accordance with previous reports, our study, based on routine clinical practice, supports the concept that selected WUS patients may be appropriate candidates for thrombolysis. Additional prospective randomized studies are urgently needed to evaluate multimodal imaging parameters and to further assess the safety and outcome of thrombolysis in WUS patients. We are currently working on a proofof concept randomized trial to address the issues of safety and efficacy of rt-PA treatment in patients with WUS. Acknowledgment: statistical advice.

We thank Dr Mariapia Sormani for

References 1. Reeves MJ, Arora S, Broderick JP, et al. Acute stroke care in the US: Results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke Registry. Stroke 2005;36: 1232-1240. 2. Albers GW, Amarenco P, Easton JD, et al. Antithrombotic and thrombolytic therapy for ischemic stroke: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008;133(6 Suppl):630S-669S. 3. Chaturvedi S, Adams HP Jr, Woolson RF. Circadian variation in ischemic stroke subtypes. Stroke 1999;30: 1792-1795. 4. Elliott WJ. Circadian variation in the timing of stroke onset: A meta-analysis. Stroke 1998;29:992-996. 5. Lago A, Geffner D, Tembl J, et al. Circadian variation in acute ischemic stroke: A hospital-based study. Stroke 1998;29:1873-1875. 6. Marler JR, Price TR, Clark GL, et al. Morning increase in onset of ischemic stroke. Stroke 1989;20:473-476. 7. Mackey J, Kleindorfer D, Sucharew H, et al. Populationbased study of wake-up strokes. Neurology 2011; 76:1662-1667. 8. Huisa BN, Raman R, Ernstrom K, et al. Alberta Stroke Program Early CT Score (ASPECTS) in patients with wake-up stroke. J Stroke Cerebrovasc Dis 2010;19:475-479.

9. Serena J, Davalos A, Segura T, et al. Stroke on awakening: Looking for a more rational management. Cerebrovasc Dis 2003;16:128-133. 10. Silva GS, Lima FO, Camargo EC, et al. Wake-up stroke: Clinical and neuroimaging characteristics. Cerebrovasc Dis 2010;29:336-342. 11. Todo K, Moriwaki H, Saito K, et al. Early CT findings in unknown-onset and wake-up strokes. Cerebrovasc Dis 2006;21:367-371. 12. Barber PA, Demchuk AM, Zhang J, et al. Validity and reliability of a quantitative computed tomography score in predicting outcome of hyperacute stroke before thrombolytic therapy. ASPECTS Study Group. Alberta Stroke Programme Early CT Score. Lancet 2000; 355:1670-1674. 13. Hirano T, Yonehara T, Inatomi Y, et al. Presence of early ischemic changes on computed tomography depends on severity and the duration of hypoperfusion: A single photon emission-computed tomographic study. Stroke 2005;36:2601-2608. 14. Adams HP Jr, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24: 35-41. 15. Wahlgren N, Ahmed N, Davalos A, et al. Thrombolysis with alteplase for acute ischaemic stroke in the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST): An observational study. Lancet 2007;369:275-282. 16. Davis SM, Donnan GA. CT screening for thrombolysis: Uncertainties remain. Stroke 2003;34:822-823. 17. Demchuk AM, Hill MD, Barber PA, et al. Importance of early ischemic computed tomography changes using ASPECTS in NINDS rtPA Stroke Study. Stroke 2005; 36:2110-2115. 18. Lyden P. Early major ischemic changes on computed tomography should not preclude use of tissue plasminogen activator. Stroke 2003;34:821-822. 19. Patel SC, Levine SR, Tilley BC, et al. Lack of clinical significance of early ischemic changes on computed tomography in acute stroke. JAMA 2001;286:2830-2838. 20. von Kummer R. Early major ischemic changes on computed tomography should preclude use of tissue plasminogen activator. Stroke 2003;34:820-821. 21. Wardlaw JM. Neuroimaging in acute ischaemic stroke: insights into unanswered questions of pathophysiology. J Intern Med 2010;267:172-190. 22. Fink JN, Kumar S, Horkan C, et al. The stroke patient who woke up: Clinical and radiological features, including diffusion and perfusion MRI. Stroke 2002;33:988-993. 23. Effect of intravenous recombinant tissue plasminogen activator on ischemic stroke lesion size measured by computed tomography. NINDS; The National Institute of Neurological Disorders and Stroke (NINDS) rt-PA Stroke Study Group. Stroke 2000;31:2912-2919. 24. Davis SM, Donnan GA, Parsons MW, et al. Effects of alteplase beyond 3 h after stroke in the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET): A placebo-controlled randomised trial. Lancet Neurol 2008;7:299-309. 25. Barreto AD, Martin-Schild S, Hallevi H, et al. Thrombolytic therapy for patients who wake-up with stroke. Stroke 2009;40:827-832. 26. Adams HP Jr, Leira EC, Torner JC, et al. Treating patients with ‘wake-up’ stroke: The experience of the AbESTT-II trial. Stroke 2008;39:3277-3282.