Cerebral Computed Tomography-Graded White Matter Lesions Are Associated With Worse Outcome After Thrombolysis in Patients With Stroke

Transcription

1 Cerebral Computed Tomography-Graded White Matter Lesions Are Associated With Worse Outcome After Thrombolysis in Patients With Stroke Sami Curtze, MD, PhD; Susanna Melkas, MD, PhD; Gerli Sibolt, MD; Elena Haapaniemi, MD, PhD; Satu Mustanoja, MD, PhD; Jukka Putaala, MD, PhD; Tiina Sairanen, MD, PhD; Marjaana Tiainen, MD, PhD; Turgut Tatlisumak, MD, PhD; Daniel Strbian, MD, PhD Background and Purpose Compared with other stroke causes, small-vessel disease is associated with better 3-month outcomes in patients with acute ischemic stroke treated with intravenous thrombolysis. Another question is the impact of coexisting cerebral white matter lesions (WMLs; a surrogate marker of small-vessel disease) on outcome, which was addressed in the current study. Methods We analyzed 2485 consecutive intravenous thrombolysis treated patients at the Helsinki University Central Hospital, 2001 to WMLs were scored according to 4 previously published computed tomographic visual rating scales from all baseline head scans. The inter-rater agreement was calculated. The primary outcome measure was shift analysis, and the secondary examined all possible binary cutoffs in the modified Rankin Scale at 3 months. The associations of modified Rankin Scale with nominal, ordinal, and continuous variables were analyzed in univariate and adjusted in multivariate binary and ordinal regression (shift analysis) models. Results In univariate and multivariate regression analyses, all 4 tested visual WML rating scales (as continuous variables, or dichotomized at different cutoff points) were associated with worse outcome at all binary levels and in shift analyses of the modified Rankin Scale. After adjusting for confounders, the statistically strongest association in shift analyses remained for the Blennow scale dichotomized at >3 points, reflecting at least moderate WMLs (odds ratio, 1.90; 95% confidence interval, ). Conclusions WMLs on admission computed tomographic scan are independently associated with worse outcome in intravenous thrombolysis treated patients with stroke. (Stroke. 2015;46: DOI: /STROKEAHA ) Reduced areas of x-ray attenuation on computed tomography (CT) are frequently seen on brain images of patients with stroke representing leukoaraiosis. 1 As proposed by the STandards for ReportIng Vascular changes on neuroimaging position paper, the term cerebral white matter lesions (WMLs) of presumed vascular origin will be used here as a surrogate marker of small-vessel disease (SVD). 1,2 Compared with other stroke causes, SVD relates to better outcome at 3 months in patients with acute ischemic stroke treated with intravenous thrombolysis (IVT). 3,4 However, in long term, SVD and WMLs as its surrogate marker are known to predict functional decline, morbidity, and death in independent outpatients and also in patients with stroke. 5 8 An increased Key Words: outcome measures thrombolytic therapy risk of worse outcome with increased WMLs was shown at 6 months already. 9 SVD coexists frequently with embolic and large artery stroke cause. 10 Coexisting WMLs in patients with large artery anterior circulation strokes relate to worse outcomes at 3 months. 11 In IVT-treated patients with acute stroke, the coexistence of WMLs is associated with worse outcome in 3 studies ; however, a lack of significance after multivariable adjustments has been found in the largest cohort to date. 15 We aimed to investigate the 3-month outcomes in the presence of WMLs in a large single-center cohort of IVT-treated patients with ischemic stroke. A second objective, to compare the predictive value and inter-rater agreement between 4 previously published CT visual rating scales for WMLs. Received January 26, 2015; final revision received March 3, 2015; accepted March 24, From the Department of Neurology, Helsinki University Central Hospital and Department of Neurological Sciences, University of Helsinki, Helsinki, Finland. Presented in part at the European Stroke Organization Conference, Glasgow, Scotland, April 17 19, The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Sami Curtze, MD, PhD, Department of Neurology, Helsinki University Central Hospital, PO Box 340, FI Helsinki, Finland. sami.curtze@hus.fi 2015 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 Curtze et al White Matter Lesions and Thrombolysis Outcome 1555 Methods Our cohort includes 2485 consecutive patients treated with IVT between December 2001 and February 2014 at the Helsinki University Central Hospital, which is the only comprehensive stroke center in the region of Helsinki, Finland, serving a population of 1.7 million. 16 Ethical review for retrospective analysis of data collected prospectively as a part of routine clinical care is not required at our institution. Informed consent was obtained from all patients (or next of kin) treated in a trial setting in the period 2001 to 2002; thereafter, IVT for acute ischemic stroke became a routine treatment, and no consent was required. Patients were treated based on our department s written guidelines for acute stroke, which are updated biannually and whenever new scientific evidence becomes available. 16 The primary outcome measure was clinical outcome at 3 months assessed by the modified Rankin Scale (mrs). 17 Pretreatment National Institutes of Health Stroke Scale scores and mrs were assessed by a certified and video-trained stroke neurologist. The baseline data on hypertension, diabetes mellitus, atrial fibrillation, hyperlipidemia, coronary artery disease, and congestive heart failure refer to the condition before the index stroke. Admission head scans before IVT were available for the whole cohort (3 magnetic resonance imaging [MRI] and 2482 CT). Before administration of IVT, a stroke neurologist and a radiologist interpreted brain CT scans, focusing on exclusion of other causes, the presence of early infarct signs, and the presence of a hyperdense cerebral artery sign. A follow-up scan was routinely performed at 24 hours post IVT either with CT or with MRI and whenever a hemorrhage was suspected. 18 Intracranial hemorrhage was considered symptomatic intracerebral hemorrhage (sich) applying the European Cooperative Acute Stroke Study II (ECASS-2) criteria for sich. 18,19 Adjudication of whether an sich had occurred, was performed by experienced stroke neurologists independently from the WML evaluation. A literature review was performed to identify existing CT visual rating scales for WMLs. Six previously used CT visual rating scales were identified. We found the rating scale of Rezek with scores from 0 to 180 inappropriate for use in acute settings of IVT candidates. 20 The van Swieten scale has been used before in a modified version scoring both hemispheres separately. The modified van Swieten scale was initially calculated for this publication, but omitted because the results were redundant to the original van Swieten scale. 15,17 Eight experienced stroke neurologists were trained to score WMLs according to 4 previously published CT visual rating scales from all baseline CT head scans in a blinded fashion without knowledge of patients clinical data or outcome. 1. The Gorter scale: White matter abnormalities were rated in 2 grades: in grade 1 (moderate leukoaraiosis), the hypodensity of the white matter on CT was restricted to the region adjacent to the ventricles; in grade 2 (severe leukoaraiosis), the hypodensity extended as far as the cortex The van Swieten scale: The following 3 subsequent CT slices are rated toward severity of the hypodensity: Through the choroid plexus of the posterior horns, through the cella media, and through the centrum semiovale. The severity reaches from not present (0), over abnormality restricted to the region adjoining the ventricles, 1 to increased hypodensity involving the entire region from lateral ventricle to the cortex, 2 and is scored according to the more affected slice. Anterior and posterior regions were rated separately giving overall scores from 0 to The Blennow rating scale: WMLs were defined as areas of decreased attenuation, with periventricular distribution. The extension and intensity of WMLs were rated independently of each other. The extension from normal (0), over decreased attenuation of white matter at the margins of the frontal and occipital horns, 1 over decreased attenuation of white matter around the frontal and occipital horns of the lateral ventricle, with some extension toward the semioval center, 2 to decreased attenuation around the lateral ventricles and coalescing in the semioval center. 3 The intensity was graded from no (0), over mild, 1 moderate, 2 to marked. 3,23 4. The Wahlund rating scale: Five anatomic areas (frontal, parietooccipital, temporal, infratentorial, and basal ganglia) in each hemisphere are rated separately on WMLs from no lesions (0), over focal lesions, 1 over beginning confluence of lesions, 2 to diffuse involvement of the entire region. 3,24 The inter-rater agreement of the infratentorial, temporal, and basal ganglia sites is known to be less than substantial in CT scans; therefore, the score will be truncated to frontal and parieto-occipital scores. 24 To determine the inter-rater agreement, the brain CT scans of 50 patients were rated by all 8 raters independently. Statistical Analyses The inter-rater agreement for all 4 used visual rating scales was calculated as a single measure and average measures interclass correlation for consistency in a 2-way mixed model with fully crossed design. The associations of the mrs at 3 months with categorical, ordinal, and continuous variables were analyzed in a univariate ordinal logistic regression as suggested for stroke outcomes. 25 To investigate whether WMLs were independently associated with mrs at 3 months, we created a multivariate ordinal logistic regression model based on the predictors of outcome known from our cohort. 26 Proportional odds assumptions were tested and met. The visual WML rating scores were separately added to each model as continuous variables and as dichotomized ones with different cutoffs. Results are expressed as adjusted odds ratio (OR) and corresponding 95% confidence intervals (CI). For all other analyses, statistical significance was set at P<0.05. All statistical analyses were performed with SPSS Statistics 22 for Linux (IBM Corp, Armonk, NY). Results In the whole cohort of 2485 patients, the mrs at 3 months was not available for 34 patients (1.4%). Baseline characteristics of the study population stratified for patients with a Blennow cerebral white matter rating scale of 0 to 3 and 4 to 6 are reported in Table 1. Additional endovascular treatment after IVT was performed in 153 patients (6.2%). The interclass correlation of raters indicated at least substantial agreement (defined as ) for all 4 CT visual rating scales for WMLs (Table 2). Almost perfect or perfect agreement was reached for the van Swieten and for the Blennow scores. In addition, the Wahlund score reached almost perfect agreement at frontal and parieto-occipital sites, but only fair agreement (defined as ) for the other sites (data not shown). The mrs categories at 3 months according to the tested visual WML rating scales are illustrated in the Figure. In univariate analyses, all tested visual WML rating scales as continuous variables, and as well as dichotomized at different cutoff points, were associated with increased risk of higher mrs scores, representing worse outcome (Table 3). We found the highest binary association (OR, 2.66; 95% CI, ) with the Blennow scale with >3 points, reflecting at least moderate WMLs. All tested binary cutoff ORs ranged from 1.96 to After adjustments for confounders (age, onset-to-treatment time, baseline National Institutes of Health Stroke Scale, hyperdense artery sign on admission imaging, early infarct signs on admission imaging, glucose on admission, and prestroke mrs>1) in multivariable ordinal regression analyses, all 4 visual WML rating scales as continuous variables, and as well as dichotomized at different cutoff points, were

3 1556 Stroke June 2015 Table 1. Differences in Baseline Characteristics Between Patients With a Blennow Cerebral White Matter Rating Scale Cutoff Score of 3, n=2485 Blennow 0 2 (n=2004) associated with increased risk of higher mrs scores (Table 3). All confounders remained independent predictors after adjustment. After adjusting for confounders, the statistically strongest association remained for the Blennow scale with >3 points, reflecting at least moderate WMLs (OR, 1.90; 95% CI, ). Removing all sich cases resulted in an OR of 1.76 (95% CI, ). The OR for all dichotomized scores was in the range of 1.43 to In multivariate shift models including the raw WMLs rating score points as categorical variables, the Blennow score of 6 had the statistically strongest association with increasing mrs scores (OR, 3.59; 95% CI, ). Removing all sich cases resulted in an OR of 2.37 (95% CI, ). Exclusion of those with basilar Blennow 3 6 (n=481) OR (CI) P Value Age, y, median (IQR) 67 (58 75) 78 (72 83) 1.10 ( ) <0.001* Age, y < % 11.2% % 47.8% 4.05 ( ) <0.001* % 41.0% ( ) <0.001* Female sex, % 41.8% 49.5% 1.37 ( ) 0.002* OTT, min, median (IQR) 120 (85 170) 120 (88 167) 1.00 ( ) OTT 90 min, % ( ) Baseline NIHSS, median (IQR) 8 (5 14) 10 (6 16) 1.03 ( ) <0.001* mrs >1, prestroke 4.30% 11.00% 2.76 ( ) <0.001* Admission imaging, % Hyperdense artery sign ( ) Early infarct signs ( ) Glucose on admission >8.0 mmol/l, % ( ) <0.001* Glucose on admission, mmol/l, 6.5 ( ) 6.9 ( ) 1.53 ( ) <0.001* median (IQR) INR on admission >1.7, % ( ) BP over 185/110 mm Hg before ( ) thrombolysis, % Previous medication, % Antihypertensive ( ) <0.001* Antithrombotic ( ) 0.029* Statin ( ) 0.001* Anticoagulation ( ) 0.030* Medical history, % Hyperlipidemia ( ) Hypertension ( ) <0.001* Atrial fibrillation ( ) 0.001* Myocardial infarction ( ) Coronary disease ( ) <0.001* Diabetes mellitus ( ) <0.001* Previous ischemic stroke ( ) <0.001* Univariate binary logistic regression with OR and the 95% CI. BP indicates blood pressure; CI, confidence interval; INR, international normalized ratio; IQR, interquartile range; mrs, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; and OTT, onset-to-treatment time. *Indicates P<0.05. artery occlusion, additional endovascular procedures, or both, did not affect our main results (data not shown). To allow comparison with previous literature, we additionally calculated the association of the van Swieten score >0 with dichotomized mrs 2 to 6 (OR, 1.59; 95% CI, ) and 3 to 6 (OR, 1.45; 95% CI, ). 14 To illustrate the proportional odds for univariate (Table I in the online-only Data Supplement) and multivariate (Table II in the online-only Data Supplement) ordinal logistic regression, ORs are presented for all tested visual WML rating scales for each cutoff on the mrs. Sensitivity analyses were performed by splitting the cohort at a cutoff date on December 2009 resulting in 2 cohorts with 1242 and 1243

4 Curtze et al White Matter Lesions and Thrombolysis Outcome 1557 Table 2. Comparison of the Different Visual Rating Scales for Cerebral WMLs on Computed Tomography SM ICC (CI) AM ICC (CI) Any WMLs 0.67 ( ) 0.95 ( ) Gorter WMLs 0.78 ( ) 0.97 ( ) Gorter WMLs ( ) 0.96 ( ) van Swieten anterior 0.76 ( ) 0.96 ( ) van Swieten posterior 0.77 ( ) 0.96 ( ) van Swieten 0.82 ( ) 0.97 ( ) van Swieten 3 or ( ) 0.95 ( ) (severe) Blennow extension 0.82 ( ) 0.97 ( ) Blennow intension 0.76 ( ) 0.96 ( ) Blennow 0.81 ( ) 0.97 ( ) Blennow > ( ) 0.94 ( ) Blennow > ( ) 0.94 ( ) Blennow > ( ) 0.94 ( ) Wahlund frontal left (0 3) 0.82 ( ) 0.97 ( ) Wahlund frontal right 0.82 ( ) 0.97 ( ) (0 3) Wahlund parieto-occipital 0.82 ( ) 0.97 ( ) left (0 3) Wahlund Parieto-occipital 0.78 ( ) 0.97 ( ) right (0 3) Wahlund sum (0 12) 0.87 ( ) 0.98 ( ) Wahlund any WMLs 0.69 ( ) 0.95 ( ) Wahlund any site score > ( ) 0.95 ( ) SM and AM ICC for consistency in a 2-way mixed model with fully crossed design of 50 patients and 8 raters. AM indicates average measures; CI, confidence interval; ICC, interclass correlation; SM, single measure; and WML, white matter lesion. patients, respectively. Additional sensitivity analyses were performed excluding patients with a preadmission mrs>1 and by adjusting multivariable analyses additionally for occurrence of any bleeding on control CT scan. All results of the present study were robust in the sensitivity analyses (data not shown). Discussion In the present cohort of patients with ischemic stroke treated with IVT, WMLs visible on baseline noncontrast CT scan were associated with a worse outcome 3 months after stroke. The inter-rater agreement for 3 visual rating scales of WMLs in the present study was substantial, and for 2 scales almost perfect. This is within the same range, or even slightly better than in previous CT or MRI comparisons. 24,27 Because of regional artifacts on CT, the Wahlund score produced expectedly, only fair inter-rater agreement for temporal, infratentorial, and the basal ganglia sites. 24 Our inter-rater agreement was slightly better than in the original Wahlund paper; however, because of the expectedly fair agreement for temporal, infratentorial, and the basal ganglia sites, we only used frontal and parieto-occipital sites for further analyses. 24 All confounders of outcome known from our cohort remained independent predictors of outcome in all multivariable models. The association of total National Institutes of Health Stroke Scale score, glucose on admission, early infarct signs, hyperdense artery signs, onset-to-treatment time, and prestroke mrs with outcome is in line with previous studies. 28 In univariate and multivariate models, all used WML visual rating scales showed association of increasing WML scores with a shift to higher mrs scores in ordinal analyses, and worse outcome for each possible dichotomization of the mrs scale. The ORs and CIs of all tested variables were stable and concordant over the whole analyzed ranges. These results are reassuring about the association between WMLs and patients outcome. Our findings also confirm that the presence of particularly moderate-to-severe WMLs on admission CT, otherwise IVTeligible patients correlate to worse outcome. This finding was consistent regardless of the used visual WML rating scale, which emphasizes that a true biological association exists between WMLs and poorer brain recovery in stroke. Previous research has shown that WMLs in ischemic stroke patients with, and without IVT treatment, 9,11,13 are associated with poorer prognosis that in those without WMLs. Three studies reported worse outcomes in ischemic stroke patients with WMLs after IVT in 400, , patients. 12 However, in 1 study of 800 IVT-treated patients, severe van Swieten scores 3,4 were only associated with poor outcome (mrs, 2 6) or death in univariate, but not after adjusted analyses, whereas there remained a significant association after adjustments in the present study. 15 This difference is probably because of sample size because the cutoff values for severe WMLs on van Swieten scale covered only 71 (8.6%) patients in the cited study, whereas there were 481 (18.8%) patients with severe WMLs in the present study. 15 Our single-center patient population alone is larger than all the 1564 patients of all previous studies together. Our dichotomized outcome measures at 3 months are in line with a previous study of 400 IVT-treated patients using the van Swieten score >0 as cutoff for WMLs. 14 In a third study comparing 200 IVT-treated with 203 non IVT-treated patients, the OR for mrs 2 to 6 is not presented but can be calculated as univariate logistic regression value as OR 1.21 (95% CI, ), which is in accordance with the present study (OR, 1.34; 95% CI, ). 13 In an MRI IVT cohort of 164 patients, any WMLs were associated with mrs 3 to 6 (OR, 3.22; 95% CI, ) after adjustments for confounders, this is in line with the present study (OR, 1.39; 95% CI, ). 12 Our study has limitations and strengths. One of the shortcomings include that a MRI cohort would have delivered more precise data on mild WMLs than on CT scans. 24,27 In addition, differentiating other causes such as old subcortical infarcts is more uncertain in CT than in MRI evaluation. 2 However, in our cohort, low scores on ordinal scales, indicating none or mild WMLs, had significantly better outcomes, suggesting that CT modality is sufficient to determine the clinically relevant extent of WMLs that translate into worse outcomes regardless of the used visual rating scale. Noncontrast CT scan remains the universally dominant imaging modality when selecting stroke thrombolysis candidates.

5 1558 Stroke June 2015 Figure. Outcome at 3 months on full range of modified Rankin Scale in 2451 patients with ischemic stroke treated with intravenous thrombolysis stratified by different visual rating scales for cerebral white matter lesions (WML). Strengths of our study include its large sample size. To the best of our knowledge, this cohort is the largest to date where WMLs have been analyzed in thrombolyzed patients. Furthermore, our sample represents all consecutive patients, currently covering 30% of all patients with ischemic stroke treated at our institution. Finally, we assessed 4 different WML visual rating scales and were able to compare their performance. Because removing all patients with sich from the analyses did not change the findings significantly, the worse

6 Curtze et al White Matter Lesions and Thrombolysis Outcome 1559 Table 3. Different Visual Rating Scales for Cerebral WMLs and Their Association With mrs After Thrombolysis for Ischemic Stroke, n=2451 OR (CI) for Scoring Higher on the mrs Univariate Multivariate Age categorical (<65, 1.53 ( )* 65 79, 80 y) OTT 90 min 1.51 ( )* Baseline NIHSS 2.27 ( )* categorical (0 4, 5 9, 10 15, >15) mrs>1, prestroke 2.19 ( )* Admission imaging Hyperdense artery sign 1.58 ( )* Early infarct signs 1.58 ( )* Glucose on admission 1.68 ( )* >8.0 mmol/l Wahlund any site 2.43 ( ) 1.70 ( ) score >1 Any WMLs 1.96 ( ) 1.43 ( ) Blennow 1.25 ( ) 1.16 ( ) Blennow score categories ( ) 1.29 ( ) ( ) 1.30 ( ) ( ) 1.82 ( ) ( ) 2.11 ( ) ( ) 2.59 ( ) Blennow > ( ) 1.66 ( ) Blennow > ( ) 1.82 ( ) Blennow > ( ) 1.90 ( ) Gorter WMLs 1.72 ( ) 1.40 ( ) Gorter score categories ( ) 1.39 ( ) ( ) 1.98 ( ) Gorter ( ) 1.66 ( ) van Swieten 1.34 ( ) 1.21 ( ) van Swieten categories ( ) 1.19 ( ) ( ) 1.40 ( ) ( ) 1.66 ( ) ( ) 2.23 ( ) van Swieten 3 or ( ) 1.77 ( ) (severe) Wahlund (0 12) 1.12 ( ) 1.08 ( ) Wahlund any WMLs 2.04 ( ) 1.54 ( ) Univariate and adjusted multivariate ordinal logistic regression with OR and the 95% CI for scoring higher (worse outcome) on the mrs. Reference category is indicated with OR=1. CI indicates confidence interval; mrs, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; OTT, onset-to-treatment time; and WML, white matter lesion. *The covariates are given only for the model including the Blennow scale dichotomized at 3 points (Blennow >3). outcome of patients with WMLs is not only because of bleeding complications of IVT-treated patients. The reason for patients with higher grades of WMLs to have poorer recovery compared with patients with lower WML load cannot be answered by our study. However, WMLs relate to dementia, 29 depression, 30 death, 29 hip-fractures, 7 and recurrent stroke, 31 which all support the idea that the underlying SVD is a progressing disease interfering with recovery. Furthermore, the association of worse outcome with WMLs has been shown as well for patients with acute stroke not treated with IVT, 9,11 supporting the idea that the worse outcome is not related, or only partially related, to the IVT treatment. Because WMLs were associated with higher baseline National Institutes of Health Stroke Scale scores another contributing mechanism for our results could be that WMLs are associated with larger initial infarct volumes translating to worse outcome. 32,33 The overall findings of our study can be expected to extend to patients with endovasculary-treated stroke, as a recent study indicated. 34 Conclusions WMLs visible on admission CT scan were an independent predictor of worse outcome in our patients with IVT-treated stroke. Sources of Funding Helsinki University Central Hospital governmental subsidiary funds for clinical research, the Finnish Medical Foundation, and the Finnish Academy. None. Disclosures References 1. Hachinski VC, Potter P, Merskey H. Leuko-araiosis. Arch Neurol. 1987;44: Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, Fazekas F, Frayne R, et al; STandards for ReportIng Vascular changes on neuroimaging (STRIVE v1). Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol. 2013;12: doi: /S (13) Mustanoja S, Meretoja A, Putaala J, Viitanen V, Curtze S, Atula S, et al; Helsinki Stroke Thrombolysis Registry Group. Outcome by stroke etiology in patients receiving thrombolytic treatment: descriptive subtype analysis. Stroke. 2011;42: doi: / STROKEAHA Pantoni L, Fierini F, Poggesi A. Thrombolysis in acute stroke patients with cerebral small vessel disease. Cerebrovasc Dis. 2014;37:5 13. doi: / Oksala NK, Oksala A, Pohjasvaara T, Vataja R, Kaste M, Karhunen PJ, et al. Age related white matter changes predict stroke death in long term follow-up. J Neurol Neurosurg Psychiatry. 2009;80: doi: /jnnp Inzitari D, Pracucci G, Poggesi A, Carlucci G, Barkhof F, Chabriat H, et al; LADIS Study Group. Changes in white matter as determinant of global functional decline in older independent outpatients: three year follow-up of LADIS (leukoaraiosis and disability) study cohort. BMJ. 2009;339:b Sibolt G, Curtze S, Melkas S, Pohjasvaara T, Kaste M, Karhunen PJ, et al. White matter lesions are associated with hospital admissions because of hip-fractures and trauma after ischemic stroke. Stroke. 2014;45: doi: /STROKEAHA Smith EE. Leukoaraiosis and stroke. Stroke. 2010;41(10 suppl):s139 S143. doi: /STROKEAHA Arsava EM, Rahman R, Rosand J, Lu J, Smith EE, Rost NS, et al. Severity of leukoaraiosis correlates with clinical outcome after

7 1560 Stroke June 2015 ischemic stroke. Neurology. 2009;72: doi: / WNL.0b013e3181a Fu JH, Chen YK, Chen XY, Mok V, Wong KS. Coexisting small vessel disease predicts poor long-term outcome in stroke patients with intracranial large artery atherosclerosis. Cerebrovasc Dis. 2010;30: doi: / Henninger N, Lin E, Baker SP, Wakhloo AK, Takhtani D, Moonis M. Leukoaraiosis predicts poor 90-day outcome after acute large cerebral artery occlusion. Cerebrovasc Dis. 2012;33: doi: / Choi JH, Bae HJ, Cha JK. Leukoaraiosis on magnetic resonance imaging is related to long-term poor functional outcome after thrombolysis in acute ischemic stroke. J Korean Neurosurg Soc. 2011;50: doi: /jkns Demchuk AM, Khan F, Hill MD, Barber PA, Silver B, Patel S, et al; NINDS rt-pa Stroke Study Group. Importance of leukoaraiosis on CT for tissue plasminogen activator decision making: evaluation of the NINDS rt-pa Stroke Study. Cerebrovasc Dis. 2008;26: doi: / Ariës MJ, Uyttenboogaart M, Vroomen PC, De Keyser J, Luijckx GJ. tpa treatment for acute ischaemic stroke in patients with leukoaraiosis. Eur J Neurol. 2010;17: doi: /j x. 15. Palumbo V, Boulanger JM, Hill MD, Inzitari D, Buchan AM; CASES Investigators. Leukoaraiosis and intracerebral hemorrhage after thrombolysis in acute stroke. Neurology. 2007;68: doi: /01. wnl Meretoja A, Putaala J, Tatlisumak T, Atula S, Artto V, Curtze S, et al. Off-label thrombolysis is not associated with poor outcome in patients with stroke. Stroke. 2010;41: doi: / STROKEAHA van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke. 1988;19: Strbian D, Engelter S, Michel P, Meretoja A, Sekoranja L, Ahlhelm FJ, et al. Symptomatic intracranial hemorrhage after stroke thrombolysis: the SEDAN score. Ann Neurol. 2012;71: doi: /ana Hacke W, Kaste M, Fieschi C, von Kummer R, Davalos A, Meier D, et al. Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second European-Australasian Acute Stroke Study Investigators. Lancet. 1998;352: Rezek DL, Morris JC, Fulling KH, Gado MH. Periventricular white matter lucencies in senile dementia of the Alzheimer type and in normal aging. Neurology. 1987;37: Gorter JW. Major bleeding during anticoagulation after cerebral ischemia: patterns and risk factors. Stroke Prevention In Reversible Ischemia Trial (SPIRIT). European Atrial Fibrillation Trial (EAFT) study groups. Neurology. 1999;53: van Swieten JC, Hijdra A, Koudstaal PJ, van Gijn J. Grading white matter lesions on CT and MRI: a simple scale. J Neurol Neurosurg Psychiatry. 1990;53: Blennow K, Wallin A, Uhlemann C, Gottfries CG. White-matter lesions on CT in Alzheimer patients: relation to clinical symptomatology and vascular factors. Acta Neurol Scand. 1991;83: Wahlund LO, Barkhof F, Fazekas F, Bronge L, Augustin M, Sjögren M, et al; European Task Force on Age-Related White Matter Changes. A new rating scale for age-related white matter changes applicable to MRI and CT. Stroke. 2001;32: Optimising Analysis of Stroke Trials (OAST) Collaboration; Bath PMW, Gray LJ, Collier T, Pocock S, Carpenter J. Can we improve the statistical analysis of stroke trials? Statistical reanalysis of functional outcomes in stroke trials. Stroke. 2007;38: Strbian D, Meretoja A, Ahlhelm FJ, Pitkäniemi J, Lyrer P, Kaste M, et al. Predicting outcome of IV thrombolysis-treated ischemic stroke patients: the DRAGON score. Neurology. 2012;78: doi: / WNL.0b013e318245d2a Kapeller P, Barber R, Vermeulen RJ, Adèr H, Scheltens P, Freidl W, et al; European Task Force of Age Related White Matter Changes. Visual rating of age-related white matter changes on magnetic resonance imaging: scale comparison, interrater agreement, and correlations with quantitative measurements. Stroke. 2003;34: Strbian D, Seiffge DJ, Breuer L, Numminen H, Michel P, Meretoja A, et al. Validation of the DRAGON score in 12 stroke centers in anterior and posterior circulation. Stroke. 2013;44: doi: / STROKEAHA Debette S, Markus HS. The clinical importance of white matter hyperintensities on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ. 2010;341:c Sibolt G, Curtze S, Melkas S, Pohjasvaara T, Kaste M, Karhunen PJ, et al. Post-stroke depression and depression-executive dysfunction syndrome are associated with recurrence of ischaemic stroke. Cerebrovasc Dis. 2013;36: doi: / Melkas S, Sibolt G, Oksala NK, Putaala J, Pohjasvaara T, Kaste M, et al. Extensive white matter changes predict stroke recurrence up to 5 years after a first-ever ischemic stroke. Cerebrovasc Dis. 2012;34: doi: / Henninger N, Lin E, Haussen DC, Lehman LL, Takhtani D, Selim M, et al. Leukoaraiosis and sex predict the hyperacute ischemic core volume. Stroke. 2013;44: doi: /STROKEAHA Henninger N, Khan MA, Zhang J, Moonis M, Goddeau RP Jr. Leukoaraiosis predicts cortical infarct volume after distal middle cerebral artery occlusion. Stroke. 2014;45: doi: / STROKEAHA Zhang J, Puri AS, Khan MA, Goddeau RP Jr, Henninger N. Leukoaraiosis predicts a poor 90-day outcome after endovascular stroke therapy. AJNR Am J Neuroradiol. 2014;35: doi: /ajnr.A4029.