- Email: [email protected]
Validation of Serial Alberta Stroke Program Early CT Score as an Outcome Predictor in Thrombolyzed Stroke Patients
ARTICLE IN PRESS
Validation of Serial Alberta Stroke Program Early CT Score as an Outcome Predictor in Thrombolyzed Stroke Patients Wan-Yee Kong, MBBS,* Benjamin Y.Q. Tan, MBBS, MRCP,* Nicholas J.H. Ngiam, MBBS,* Deborah Y.C. Tan,† Christine H. Yuan,† Staffan Holmin, MD, PhD,‡ Tommy Andersson, MD, PhD,‡ Erik Lundström, MD, PhD,§ Hock Luen Teoh, MBChB,* Bernard P.L. Chan, MBBS,* Rahul Rathakrishnan, BM, MBBS,* Eric Y.S. Ting, MBBS,‖ Vijay K. Sharma, MBBS,*,† and Leonard L.L. Yeo, MBBS*,‡
Background: The Alberta Stroke Program Early CT Score (ASPECTS) on baseline imaging is an established predictor of functional outcome in anterior circulation acute ischemic stroke (AIS). We studied ASPECTS before intravenous thrombolysis (IVT) and at 24 hours to assess its prognostic value. Methods: Data for consecutive anterior circulation AIS patients treated with IVT from 2006 to 2013 were extracted from a prospectively managed registry at our tertiary center. Prethrombolysis and 24-hour ASPECTS were evaluated by 2 independent neuroradiologists. Outcome measures included symptomatic intracranial hemorrhage (SICH), modified Rankin Scale (mRS) at 90 days, and mortality. Unfavorable functional outcome was defined by mRS >1. Dramatic ASPECTS progression (DAP) was defined as deterioration in ASPECTS by 6 points or more. Results: Of 554 AIS patients thrombolyzed during the study period, 400 suffered from anterior circulation infarction. The median age was 65 years (interquartile range (IQR): 59-70) and the median National Institutes of Health Stroke Scale score was 18 points (IQR: 12-22). Compared with the pre-IVT ASPECTS (area under the curve [AUC] = .64, 95% confidence interval [CI]: .54-.65, P = .001), ASPECTS on the 24hour CT scan (AUC = .78, 95% CI: .73-.82, P < .001), and change in ASPECTS (AUC = .69, 95% CI: .64-.74, P < .001) were better predictors of unfavorable functional outcome at 3 months. DAP, noted in 34 (14.4%) patients with good baseline ASPECTS (8-10 points), was significantly associated with unfavorable functional outcome (odds ratio [OR]: 9.91, 95% CI: 3.37-29.19, P ≤ .001), mortality (OR: 21.99, 95% CI: 7.98-60.58, P < .001), and SICH (OR: 8.57, 95% CI: 2.87-25.59, P < .001). Conclusion: Compared with the pre-thrombolysis score, ASPECTS measured
From the *Division of Neurology, Department of Medicine, National University Health System, Singapore; †Yong Loo Lin School of Medicine, National University of Singapore, Singapore; ‡Department of Clinical Neuroscience, Karolinska Institutet and Department of Neuroradiology; §Department of Clinical Neuroscience, Karolinska Institutet and Department of Neurology, Karolinska University Hospital, Stockholm, Sweden; and ‖Department of Diagnostic Imaging, National University Health System, Singapore. Received April 25, 2017; revision received May 4, 2017; accepted May 7, 2017. Ethics approval: Ethics approval for this project was obtained from the institutional review board (IRB). Grant support: This research was sponsored by the National Medical Research Council (NMRC), Singapore (Grant number CNIG12nov001). Wan-Yee Kong and Benjamin Y.Q. Tan are co-first authors; Vijay. K Sharma and Leonard L.L. Yeo are co-senior authors. Address correspondence to Benjamin Y.Q. Tan, MBBS, MRCP, Division of Neurology, Department of Medicine, National University Health System, Singapore, 1 E Kent Ridge Road, Singapore 119228. E-mail: [email protected]. 1052-3057/$ - see front matter © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2017.05.009
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2017: pp ■■–■■
1
ARTICLE IN PRESS W.-Y. KONG ET AL.
2 at 24 hours as well as serial change in ASPECTS is a better predictor of 3-month functional outcome. Key Words: Acute ischemic stroke—CT scan—thrombolysis—ASPECTS. © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved.
Introduction Acute ischemic stroke (AIS) is a leading cause of dependency.1 Although rapid advances have occurred in endovascular therapy, these facilities are not readily available at many centers.2 Therefore, intravenous thrombolysis (IVT) still remains the primary reperfusion strategy in AIS. Various clinical and radiological assessments are commonly employed for systematic quantification of stroke severity at presentation and at 24 hours for optimal management and early prognostication.3,4 The Alberta Stroke Program Early CT Score (ASPECTS) on the pre-IVT imaging is a simple, rapid, and reliable method to determine the extent of presumed infarcted tissue and cytotoxic edema with high sensitivity and specificity.5,6 It has been validated in various intravenous reperfusion strategies to predict functional outcome as well as complications such as symptomatic intracranial hemorrhage (SICH).5,7,8 Furthermore, pre-IVT ASPECTS has been suggested as a tool for patient selection for reperfusion strategies.9-11 Initial stroke severity primarily determines the clinical outcome in patients with AIS.12 Interestingly, ASPECTS serves well to establish the stroke severity at presentation and matches well with other commonly used clinical scores like the National Institutes of Health Stroke Scale (NIHSS),13 Canadian Neurological Scale,14 and European Stroke Scale.15 However, the stroke severity measured “after” rather than “before” IVT may be a better predictor of functional outcome.16 Accordingly, longitudinal change in pre-IVT NIHSS score to the 24-hour score has been shown to be a stronger predictor of functional outcomes compared with the stroke severity at presentation.17 Because radiological scoring systems estimate the parenchymal damage, serial ASPECTS assessment and their temporal profile may serve as a robust determinant of functional recovery in thrombolyzed AIS patients. We evaluated this potential association in our anterior circulation AIS cohort.
Methods Patient Selection and Data Collection Data for consecutive AIS patients treated with intravenous tissue plasminogen activator and admitted to our tertiary center from 2006 to 2013 were entered in a prospectively maintained registry. All AIS patients underwent non-contrast computed tomography of the brain before IVT and 24 hours after. Stroke subtypes were deter-
mined using the Trial of Org 10172 in Acute Stroke Treatment.18 NIHSS scores were recorded for all cases by credentialed neurologists before IVT and 24 hours after.
Radiological Scoring and Definitions Radiological stroke severity was assessed with ASPECTS assessment by 2 independent neuroradiologists with expertise in stroke imaging. ASPECTS is a 10-point quantitative topographic score, which is determined by evaluating 2 standardized regions of the middle cerebral artery territory, the basal ganglia level and the supraganglionic level, for ischemic changes.5 These areas were divided into 10 distinct regions, each amounting to 1 point. The maximum ASPECTS was 10 points in patients with no ischemic changes on non-contrast computed tomography. One point was deducted for each area with acute ischemic changes. These changes excluded chronic changes such as leukoaraiosis, established infarcts, or atrophy.5 Scoring for pre-IVT and 24-hour CT was performed on the same sitting. ASPECT evaluators were blinded to the clinical data and any discrepancy between them was resolved by consensus. A score of 8-10 points defined good ASPECTS, whereas a score of 0-7 points was considered poor.19 Change in ASPECTS was also calculated as the 24-hour CT ASPECTS minus the initial ASPECTS. Dramatic ASPECTS progression (DAP) was defined as a reduction in ASPECTS by 6 points or more on the 24-hour CT, when compared with the baseline.20 Functional outcome was assessed by the modified Rankin Scale (mRS) at 90 days. An mRS of >1 was defined as unfavorable functional outcome.21 Additionally, we looked at functional independence, defined by mRS 0-2, as a secondary outcome. SICH was defined as the presence of new bleeding on the follow-up CT scan, which was accompanied by an increase in NIHSS by 4 points or more.22
Statistical Methods We present the numerical variables as median and range. Categorical variables were presented as percentages. Numerical predictors were assessed using Mann-Whitney U-test, and categorical variables were evaluated using chisquare test or Fisher exact test where applicable. Variables that were found to have a significant association (P < .05) were entered into the multivariable model to perform logistic regression for determining the independent predictors of the prespecified favorable and unfavorable functional outcome at 90 days. Where applicable, separate multivariate models were used to evaluate the effect of
ARTICLE IN PRESS VALIDATION OF ASPECTS AS OUTCOME PREDICTOR IN STROKE PATIENTS
initial ASPECTS, 24-hour ASPECTS, and change in ASPECTS in predicting clinical or functional outcomes to account for multicollinearity of the variables. Associations were presented as odds ratios (OR) with corresponding 95% confidence intervals (CI). Receiver operating characteristic curves were generated for evaluation and comparison of scores by DeLong’s method23 in predicting respective functional outcomes. Statistical analyses were performed using the IBM Statistical Package for Social Sciences (SPSS) Statistics Version 21.0. Armonk, NY: IBM Corp, 2012.
Results From 2006 to 2013, 554 patients received IVT for AIS at our center. Of these, only 400 patients who suffered from anterior circulation AIS were included in this study. We excluded 144 patients due to their AIS in the posterior circulation. Another 10 patients were excluded because a repeat CT scan was not performed. The median age of our patients
Table 1. Baseline characteristics of the study population Characteristics
Value
Median age in years (IQR) Female sex, n (%) Hypertension, n (%) Dyslipidemia, n (%) Diabetes mellitus, n (%) Smoker, n (%) Atrial fibrillation, n (%) Median pre-tPA systolic BP in mm Hg (IQR) Median pre-tPA diastolic BP in mm Hg (IQR) Median pre-tPA NIHSS score (IQR) Median onset-to-treatment time in minutes (IQR) Median door-to-needle time in minutes (IQR) Modified Rankin Scale 0-1 at 3 months, n (%) Symptomatic intracranial hemorrhage, n (%) Mortality, n (%) Baseline ASPECTS <8, n (%) 24-hour ASPECTS <8, n (%) Dramatic infarct progression, n (%) TOAST classification, n (%) Large artery atherosclerosis Cardioembolism Lacunar Stroke of undetermined cause Stroke of other etiology
65 (54-75) 163 (40.8%) 276 (69.0%) 232 (58.0%) 124 (31.0%) 97 (24.3%) 145 (36.3%) 153 (135-169)
3
was 65 years (interquartile range: 54-75), 60% were male, and the median NIHSS score was 18 points (interquartile range: 12-22) (Table 1). The distributions of baseline and 24-hour ASPECTS are reflected in Figure 1, A. Older age, higher NIHSS score, higher systolic blood pressure at onset, diabetes mellitus, and poor ASPECTS on 24-hour CT were independent predictors of unfavorable functional outcome (Table 2). Baseline ASPECTS as well as change in ASPECTS (between the 24-hour CT and the initial CT) were also independently associated with unfavorable functional outcomes (mRS >1 at 3 months) when studied in separate multivariate models. Similarly, we found independent predictors of functional dependence (mRS >2 at 3 months) to be older age, diabetes mellitus, atrial fibrillation, higher NIHSS at onset, and poor 24-hour ASPECTS (Table 3). In separate models, baseline ASPECTS and change in ASPECTS were also found to be independent predictors of functional dependence. SICH was observed in 36 (9%) of patients, whereas 54 (13.5%) patients died within the first 3 months after AIS. Mortality at 3 months was found to be associated with older age and longer door-to-needle time. Baseline poor ASPECTS, poor 24-hour ASPECTS, and change in
80 (70-91) 18 (12-22) 160 (131-194) 83 (68-106) 170 (42.5%) 36 (9.0%) 54 (13.5%) 164 (41.0%) 250 (62.5%) 43 (10.8%) 137 (34.3%) 172 (43.0%) 31 (7.8%) 56 (14.0%) 4 (1.0%)
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; BP, blood pressure; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; TOAST, trial of Org in acute stroke treatment; tPA, tissue plasminogen activator.
Figure 1. (A) Baseline and post-thrombolysis 24-hour ASPECTS distribution. (B) Comparing the performance of baseline ASPECTS (AUC = .64, 95% CI: .59-.70, P < .001), 24-hour ASPECTS (AUC = .78, 95% CI: .73.82, P < .001), and change in ASPECTS (AUC = .69, 95% CI: .64-.74) in predicting unfavorable mRS outcome using receiver operating characteristic curves. Difference in AUC 24-hour ASPECTS versus baseline ASPECTS (by DeLong’s method) = .14, z = −2.936, P < .001; difference in AUC change in ASPECTS versus baseline ASPECTS (by DeLong’s method) = .05, z = −2.831, P = .002. Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; AUC, area under the curve; CI, confidence interval; mRS, modified Rankin Scale.
4
Table 2. Predictors of unfavorable functional outcomes
Univariate analysis Unfavorable functional outcome (mRS >1) (n = 230)
P value
60 (50-70) 53 (31.2%) 95 (55.9%) 88 (51.8%) 39 (22.9%) 52 (30.6%) 41 (24.1%) 160 (130-200) 81 (67-104) 13 (8-19) 148 (130-162) 80 (70-89) 51 (30.0%) 70 (41.2%) −1 (−3 to 1)
70 (60-77) 110 (47.8%) 181 (78.7%) 144 (62.6%) 85 (37.0%) 45 (19.6%) 104 (45.2%) 160 (134-190) 84 (70-108) 20 (16-23) 156 (139-174) 82 (70-95) 113 (49.1%) 180 (78.3%) −3 (−6 to 0)
<.001 .001 <.001 .030 .003 .011 <.001 .802 .485 <.001 <.001 .130 <.001 <.001 <.001
Multivariate analysis (baseline ASPECTS)
Adjusted odds ratio (95% CI)
P value
Adjusted odds ratio (95% CI)
1.040 (1.020-1.060)
<.001
2.304 (1.339-3.962)
Multivariate analysis (change in ASPECTS)
P value
Adjusted odds ratio (95% CI)
P value
1.039 (1.019-1.058)
<.001
1.036 (1.016-1.056)
<.001
.003
2.053 (1.218-3.461)
.007
2.298 (1.340-3.941)
.003
1.109 (1.061-1.159) 1.012 (1.002-1.023)
<.001 .020
1.129 (1.083-1.177) 1.013 (1.003-1.024)
<.001 .010
1.132 (1.086-1.178) 1.011 (1.001-1.021)
<.001 .034
NA 2.681 (1.558-4.608) NA
NA <.001 NA
1.112 (1.003-1.233) NA NA
.045 NA NA
NA NA .786 (.689-.857)
NA NA <.001
W.-Y. KONG ET AL.
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; CI, confidence interval; DBP, diastolic blood pressure; DNT, door-to-needle time; IQR, interquartile range; mRS, modified Rankin Scale; NA, not applicable; NIHSS, National Institutes of Health Strokes Scale; OTT, onset-to-treatment time; SBP, systolic blood pressure. Bold indicates P values < 0.05 are statistically significant.
ARTICLE IN PRESS
Median age in years (IQR) Female, n (%) Hypertension, n (%) Hyperlipidemia, n (%) Diabetes mellitus, n (%) Smoking, n (%) Atrial fibrillation, n (%) Median OTT in minutes (IQR) Median DNT in minutes (IQR) Median NIHSS at onset, (IQR) Median SBP at onset in mm Hg (IQR) Median DBP at onset in mm Hg (IQR) Baseline ASPECTS <8, n (%) 24-hour ASPECTS <8, n (%) Median change in ASPECTS (IQR)
Favorable functional outcome (mRS 0-1) (n = 170)
Multivariate analysis (24-hour ASPECTS)
Univariate analysis Functional dependence mRS >2 (n = 199)
P value
60 (51-70) 61 (30.3%) 118 (58.7%) 107 (53.2%) 51 (25.4%) 35 (16.7%) 47 (23.4%) 157 (130-195)
72 (61-78) 102 (51.3%) 158 (79.4%) 125 (62.8%) 73 (36.7%) 62 (30.8%) 98 (49.2%) 167 (138-190)
<.001 <.001 <.001 .052 .014 .002 <.001 .318
81 (67-101)
85 (70-112)
.112
14 (8-20) 150 (134-165) 80 (70-90) 59 (29.4%) 86 (42.8%) −1 (−3 to 1)
20 (17-23) 155 (137-172) 82 (70-92) 105 (52.8%) 164 (82.4%) −3 (−6 to 0)
<.001 .031 .371 <.001 <.001 <.001
Multivariate analysis (baseline ASPECTS)
Adjusted odds ratio (95% CI)
P value
Adjusted odds ratio (95% CI)
1.036 (1.014-1.058) .625 (.370-1.055) 1.524 (.853-2.723)
.001 .078 .155
1.758 (1.007-3.066) 1.045 (.571-1.912) 1.772 (1.055-3.010)
Multivariate analysis (change in ASPECTS
P value
Adjusted odds ratio (95% CI)
P value
1.035 (1.014-1.056) .542 (.325-.906) 1.730 (.981-3.049)
.001 .019 .058
1.031 (1.010-1.052) .616 (.367-1.036) 1.484 (.839-2.623)
.004 .068 .175
.047 .886 .031
1.469 (.855-2.524) 1.112 (.618-2.000) 1.710 (1.026-2.851)
.163 .723 .040
1.775 (1.025-3.071) 1.197 (.652-2.203) 1.799 (1.066-3.036)
.040 .560 .028
1.098 (1.050-1.148) 1.003 (.993-1.013)
<.001 .596
1.107 (1.061-1.154) 1.004 (.994-1.014)
<.001 .467
1.119 (1.075-1.166) 1.002 (.992-1.012)
<.001 .715
NA 2.865 (1.639-5.00) NA
NA <.001 NA
1.182 (1.065-1.311) NA NA
.002 NA NA
NA NA .786 (.708-.873)
NA NA <.001
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; CI, confidence interval; DBP, diastolic blood pressure; IQR, interquartile range; mRS, modified Rankin Scale; NA, not applicable; NIHSS, National Institutes of Health Strokes Scale; SBP, systolic blood pressure. Bold indicates P values < 0.05 are statistically significant.
ARTICLE IN PRESS
Median age in years (IQR) Female, n (%) Hypertension, n (%) Hyperlipidemia, n (%) Diabetes mellitus, n (%) Smoking, n (%) Atrial fibrillation, n (%) Median onset-to-treatment time in minutes (IQR) Median door-to-needle time in minutes (IQR) Median NIHSS at onset (IQR) Median SBP at onset in mm Hg (IQR) Median DBP at onset in mm Hg (IQR) Baseline ASPECTS <8, n (%) 24-hour ASPECTS <8, n (%) Median change in ASPECTS (IQR)
Functional independence mRS 0-2 (n = 201)
Multivariate analysis (24-hour ASPECTS)
VALIDATION OF ASPECTS AS OUTCOME PREDICTOR IN STROKE PATIENTS
Table 3. Results of univariate and multivariate analyses: predictors of functional dependence (mRS >2)
5
ARTICLE IN PRESS W.-Y. KONG ET AL.
6
Table 4. Results of univariate and multivariate analyses: predictors of dramatic ASPECTS progression in patients with good baseline ASPECTS 8-10 points Univariate analysis
Multivariate analysis
No dramatic ASPECTS Dramatic ASPECTS Adjusted odds ratio progression (n = 202) progression (n = 34) P value (95% CI) P value Median age in years (IQR) Female, n (%) Hypertension, n (%) Hyperlipidemia, n (%) Diabetes mellitus, n (%) Smoking, n (%) Atrial fibrillation, n (%) Median onset-to-treatment time, in minutes (IQR) Median door-to-needle time, in minutes (IQR) Median NIHSS at onset (IQR) Median SBP at onset in mm Hg (IQR) Median DBP at onset in mm Hg (IQR)
62 (53-72) 68 (33.7%) 133 (65.8%) 122 (60.4%) 50 (24.8%) 58 (28.7%) 57 (28.2%) 165 (135-201)
75 (66-79) 22 (64.7%) 26 (76.5%) 19 (55.9%) 14 (41.2%) 6 (17.6%) 20 (58.8%) 144 (114-181)
<.001 .001 .221 .620 .046 .179 <.001 .019
81 (69-103)
82 (71-111)
.681
13 (9-19) 152 (138-168) 84 (74-92)
21 (16-23) 159 (149-173) 75 (67-92)
<.001 .087 .104
1.05 (1.01-1.09) 3.48 (1.43-8.43)
.013 .006
2.96 (1.16-7.55)
.023
1.17 (1.08-1.26)
<.001
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; CI, confidence interval; DBP, diastolic blood pressure; IQR, interquartile range; NIHSS, National Institutes of Health Strokes Scale; SBP, systolic blood pressure. Bold indicates P values < 0.05 are statistically significant.
ASPECTS were significant predictors of mortality and SICH on both univariate and multivariate analyses (Supplementary Tables S1 and S2). In comparing the prognostic utility for predicting unfavorable functional outcome of baseline ASPECTS, 24hour ASPECTS, and change in ASPECTS, receiver operating characteristic curves were constructed. ASPECTS on the 24-hour scan (area under the curve [AUC] = .78, 95% CI: .73-.82, P < .001) was a better predictor compass compared with baseline ASPECTS (AUC = .64, 95% CI: .59.70, P < .001). This significance was further quantified by DeLong’s method (difference in AUC = .14, z = −2.936, P = .001) (Fig 1, B). Change in ASPECTS was also superior to baseline ASPECTS in predicting functional outcome (AUC = .69, 95% CI: .59-.70, P < .001; difference in AUC = .05, z = −2.831, P = .002). Interestingly, a good ASPECTS (8-10 points) was observed on the initial CT in 236 (59%) patients. This score deteriorated in a fair proportion (104 case, 44%) of these patients on their 24-hour CT scan, and 34 (14%) of them developed DAP. Older age (OR: 1.05, 95% CI: 1.01-1.09, P = .013), female gender (OR: 3.48, 95%CI: 1.43-8.43, P = .006), higher NIHSS at onset (OR: 1.17, 95% CI: 1.081.26, P < .001), and diabetes mellitus (OR: 2.96, 95% CI: 1.16-7.55, P = .023) were found to be independently associated with DAP (Table 4). Furthermore, in patients with good baseline ASPECTS, poor 24-hour ASPECTS as well as DAP were associated with unfavorable functional outcome (OR: 5.11, 95% CI: 2.91-8.93, P < .001; OR: 9.91, 95% CI: 3.37-29.19, P < .001, respectively), functional dependence (OR: 6.45, 95% CI:
3.62-11.49, P < .001; OR: 12.23, 95% CI: 4.53-33.03, P < .001, respectively), mortality (OR: 6.70, 95% CI: 2.19-20.47, P < .001; OR: 21.99, 95% CI: 7.98-60.58, P < .001, respectively), and SICH (OR: 9.29, 95% CI: 2.05-42.14, P = .001; OR: 8.57, 95% CI: 2.87-25.59, P < .001, respectively) (Supplementary Table S3).
Discussion ASPECTS on baseline imaging has an irrefutable role in initial patient selection for reperfusion strategies in AIS.10-12 However, our study demonstrates that 24-hour ASPECTS on CT and change in ASPECTS (between the baseline and 24-hour CT) were more accurate predictors of functional outcome at 3 months, when compared with baseline ASPECTS. Furthermore, in the subgroup of patients with initially good baseline ASPECTS, DAP is associated with unfavorable functional outcome, functional dependence, mortality, as well as SICH. Our findings are consistent with prior studies that have established that early, rather than baseline, markers of clinical stroke severity are important predictors of clinical outcomes.24 Accordingly, 24-hour ASPECTS was superior to baseline ASPECTS in predicting clinical outcomes. Studies on clinical stroke severity scores continued to trend scores to 48-hour and beyond, and identified that it was the latent trajectory subgroups based on longitudinal scores that were the best predictors of 90-day outcomes.25 Score trajectory in a radiological score like ASPECTS would be difficult to establish, as it would require at least 3 serial scans, which is not done in most
ARTICLE IN PRESS VALIDATION OF ASPECTS AS OUTCOME PREDICTOR IN STROKE PATIENTS
of the cases. Nevertheless, we compare the scores at these 2 time points (baseline and 24-hours) and established that the 24-hour ASPECTS or change in score from the baseline CT had greater prognostic utility. ASPECTS on the pre-thrombolysis CT scan has been used for risk stratification, estimation of time of symptom onset, and selection of patients eligible for endovascular treatment. Furthermore, it correlates well with functional outcome at 3 months. On the contrary, 24-hour or serial ASPECTS has been evaluated scarcely, perhaps due to factors like performing of magnetic resonance imaging at 24 hours instead of CT scan at some centers, reliability of serial NIHSS scores, or fear of overestimation of ASPECTS due to cerebral edema. Our study suggests that 24-hour as well as serial CT ASPECTS may identify patients who develop SICH to aid in appropriate antithrombotic therapy, serve as reliable predictors of functional outcome at 3 months, and help in early prognostication, planning optimal rehabilitation strategies, as well as proper disposition of AIS patients. These results are concordant with a prior study demonstrating that 24-hour ASPECTS was superior to baseline ASPECTS in predicting clinical outcomes of patients undergoing endovascular treatment for AIS.20 However, there are some important differences in our study design compared with the solitaire flow restoration with the intention for thrombectomy (SWIFT) trial. The SWIFT trial analyses included solely patients with large vessel occlusion undergoing endovascular stroke treatment, whereas in our study we have included a broader cohort of patients (different stroke types) undergoing thrombolysis for anterior circulation stroke. It can be rationalized that the fundamental principle of the ASPECTS score (ordinal categorization of the extent of CT-based infarct signs) is independent of the underlying stroke type (large vessel occlusion versus all strokes) and hence supportive of our findings. One interesting observation in our study was regarding a specific subgroup of patients with good baseline ASPECTS (ASPECTS 8-10), which eventually fared poorly. In our study, almost 50% (117 out of 236) of such patients subsequently experienced unfavorable functional outcomes at 90 days. Approximately, 9% expired and 6% developed SICH. Our study subsequently demonstrated that both 24-hour ASPECTS and DAP (ASPECTS deterioration of 6 points or more) were reliable predictors of poor functional outcome, even in patients with good baseline ASPECTS. Other clinical predictors of unfavorable functional outcome were in accordance with the conventional understanding of stroke pathophysiology. For example, higher systolic blood pressure at onset, presence of diabetes mellitus, higher age, and higher NIHSS scores were found to be independent predictors of poorer outcome in our AIS cohort.26 One surprising finding in our study was the lack of association between poorer outcome and doorto-needle time and onset-to-treatment time.27,28 Perhaps,
7
higher stroke severity (baseline median NIHSS score: 18 points) at presentation counterbalanced these important variables in our study population. Certain limitations of our study need to be acknowledged. Although this study employed a moderately large sample size of 400 patients with clinical outcomes at 90 days, it remained limited by the retrospective design of the study. Furthermore, 10 patients were excluded from the analyses as they could not have the repeat CT scan due to poor neurologic status or death within 24 hours. Other inherent problems with the ASPECTS methodology were also present in our study. For example, we included only anterior circulation strokes, but a small proportion of them might have suffered from a stroke in the posterior circulation. There also remain technical limitations where a CT scan done very early in the course of stroke may not show the changes of ASPECTS demonstrating ischemic changes or associated vasogenic and cytotoxic edema.29 Nonetheless, our findings proposing the utility of serial imaging would compensate for these limitations. In summary, we found that serial changes of tissue injury on routinely acquired imaging, such as non-contrast CT, may provide critical data on the response of brain tissue to varying degrees of reperfusion that has occurred. It may not only reveal the extent of ischemic injury but also outline the trajectory of expected clinical sequelae. All patients receive repeat cerebral imaging at 24 hours after IVT, and this can serve an important role for early identification of malignant infarction and help determine which patients require intensive neurologic monitoring for better resource planning.
Conclusion Although baseline ASPECTS remains important as a patient selection tool for thrombolysis, ASPECTS measured at 24 hours as well as serial change in ASPECTS may demonstrate greater clinical utility as more accurate prognostic markers of clinical outcomes in patients with AIS.
Appendix: Supplementary Material Supplementary data to this article can be found online at doi:10.1016/j.jstrokecerebrovasdis.2017.05.009.
References 1. CDC, NCHS. Underlying cause of death 1999-2013 on CDC WONDER online database Released 2015. Data are from the Multiple Cause of Death Files, 1999-2013, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative Program. Available at: https://wonder.cdc.gov/wonder/ help/ucd.html. Accessed February 3, 2015. 2. Campbell BCV, Mitchell PJ, Kleinig TJ, et al. Endovascular therapy for ischemic stroke with perfusion-imaging selection. N Engl J Med 2015;372:1009-1018.
ARTICLE IN PRESS W.-Y. KONG ET AL.
8 3. Schwamm LH, Pancioli A, Acker JE, et al. Recommendations for the establishment of stroke systems of care: recommendations from the American Stroke Association’s Task Force on the Development of Stroke Systems. Stroke 2005;36:690-703. 4. Alberts MJ, Latchaw RE, Selman WR, et al. Recommendations for comprehensive stroke centers: a consensus statement from the Brain Attack Coalition. Stroke 2005;36:1597-1616. 5. Pexman JH, Barber PA, Hill MD, et al. Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR Am J Neuroradiol 2001;22:1534-1542. 6. Puetz V, Dzialowski I, Hill MD, et al. The Alberta Stroke Program Early CT Score in clinical practice: what have we learned? Int J Stroke 2009;4:354-364. 7. 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. 8. Coutts SB, Demchuk AM, Barber PA, et al. Interobserver variation of ASPECTS in real time. Stroke 2004;35:e103e105. 9. Dzialowski I, Hill MD, Coutts SB, et al. Extent of early ischemic changes on computed tomography (CT) before thrombolysis: prognostic value of the Alberta Stroke Program Early CT Score in ECASS II. Stroke 2006;37:973978. 10. Hill MD, Rowley HA, Adler F, et al. PROACT-II Investigators. Selection of acute ischemic stroke patients for intra-arterial thrombolysis with pro-urokinase by using ASPECTS. Stroke 2003;34:1925-1931. 11. Tsivgoulis G, Saqqur M, Sharma VK, et al. Association of pretreatment ASPECTS scores with tPA-induced arterial recanalization in acute middle cerebral artery occlusion. J Neuroimaging 2008;18:56-61. 12. Dewey HM, Donnan GA, Freeman EJ, et al. Interrater reliability of the National Institutes of Health Stroke Scale: rating by neurologists and nurses in a community-based stroke incidence study. Cerebrovasc Dis 1999;9:323-327. 13. Goldstein LB, Samsa GP. Reliability of the National Institutes of Health Stroke Scale. Extension to nonneurologists in the context of a clinical trial. Stroke 1997;28:307-310. 14. Côté R, Battista RN, Wolfson C, et al. The Canadian Neurological Scale: validation and reliability assessment. Neurology 1989;39:638-643. 15. Hantson L, De Weerdt W, De Keyser J, et al. The European Stroke Scale. Stroke 1994;25:2215-2219. 16. Govan L, Langhorne P, Weir CJ. Categorizing stroke prognosis using different stroke scales. Stroke 2009;40:3396-3399. 17. Al-Ajlan FS, Goyal M, Demchuk AM, et al. Intra-arterial therapy and post-treatment infarct volumes: insights from
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28.
29.
the ESCAPE randomized controlled trial. Stroke 2016;47:777-781. 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. MacCallum C, Churilov L, Mitchell P, et al. Low Alberta Stroke Program Early CT Score (ASPECTS) associated with malignant middle cerebral artery infarction. Cerebrovasc Dis 2014;38:39-45. Liebeskind DS, Jahan R, Nogueira RG, et al. Serial Alberta Stroke Program Early CT Score from baseline to 24 hours in solitaire flow restoration with the intention for thrombectomy (SWIFT) study: a novel surrogate end point for revascularization in acute stroke. Stroke 2014;45:723727. Aviv RI, Mandelcorn J, Chakraborty S, et al. Alberta Stroke Program Early CT Scoring of CT perfusion in early stroke visualization and assessment. AJNR Am J Neuroradiol 2007;28:1975-1980. Chimowitz MI. Endovascular treatment for acute ischemic stroke–still unproven. N Engl J Med 2013;368:952-955. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837-845. Broderick JP, Lu M, Kothari R, et al. Finding the most powerful measures of the effectiveness of tissue plasminogen activator in the NINDS tPA stroke trial. Stroke 2000;31:2335-2341. Sajobi TT, Menon BK, Wang M, et al. Early trajector of stroke severity predicts long-term functional outcomes in ischemic stroke subjects. Results from the ESCAPE trial (Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion With Emphasis on Minimizing CT to Recanalization Times). Stroke 2017;48: 105-110. Demchuk AM, Goyal M, Menon BK, et al. Endovascular treatment for Small Core and Anterior circulation Proximal occlusion with Emphasis on minimizing CT to recanalization times (ESCAPE) trial: methodology. Int J Stroke 2015;10:429-438. Wardlaw JM, Murray V, Berge E, et al. Recombinant tissue plasminogen activator for acute ischaemic stroke: an updated systematic review and meta-analysis. Lancet 2012;39:2364-2372. National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke: the National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med 1995;333:1581-1587. Butcher KS, Lee SB, Parsons MW, et al. Differential prognosis of isolated cortical swelling and hypoattenuation on CT in acute stroke. Stroke 2007;38:941-947.
Validation of Serial Alberta Stroke Program Early CT Score as an Outcome Predictor in Thrombolyzed Stroke Patients Wan-Yee Kong, MBBS,* Benjamin Y.Q. Tan, MBBS, MRCP,* Nicholas J.H. Ngiam, MBBS,* Deborah Y.C. Tan,† Christine H. Yuan,† Staffan Holmin, MD, PhD,‡ Tommy Andersson, MD, PhD,‡ Erik Lundström, MD, PhD,§ Hock Luen Teoh, MBChB,* Bernard P.L. Chan, MBBS,* Rahul Rathakrishnan, BM, MBBS,* Eric Y.S. Ting, MBBS,‖ Vijay K. Sharma, MBBS,*,† and Leonard L.L. Yeo, MBBS*,‡
Background: The Alberta Stroke Program Early CT Score (ASPECTS) on baseline imaging is an established predictor of functional outcome in anterior circulation acute ischemic stroke (AIS). We studied ASPECTS before intravenous thrombolysis (IVT) and at 24 hours to assess its prognostic value. Methods: Data for consecutive anterior circulation AIS patients treated with IVT from 2006 to 2013 were extracted from a prospectively managed registry at our tertiary center. Prethrombolysis and 24-hour ASPECTS were evaluated by 2 independent neuroradiologists. Outcome measures included symptomatic intracranial hemorrhage (SICH), modified Rankin Scale (mRS) at 90 days, and mortality. Unfavorable functional outcome was defined by mRS >1. Dramatic ASPECTS progression (DAP) was defined as deterioration in ASPECTS by 6 points or more. Results: Of 554 AIS patients thrombolyzed during the study period, 400 suffered from anterior circulation infarction. The median age was 65 years (interquartile range (IQR): 59-70) and the median National Institutes of Health Stroke Scale score was 18 points (IQR: 12-22). Compared with the pre-IVT ASPECTS (area under the curve [AUC] = .64, 95% confidence interval [CI]: .54-.65, P = .001), ASPECTS on the 24hour CT scan (AUC = .78, 95% CI: .73-.82, P < .001), and change in ASPECTS (AUC = .69, 95% CI: .64-.74, P < .001) were better predictors of unfavorable functional outcome at 3 months. DAP, noted in 34 (14.4%) patients with good baseline ASPECTS (8-10 points), was significantly associated with unfavorable functional outcome (odds ratio [OR]: 9.91, 95% CI: 3.37-29.19, P ≤ .001), mortality (OR: 21.99, 95% CI: 7.98-60.58, P < .001), and SICH (OR: 8.57, 95% CI: 2.87-25.59, P < .001). Conclusion: Compared with the pre-thrombolysis score, ASPECTS measured
From the *Division of Neurology, Department of Medicine, National University Health System, Singapore; †Yong Loo Lin School of Medicine, National University of Singapore, Singapore; ‡Department of Clinical Neuroscience, Karolinska Institutet and Department of Neuroradiology; §Department of Clinical Neuroscience, Karolinska Institutet and Department of Neurology, Karolinska University Hospital, Stockholm, Sweden; and ‖Department of Diagnostic Imaging, National University Health System, Singapore. Received April 25, 2017; revision received May 4, 2017; accepted May 7, 2017. Ethics approval: Ethics approval for this project was obtained from the institutional review board (IRB). Grant support: This research was sponsored by the National Medical Research Council (NMRC), Singapore (Grant number CNIG12nov001). Wan-Yee Kong and Benjamin Y.Q. Tan are co-first authors; Vijay. K Sharma and Leonard L.L. Yeo are co-senior authors. Address correspondence to Benjamin Y.Q. Tan, MBBS, MRCP, Division of Neurology, Department of Medicine, National University Health System, Singapore, 1 E Kent Ridge Road, Singapore 119228. E-mail: [email protected]. 1052-3057/$ - see front matter © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2017.05.009
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2017: pp ■■–■■
1
ARTICLE IN PRESS W.-Y. KONG ET AL.
2 at 24 hours as well as serial change in ASPECTS is a better predictor of 3-month functional outcome. Key Words: Acute ischemic stroke—CT scan—thrombolysis—ASPECTS. © 2017 National Stroke Association. Published by Elsevier Inc. All rights reserved.
Introduction Acute ischemic stroke (AIS) is a leading cause of dependency.1 Although rapid advances have occurred in endovascular therapy, these facilities are not readily available at many centers.2 Therefore, intravenous thrombolysis (IVT) still remains the primary reperfusion strategy in AIS. Various clinical and radiological assessments are commonly employed for systematic quantification of stroke severity at presentation and at 24 hours for optimal management and early prognostication.3,4 The Alberta Stroke Program Early CT Score (ASPECTS) on the pre-IVT imaging is a simple, rapid, and reliable method to determine the extent of presumed infarcted tissue and cytotoxic edema with high sensitivity and specificity.5,6 It has been validated in various intravenous reperfusion strategies to predict functional outcome as well as complications such as symptomatic intracranial hemorrhage (SICH).5,7,8 Furthermore, pre-IVT ASPECTS has been suggested as a tool for patient selection for reperfusion strategies.9-11 Initial stroke severity primarily determines the clinical outcome in patients with AIS.12 Interestingly, ASPECTS serves well to establish the stroke severity at presentation and matches well with other commonly used clinical scores like the National Institutes of Health Stroke Scale (NIHSS),13 Canadian Neurological Scale,14 and European Stroke Scale.15 However, the stroke severity measured “after” rather than “before” IVT may be a better predictor of functional outcome.16 Accordingly, longitudinal change in pre-IVT NIHSS score to the 24-hour score has been shown to be a stronger predictor of functional outcomes compared with the stroke severity at presentation.17 Because radiological scoring systems estimate the parenchymal damage, serial ASPECTS assessment and their temporal profile may serve as a robust determinant of functional recovery in thrombolyzed AIS patients. We evaluated this potential association in our anterior circulation AIS cohort.
Methods Patient Selection and Data Collection Data for consecutive AIS patients treated with intravenous tissue plasminogen activator and admitted to our tertiary center from 2006 to 2013 were entered in a prospectively maintained registry. All AIS patients underwent non-contrast computed tomography of the brain before IVT and 24 hours after. Stroke subtypes were deter-
mined using the Trial of Org 10172 in Acute Stroke Treatment.18 NIHSS scores were recorded for all cases by credentialed neurologists before IVT and 24 hours after.
Radiological Scoring and Definitions Radiological stroke severity was assessed with ASPECTS assessment by 2 independent neuroradiologists with expertise in stroke imaging. ASPECTS is a 10-point quantitative topographic score, which is determined by evaluating 2 standardized regions of the middle cerebral artery territory, the basal ganglia level and the supraganglionic level, for ischemic changes.5 These areas were divided into 10 distinct regions, each amounting to 1 point. The maximum ASPECTS was 10 points in patients with no ischemic changes on non-contrast computed tomography. One point was deducted for each area with acute ischemic changes. These changes excluded chronic changes such as leukoaraiosis, established infarcts, or atrophy.5 Scoring for pre-IVT and 24-hour CT was performed on the same sitting. ASPECT evaluators were blinded to the clinical data and any discrepancy between them was resolved by consensus. A score of 8-10 points defined good ASPECTS, whereas a score of 0-7 points was considered poor.19 Change in ASPECTS was also calculated as the 24-hour CT ASPECTS minus the initial ASPECTS. Dramatic ASPECTS progression (DAP) was defined as a reduction in ASPECTS by 6 points or more on the 24-hour CT, when compared with the baseline.20 Functional outcome was assessed by the modified Rankin Scale (mRS) at 90 days. An mRS of >1 was defined as unfavorable functional outcome.21 Additionally, we looked at functional independence, defined by mRS 0-2, as a secondary outcome. SICH was defined as the presence of new bleeding on the follow-up CT scan, which was accompanied by an increase in NIHSS by 4 points or more.22
Statistical Methods We present the numerical variables as median and range. Categorical variables were presented as percentages. Numerical predictors were assessed using Mann-Whitney U-test, and categorical variables were evaluated using chisquare test or Fisher exact test where applicable. Variables that were found to have a significant association (P < .05) were entered into the multivariable model to perform logistic regression for determining the independent predictors of the prespecified favorable and unfavorable functional outcome at 90 days. Where applicable, separate multivariate models were used to evaluate the effect of
ARTICLE IN PRESS VALIDATION OF ASPECTS AS OUTCOME PREDICTOR IN STROKE PATIENTS
initial ASPECTS, 24-hour ASPECTS, and change in ASPECTS in predicting clinical or functional outcomes to account for multicollinearity of the variables. Associations were presented as odds ratios (OR) with corresponding 95% confidence intervals (CI). Receiver operating characteristic curves were generated for evaluation and comparison of scores by DeLong’s method23 in predicting respective functional outcomes. Statistical analyses were performed using the IBM Statistical Package for Social Sciences (SPSS) Statistics Version 21.0. Armonk, NY: IBM Corp, 2012.
Results From 2006 to 2013, 554 patients received IVT for AIS at our center. Of these, only 400 patients who suffered from anterior circulation AIS were included in this study. We excluded 144 patients due to their AIS in the posterior circulation. Another 10 patients were excluded because a repeat CT scan was not performed. The median age of our patients
Table 1. Baseline characteristics of the study population Characteristics
Value
Median age in years (IQR) Female sex, n (%) Hypertension, n (%) Dyslipidemia, n (%) Diabetes mellitus, n (%) Smoker, n (%) Atrial fibrillation, n (%) Median pre-tPA systolic BP in mm Hg (IQR) Median pre-tPA diastolic BP in mm Hg (IQR) Median pre-tPA NIHSS score (IQR) Median onset-to-treatment time in minutes (IQR) Median door-to-needle time in minutes (IQR) Modified Rankin Scale 0-1 at 3 months, n (%) Symptomatic intracranial hemorrhage, n (%) Mortality, n (%) Baseline ASPECTS <8, n (%) 24-hour ASPECTS <8, n (%) Dramatic infarct progression, n (%) TOAST classification, n (%) Large artery atherosclerosis Cardioembolism Lacunar Stroke of undetermined cause Stroke of other etiology
65 (54-75) 163 (40.8%) 276 (69.0%) 232 (58.0%) 124 (31.0%) 97 (24.3%) 145 (36.3%) 153 (135-169)
3
was 65 years (interquartile range: 54-75), 60% were male, and the median NIHSS score was 18 points (interquartile range: 12-22) (Table 1). The distributions of baseline and 24-hour ASPECTS are reflected in Figure 1, A. Older age, higher NIHSS score, higher systolic blood pressure at onset, diabetes mellitus, and poor ASPECTS on 24-hour CT were independent predictors of unfavorable functional outcome (Table 2). Baseline ASPECTS as well as change in ASPECTS (between the 24-hour CT and the initial CT) were also independently associated with unfavorable functional outcomes (mRS >1 at 3 months) when studied in separate multivariate models. Similarly, we found independent predictors of functional dependence (mRS >2 at 3 months) to be older age, diabetes mellitus, atrial fibrillation, higher NIHSS at onset, and poor 24-hour ASPECTS (Table 3). In separate models, baseline ASPECTS and change in ASPECTS were also found to be independent predictors of functional dependence. SICH was observed in 36 (9%) of patients, whereas 54 (13.5%) patients died within the first 3 months after AIS. Mortality at 3 months was found to be associated with older age and longer door-to-needle time. Baseline poor ASPECTS, poor 24-hour ASPECTS, and change in
80 (70-91) 18 (12-22) 160 (131-194) 83 (68-106) 170 (42.5%) 36 (9.0%) 54 (13.5%) 164 (41.0%) 250 (62.5%) 43 (10.8%) 137 (34.3%) 172 (43.0%) 31 (7.8%) 56 (14.0%) 4 (1.0%)
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; BP, blood pressure; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; TOAST, trial of Org in acute stroke treatment; tPA, tissue plasminogen activator.
Figure 1. (A) Baseline and post-thrombolysis 24-hour ASPECTS distribution. (B) Comparing the performance of baseline ASPECTS (AUC = .64, 95% CI: .59-.70, P < .001), 24-hour ASPECTS (AUC = .78, 95% CI: .73.82, P < .001), and change in ASPECTS (AUC = .69, 95% CI: .64-.74) in predicting unfavorable mRS outcome using receiver operating characteristic curves. Difference in AUC 24-hour ASPECTS versus baseline ASPECTS (by DeLong’s method) = .14, z = −2.936, P < .001; difference in AUC change in ASPECTS versus baseline ASPECTS (by DeLong’s method) = .05, z = −2.831, P = .002. Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; AUC, area under the curve; CI, confidence interval; mRS, modified Rankin Scale.
4
Table 2. Predictors of unfavorable functional outcomes
Univariate analysis Unfavorable functional outcome (mRS >1) (n = 230)
P value
60 (50-70) 53 (31.2%) 95 (55.9%) 88 (51.8%) 39 (22.9%) 52 (30.6%) 41 (24.1%) 160 (130-200) 81 (67-104) 13 (8-19) 148 (130-162) 80 (70-89) 51 (30.0%) 70 (41.2%) −1 (−3 to 1)
70 (60-77) 110 (47.8%) 181 (78.7%) 144 (62.6%) 85 (37.0%) 45 (19.6%) 104 (45.2%) 160 (134-190) 84 (70-108) 20 (16-23) 156 (139-174) 82 (70-95) 113 (49.1%) 180 (78.3%) −3 (−6 to 0)
<.001 .001 <.001 .030 .003 .011 <.001 .802 .485 <.001 <.001 .130 <.001 <.001 <.001
Multivariate analysis (baseline ASPECTS)
Adjusted odds ratio (95% CI)
P value
Adjusted odds ratio (95% CI)
1.040 (1.020-1.060)
<.001
2.304 (1.339-3.962)
Multivariate analysis (change in ASPECTS)
P value
Adjusted odds ratio (95% CI)
P value
1.039 (1.019-1.058)
<.001
1.036 (1.016-1.056)
<.001
.003
2.053 (1.218-3.461)
.007
2.298 (1.340-3.941)
.003
1.109 (1.061-1.159) 1.012 (1.002-1.023)
<.001 .020
1.129 (1.083-1.177) 1.013 (1.003-1.024)
<.001 .010
1.132 (1.086-1.178) 1.011 (1.001-1.021)
<.001 .034
NA 2.681 (1.558-4.608) NA
NA <.001 NA
1.112 (1.003-1.233) NA NA
.045 NA NA
NA NA .786 (.689-.857)
NA NA <.001
W.-Y. KONG ET AL.
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; CI, confidence interval; DBP, diastolic blood pressure; DNT, door-to-needle time; IQR, interquartile range; mRS, modified Rankin Scale; NA, not applicable; NIHSS, National Institutes of Health Strokes Scale; OTT, onset-to-treatment time; SBP, systolic blood pressure. Bold indicates P values < 0.05 are statistically significant.
ARTICLE IN PRESS
Median age in years (IQR) Female, n (%) Hypertension, n (%) Hyperlipidemia, n (%) Diabetes mellitus, n (%) Smoking, n (%) Atrial fibrillation, n (%) Median OTT in minutes (IQR) Median DNT in minutes (IQR) Median NIHSS at onset, (IQR) Median SBP at onset in mm Hg (IQR) Median DBP at onset in mm Hg (IQR) Baseline ASPECTS <8, n (%) 24-hour ASPECTS <8, n (%) Median change in ASPECTS (IQR)
Favorable functional outcome (mRS 0-1) (n = 170)
Multivariate analysis (24-hour ASPECTS)
Univariate analysis Functional dependence mRS >2 (n = 199)
P value
60 (51-70) 61 (30.3%) 118 (58.7%) 107 (53.2%) 51 (25.4%) 35 (16.7%) 47 (23.4%) 157 (130-195)
72 (61-78) 102 (51.3%) 158 (79.4%) 125 (62.8%) 73 (36.7%) 62 (30.8%) 98 (49.2%) 167 (138-190)
<.001 <.001 <.001 .052 .014 .002 <.001 .318
81 (67-101)
85 (70-112)
.112
14 (8-20) 150 (134-165) 80 (70-90) 59 (29.4%) 86 (42.8%) −1 (−3 to 1)
20 (17-23) 155 (137-172) 82 (70-92) 105 (52.8%) 164 (82.4%) −3 (−6 to 0)
<.001 .031 .371 <.001 <.001 <.001
Multivariate analysis (baseline ASPECTS)
Adjusted odds ratio (95% CI)
P value
Adjusted odds ratio (95% CI)
1.036 (1.014-1.058) .625 (.370-1.055) 1.524 (.853-2.723)
.001 .078 .155
1.758 (1.007-3.066) 1.045 (.571-1.912) 1.772 (1.055-3.010)
Multivariate analysis (change in ASPECTS
P value
Adjusted odds ratio (95% CI)
P value
1.035 (1.014-1.056) .542 (.325-.906) 1.730 (.981-3.049)
.001 .019 .058
1.031 (1.010-1.052) .616 (.367-1.036) 1.484 (.839-2.623)
.004 .068 .175
.047 .886 .031
1.469 (.855-2.524) 1.112 (.618-2.000) 1.710 (1.026-2.851)
.163 .723 .040
1.775 (1.025-3.071) 1.197 (.652-2.203) 1.799 (1.066-3.036)
.040 .560 .028
1.098 (1.050-1.148) 1.003 (.993-1.013)
<.001 .596
1.107 (1.061-1.154) 1.004 (.994-1.014)
<.001 .467
1.119 (1.075-1.166) 1.002 (.992-1.012)
<.001 .715
NA 2.865 (1.639-5.00) NA
NA <.001 NA
1.182 (1.065-1.311) NA NA
.002 NA NA
NA NA .786 (.708-.873)
NA NA <.001
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; CI, confidence interval; DBP, diastolic blood pressure; IQR, interquartile range; mRS, modified Rankin Scale; NA, not applicable; NIHSS, National Institutes of Health Strokes Scale; SBP, systolic blood pressure. Bold indicates P values < 0.05 are statistically significant.
ARTICLE IN PRESS
Median age in years (IQR) Female, n (%) Hypertension, n (%) Hyperlipidemia, n (%) Diabetes mellitus, n (%) Smoking, n (%) Atrial fibrillation, n (%) Median onset-to-treatment time in minutes (IQR) Median door-to-needle time in minutes (IQR) Median NIHSS at onset (IQR) Median SBP at onset in mm Hg (IQR) Median DBP at onset in mm Hg (IQR) Baseline ASPECTS <8, n (%) 24-hour ASPECTS <8, n (%) Median change in ASPECTS (IQR)
Functional independence mRS 0-2 (n = 201)
Multivariate analysis (24-hour ASPECTS)
VALIDATION OF ASPECTS AS OUTCOME PREDICTOR IN STROKE PATIENTS
Table 3. Results of univariate and multivariate analyses: predictors of functional dependence (mRS >2)
5
ARTICLE IN PRESS W.-Y. KONG ET AL.
6
Table 4. Results of univariate and multivariate analyses: predictors of dramatic ASPECTS progression in patients with good baseline ASPECTS 8-10 points Univariate analysis
Multivariate analysis
No dramatic ASPECTS Dramatic ASPECTS Adjusted odds ratio progression (n = 202) progression (n = 34) P value (95% CI) P value Median age in years (IQR) Female, n (%) Hypertension, n (%) Hyperlipidemia, n (%) Diabetes mellitus, n (%) Smoking, n (%) Atrial fibrillation, n (%) Median onset-to-treatment time, in minutes (IQR) Median door-to-needle time, in minutes (IQR) Median NIHSS at onset (IQR) Median SBP at onset in mm Hg (IQR) Median DBP at onset in mm Hg (IQR)
62 (53-72) 68 (33.7%) 133 (65.8%) 122 (60.4%) 50 (24.8%) 58 (28.7%) 57 (28.2%) 165 (135-201)
75 (66-79) 22 (64.7%) 26 (76.5%) 19 (55.9%) 14 (41.2%) 6 (17.6%) 20 (58.8%) 144 (114-181)
<.001 .001 .221 .620 .046 .179 <.001 .019
81 (69-103)
82 (71-111)
.681
13 (9-19) 152 (138-168) 84 (74-92)
21 (16-23) 159 (149-173) 75 (67-92)
<.001 .087 .104
1.05 (1.01-1.09) 3.48 (1.43-8.43)
.013 .006
2.96 (1.16-7.55)
.023
1.17 (1.08-1.26)
<.001
Abbreviations: ASPECTS, Alberta Stroke Program Early CT Score; CI, confidence interval; DBP, diastolic blood pressure; IQR, interquartile range; NIHSS, National Institutes of Health Strokes Scale; SBP, systolic blood pressure. Bold indicates P values < 0.05 are statistically significant.
ASPECTS were significant predictors of mortality and SICH on both univariate and multivariate analyses (Supplementary Tables S1 and S2). In comparing the prognostic utility for predicting unfavorable functional outcome of baseline ASPECTS, 24hour ASPECTS, and change in ASPECTS, receiver operating characteristic curves were constructed. ASPECTS on the 24-hour scan (area under the curve [AUC] = .78, 95% CI: .73-.82, P < .001) was a better predictor compass compared with baseline ASPECTS (AUC = .64, 95% CI: .59.70, P < .001). This significance was further quantified by DeLong’s method (difference in AUC = .14, z = −2.936, P = .001) (Fig 1, B). Change in ASPECTS was also superior to baseline ASPECTS in predicting functional outcome (AUC = .69, 95% CI: .59-.70, P < .001; difference in AUC = .05, z = −2.831, P = .002). Interestingly, a good ASPECTS (8-10 points) was observed on the initial CT in 236 (59%) patients. This score deteriorated in a fair proportion (104 case, 44%) of these patients on their 24-hour CT scan, and 34 (14%) of them developed DAP. Older age (OR: 1.05, 95% CI: 1.01-1.09, P = .013), female gender (OR: 3.48, 95%CI: 1.43-8.43, P = .006), higher NIHSS at onset (OR: 1.17, 95% CI: 1.081.26, P < .001), and diabetes mellitus (OR: 2.96, 95% CI: 1.16-7.55, P = .023) were found to be independently associated with DAP (Table 4). Furthermore, in patients with good baseline ASPECTS, poor 24-hour ASPECTS as well as DAP were associated with unfavorable functional outcome (OR: 5.11, 95% CI: 2.91-8.93, P < .001; OR: 9.91, 95% CI: 3.37-29.19, P < .001, respectively), functional dependence (OR: 6.45, 95% CI:
3.62-11.49, P < .001; OR: 12.23, 95% CI: 4.53-33.03, P < .001, respectively), mortality (OR: 6.70, 95% CI: 2.19-20.47, P < .001; OR: 21.99, 95% CI: 7.98-60.58, P < .001, respectively), and SICH (OR: 9.29, 95% CI: 2.05-42.14, P = .001; OR: 8.57, 95% CI: 2.87-25.59, P < .001, respectively) (Supplementary Table S3).
Discussion ASPECTS on baseline imaging has an irrefutable role in initial patient selection for reperfusion strategies in AIS.10-12 However, our study demonstrates that 24-hour ASPECTS on CT and change in ASPECTS (between the baseline and 24-hour CT) were more accurate predictors of functional outcome at 3 months, when compared with baseline ASPECTS. Furthermore, in the subgroup of patients with initially good baseline ASPECTS, DAP is associated with unfavorable functional outcome, functional dependence, mortality, as well as SICH. Our findings are consistent with prior studies that have established that early, rather than baseline, markers of clinical stroke severity are important predictors of clinical outcomes.24 Accordingly, 24-hour ASPECTS was superior to baseline ASPECTS in predicting clinical outcomes. Studies on clinical stroke severity scores continued to trend scores to 48-hour and beyond, and identified that it was the latent trajectory subgroups based on longitudinal scores that were the best predictors of 90-day outcomes.25 Score trajectory in a radiological score like ASPECTS would be difficult to establish, as it would require at least 3 serial scans, which is not done in most
ARTICLE IN PRESS VALIDATION OF ASPECTS AS OUTCOME PREDICTOR IN STROKE PATIENTS
of the cases. Nevertheless, we compare the scores at these 2 time points (baseline and 24-hours) and established that the 24-hour ASPECTS or change in score from the baseline CT had greater prognostic utility. ASPECTS on the pre-thrombolysis CT scan has been used for risk stratification, estimation of time of symptom onset, and selection of patients eligible for endovascular treatment. Furthermore, it correlates well with functional outcome at 3 months. On the contrary, 24-hour or serial ASPECTS has been evaluated scarcely, perhaps due to factors like performing of magnetic resonance imaging at 24 hours instead of CT scan at some centers, reliability of serial NIHSS scores, or fear of overestimation of ASPECTS due to cerebral edema. Our study suggests that 24-hour as well as serial CT ASPECTS may identify patients who develop SICH to aid in appropriate antithrombotic therapy, serve as reliable predictors of functional outcome at 3 months, and help in early prognostication, planning optimal rehabilitation strategies, as well as proper disposition of AIS patients. These results are concordant with a prior study demonstrating that 24-hour ASPECTS was superior to baseline ASPECTS in predicting clinical outcomes of patients undergoing endovascular treatment for AIS.20 However, there are some important differences in our study design compared with the solitaire flow restoration with the intention for thrombectomy (SWIFT) trial. The SWIFT trial analyses included solely patients with large vessel occlusion undergoing endovascular stroke treatment, whereas in our study we have included a broader cohort of patients (different stroke types) undergoing thrombolysis for anterior circulation stroke. It can be rationalized that the fundamental principle of the ASPECTS score (ordinal categorization of the extent of CT-based infarct signs) is independent of the underlying stroke type (large vessel occlusion versus all strokes) and hence supportive of our findings. One interesting observation in our study was regarding a specific subgroup of patients with good baseline ASPECTS (ASPECTS 8-10), which eventually fared poorly. In our study, almost 50% (117 out of 236) of such patients subsequently experienced unfavorable functional outcomes at 90 days. Approximately, 9% expired and 6% developed SICH. Our study subsequently demonstrated that both 24-hour ASPECTS and DAP (ASPECTS deterioration of 6 points or more) were reliable predictors of poor functional outcome, even in patients with good baseline ASPECTS. Other clinical predictors of unfavorable functional outcome were in accordance with the conventional understanding of stroke pathophysiology. For example, higher systolic blood pressure at onset, presence of diabetes mellitus, higher age, and higher NIHSS scores were found to be independent predictors of poorer outcome in our AIS cohort.26 One surprising finding in our study was the lack of association between poorer outcome and doorto-needle time and onset-to-treatment time.27,28 Perhaps,
7
higher stroke severity (baseline median NIHSS score: 18 points) at presentation counterbalanced these important variables in our study population. Certain limitations of our study need to be acknowledged. Although this study employed a moderately large sample size of 400 patients with clinical outcomes at 90 days, it remained limited by the retrospective design of the study. Furthermore, 10 patients were excluded from the analyses as they could not have the repeat CT scan due to poor neurologic status or death within 24 hours. Other inherent problems with the ASPECTS methodology were also present in our study. For example, we included only anterior circulation strokes, but a small proportion of them might have suffered from a stroke in the posterior circulation. There also remain technical limitations where a CT scan done very early in the course of stroke may not show the changes of ASPECTS demonstrating ischemic changes or associated vasogenic and cytotoxic edema.29 Nonetheless, our findings proposing the utility of serial imaging would compensate for these limitations. In summary, we found that serial changes of tissue injury on routinely acquired imaging, such as non-contrast CT, may provide critical data on the response of brain tissue to varying degrees of reperfusion that has occurred. It may not only reveal the extent of ischemic injury but also outline the trajectory of expected clinical sequelae. All patients receive repeat cerebral imaging at 24 hours after IVT, and this can serve an important role for early identification of malignant infarction and help determine which patients require intensive neurologic monitoring for better resource planning.
Conclusion Although baseline ASPECTS remains important as a patient selection tool for thrombolysis, ASPECTS measured at 24 hours as well as serial change in ASPECTS may demonstrate greater clinical utility as more accurate prognostic markers of clinical outcomes in patients with AIS.
Appendix: Supplementary Material Supplementary data to this article can be found online at doi:10.1016/j.jstrokecerebrovasdis.2017.05.009.
References 1. CDC, NCHS. Underlying cause of death 1999-2013 on CDC WONDER online database Released 2015. Data are from the Multiple Cause of Death Files, 1999-2013, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative Program. Available at: https://wonder.cdc.gov/wonder/ help/ucd.html. Accessed February 3, 2015. 2. Campbell BCV, Mitchell PJ, Kleinig TJ, et al. Endovascular therapy for ischemic stroke with perfusion-imaging selection. N Engl J Med 2015;372:1009-1018.
ARTICLE IN PRESS W.-Y. KONG ET AL.
8 3. Schwamm LH, Pancioli A, Acker JE, et al. Recommendations for the establishment of stroke systems of care: recommendations from the American Stroke Association’s Task Force on the Development of Stroke Systems. Stroke 2005;36:690-703. 4. Alberts MJ, Latchaw RE, Selman WR, et al. Recommendations for comprehensive stroke centers: a consensus statement from the Brain Attack Coalition. Stroke 2005;36:1597-1616. 5. Pexman JH, Barber PA, Hill MD, et al. Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR Am J Neuroradiol 2001;22:1534-1542. 6. Puetz V, Dzialowski I, Hill MD, et al. The Alberta Stroke Program Early CT Score in clinical practice: what have we learned? Int J Stroke 2009;4:354-364. 7. 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. 8. Coutts SB, Demchuk AM, Barber PA, et al. Interobserver variation of ASPECTS in real time. Stroke 2004;35:e103e105. 9. Dzialowski I, Hill MD, Coutts SB, et al. Extent of early ischemic changes on computed tomography (CT) before thrombolysis: prognostic value of the Alberta Stroke Program Early CT Score in ECASS II. Stroke 2006;37:973978. 10. Hill MD, Rowley HA, Adler F, et al. PROACT-II Investigators. Selection of acute ischemic stroke patients for intra-arterial thrombolysis with pro-urokinase by using ASPECTS. Stroke 2003;34:1925-1931. 11. Tsivgoulis G, Saqqur M, Sharma VK, et al. Association of pretreatment ASPECTS scores with tPA-induced arterial recanalization in acute middle cerebral artery occlusion. J Neuroimaging 2008;18:56-61. 12. Dewey HM, Donnan GA, Freeman EJ, et al. Interrater reliability of the National Institutes of Health Stroke Scale: rating by neurologists and nurses in a community-based stroke incidence study. Cerebrovasc Dis 1999;9:323-327. 13. Goldstein LB, Samsa GP. Reliability of the National Institutes of Health Stroke Scale. Extension to nonneurologists in the context of a clinical trial. Stroke 1997;28:307-310. 14. Côté R, Battista RN, Wolfson C, et al. The Canadian Neurological Scale: validation and reliability assessment. Neurology 1989;39:638-643. 15. Hantson L, De Weerdt W, De Keyser J, et al. The European Stroke Scale. Stroke 1994;25:2215-2219. 16. Govan L, Langhorne P, Weir CJ. Categorizing stroke prognosis using different stroke scales. Stroke 2009;40:3396-3399. 17. Al-Ajlan FS, Goyal M, Demchuk AM, et al. Intra-arterial therapy and post-treatment infarct volumes: insights from
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28.
29.
the ESCAPE randomized controlled trial. Stroke 2016;47:777-781. 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. MacCallum C, Churilov L, Mitchell P, et al. Low Alberta Stroke Program Early CT Score (ASPECTS) associated with malignant middle cerebral artery infarction. Cerebrovasc Dis 2014;38:39-45. Liebeskind DS, Jahan R, Nogueira RG, et al. Serial Alberta Stroke Program Early CT Score from baseline to 24 hours in solitaire flow restoration with the intention for thrombectomy (SWIFT) study: a novel surrogate end point for revascularization in acute stroke. Stroke 2014;45:723727. Aviv RI, Mandelcorn J, Chakraborty S, et al. Alberta Stroke Program Early CT Scoring of CT perfusion in early stroke visualization and assessment. AJNR Am J Neuroradiol 2007;28:1975-1980. Chimowitz MI. Endovascular treatment for acute ischemic stroke–still unproven. N Engl J Med 2013;368:952-955. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837-845. Broderick JP, Lu M, Kothari R, et al. Finding the most powerful measures of the effectiveness of tissue plasminogen activator in the NINDS tPA stroke trial. Stroke 2000;31:2335-2341. Sajobi TT, Menon BK, Wang M, et al. Early trajector of stroke severity predicts long-term functional outcomes in ischemic stroke subjects. Results from the ESCAPE trial (Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion With Emphasis on Minimizing CT to Recanalization Times). Stroke 2017;48: 105-110. Demchuk AM, Goyal M, Menon BK, et al. Endovascular treatment for Small Core and Anterior circulation Proximal occlusion with Emphasis on minimizing CT to recanalization times (ESCAPE) trial: methodology. Int J Stroke 2015;10:429-438. Wardlaw JM, Murray V, Berge E, et al. Recombinant tissue plasminogen activator for acute ischaemic stroke: an updated systematic review and meta-analysis. Lancet 2012;39:2364-2372. National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke: the National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med 1995;333:1581-1587. Butcher KS, Lee SB, Parsons MW, et al. Differential prognosis of isolated cortical swelling and hypoattenuation on CT in acute stroke. Stroke 2007;38:941-947.