The Combination of Clinical Features, Transcranial Doppler, and Alberta Stroke Program Early Computed Tomography Score (Computed Tomography Angiography) in Predicting Outcome in Intravenous Recombinant Tissue Plasminogen Activator-Treated Patients

ARTICLE IN PRESS

The Combination of Clinical Features, Transcranial Doppler, and Alberta Stroke Program Early Computed Tomography Score (Computed Tomography Angiography) in Predicting Outcome in Intravenous Recombinant Tissue Plasminogen Activator-Treated Patients Maher Saqqur, MD, MPH, FRCPC,* Esseddeeg Ghrooda, MD,* Aftab Ahmad, Khurshid Khan, MD, FRCPC,* Muhammad S. Hussain, MD,‡ and Ashfaq Shuaib, MD, FRCPC*

MBBS,†

Background: Little data exist on using combined baseline clinical neuroimaging and transcranial Doppler (TCD) information in predicting clinical outcome in stroke patients treated with intravenous (IV) thrombolysis. Methods: Stroke patients received IV recombinant tissue plasminogen activator (rt-PA) and had diagnostic TCD within 3 hours of symptom onset. The TCD result was interpreted using the thrombolysis in brain ischemia (TIBI) flow grading system. Following multiple regression analysis, a grading system was created with 1 point for each of the following: National Institutes of Health Stroke Scale (NIHSS) score of 16 or higher, TIBI score of 1 or lower, and Alberta Stroke Program Early CT Score (ASPECTS) of 6 or lower. The patients’ scores were compared to modified Rankin Scale (mRS) scores at 90 days. Results: A total of 349 patients were included. In unvaried analysis, age of 80 years or older (P = .002), an ASPECTS of 6 or lower (P < .001), an NIHSS score of 16 or higher (P < .001), a TIBI score of 1 or lower (P < .001), and a glucose level ≥ 200 mg/dl (P = .04) were associated with poor outcome (mRS score > 2). In the multiple regression analysis, age of 80 years or older, an ASPECTS of 6 or lower, an NIHSS score of 16 or higher, and hyperglycemia were predictors of poor outcome (P < .05). Based on our scoring system, the patients’ odds ratios for poor outcome were 7 (95% confidence interval [CI]: 2-23, P = .003), 8 (95% CI: 3-25, P < .001), and 24 (95% CI: 4-151, P = .001) for scores of 1, 2, and 3, respectively, after adjustment for common stroke risk factors. The mean time to recanalization increased as the score increased (score of 0: 160 ± 45 minutes versus score of 3: 186 ± 38 (P = .70). Conclusion: A multimodal grading system is useful in predicting outcome in patients treated with IV rt-PA. Those with higher scores might be candidates for interventional therapy. Key Words: TCD—ultrasound—stroke—thrombolysis—outcome—neuroimaging. © 2016 Published by Elsevier Inc. on behalf of National Stroke Association.

From the *Department of Medicine (Neurology), University of Alberta, Edmonton, Alberta, Canada; †Division of Neurology, Department of Medicine, National University Health System, National Neuroscience Institute, Singapore; and ‡Endovascular Surgical Neuroradiology, Cleveland Clinic Foundations, Cleveland, Ohio. Received September 30, 2015; accepted December 9, 2015. Address corespondence to Maher Saqqur, MD, MPH, FRCPC, Department of Medicine, Division of Neurology, University of Alberta, Edmonton, Alberta, Canada. E-mail: [email protected]. 1052-3057/$ - see front matter © 2016 Published by Elsevier Inc. on behalf of National Stroke Association. http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2015.12.013

Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2016: pp ■■–■■

1

ARTICLE IN PRESS M. SAQQUR ET AL.

2

Background Intravenous (IV) recombinant tissue plasminogen activator (rt-PA) remains the only treatment approved for acute stroke treatment in the 3-hour window based in the National Institute of Neurological Disorders and Stroke rt-PA Stroke Study,1 with the number needed to treat of 8 to reverse 1 stroke completely at 3 months. However, half of rt-PA-treated patients remain severely disabled or die within this period of time. The modest recanalization rate of 25%-30% observed with proximal largevessel occlusion may explain the limited effect of systemic thrombolysis alone and the poor outcome apparently associated with persisting occlusion.2 Lately, there were several major trials that showed there is a major role for interventional mechanical thrombectomy once the IV thrombolysis failed to open the vessel.3-5 Several clinical scoring systems have been developed to predict outcome in IV tissue plasminogen activator (tPA) treatment patients.6-8 However, these scores did not take into account functional imaging such as transcranial Doppler (TCD). While each score adds useful information, little data exists on using combined information from these modalities to predict clinical outcome. Our study’s aim is to develop a multimodal grading system, from information available at initial presentation, to predict outcome in stroke patients receiving IV t-PA.

Methods The Combined Lysis of Thrombus in Brain Ischemia with Transcranial Ultrasound and Systemic TPA (CLOTBUST) databank was used to develop a multimodal grading system, from information available at initial presentation, to predict outcome in stroke patients receiving IV t-PA. Our multicenter database was designed using the CLOTBUST trial methodology.9 The databank was a retrospective study of patients who presented with acute stroke and underwent systemic thrombolysis and had an intracranial arterial occlusion on their TCD at 4 stroke centers (University of Texas–Houston, Universitat Autònoma de Barcelona, University of Calgary, and University of Alberta) from 2000 to 2003. The patients received either continuous or intermittent 2-MHz TCD assessment of recanalization within 2 hours of t-PA bolus. All patients had evidence of obstructive residual flow signals in proximal intracranial arteries on baseline TCD assessment before t-PA bolus. An experienced physician-sonographer diagnosed these occlusions using previously validated criteria.10,11 Arterial recanalization was determined using the thrombolysis in brain ischemia system by the site investigators who gave t-PA and monitored residual blood flow signals, and based on our previous published criteria.12 Recanalization was determined at 2 hours following t-PA bolus.

Neurological status (NIHSS stroke score) was repeatedly assessed at baseline and during the first 2 hours after rt-PA bolus by the treating neurologist who, despite not being directly involved in TCD performance, was informed about worsening of flow signals on diagnostic TCD, if these occurred. All neurologists who performed serial neurological examinations in the emergency room were certified in the NIHSS scoring. The NIHSS scores at 24 hours and modified Rankin Scale (mRS) scores at 3 months were obtained by a neurologist who was not aware of the TCD findings and the purposes of the present study. Poor long-term outcome was defined as an mRS score of 3-6 at 3 months’ follow-up in the CLOTBUST trial or at 3-5 months’ follow-up clinic visit or telephone call in the rest of the cases. Safety was assessed by detecting symptomatic intracerebral hemorrhage within 72 hours of the onset of stroke. Symptomatic intracerebral hemorrhage was defined as brain imaging evidence of intracerebral hemorrhage with clinical worsening by an increase in the NIHSS score of 4 points or higher.9

Statistical Analysis Univariate and multivariate regression analyses were performed to determine which clinical, radiological, and TCD factors were significant and, based on these factors, a scoring system was created. Following multiple regression analysis, a grading system was created with 1 point for each of the following: baseline NIHSS score of 16 or higher, proximal arterial occlusion (M1, terminal internal carotid artery, and tandem internal carotid artery/ middle cerebral artery), and Alberta Stroke Program Early CT Score (ASPECTS) of 6 or lower. The patients’ scores were compared to the recanalization rate, 24 hours’ NIHSS score, and mRS scores at 90 days. The Statistical Package for Social Science (version 17 for Windows; SPSS, Inc.) was used for statistical analyses.

Results A total of 342 patients with acute anterior circulation ischemic stroke treated in 4 academic centers were included. The median NIHSS score is 16 (interquartile range: 3-34). The mean time to IV rt-PA treatment is 134 ± 32 minutes from symptom onset. The mean time to TCD is 129 ± 40 minutes. The mean systolic blood pressure is 157 ± 22 mmHg and the baseline glucose level is 148 ± 73 mg/dL. Unvaried analysis for poor long-term outcome (mRS score ≥ 3) as a dependent variable is summarized in Table 1. In the multiple logistic regression analysis, an ASPECTS of 6 or lower, an NIHSS score of 16 or higher, and proximal occlusion on baseline TCD (terminal internal carotid artery, M1, tandem internal carotid artery/middle cerebral artery occlusion) were significantly correlated with poor outcome (P < .05). Based on these criteria, a baseline computed tomography angiography (CTA) score

ARTICLE IN PRESS COMBINE CLINICAL, TCD AND CT SCORE IN PREDICTING OUTCOME IN STROKE

3

Table 1. Univariate statistics between patients with and without poor long-term outcome Factors

Poor (N = 148)

Good (N = 140)

P value

Age (mean ± SD, years) 80 years or older Younger than 80 years Sex, n (%) Male Female Baseline NIHSS score (mean ± SD) Lower than 16, n (%) 16 or higher, n (%) Time to rt-PA treatment (mean ± SD, min) SBP (mean ± SD, mmHg) Glucose level (mean ± SD, mg/dL) HTN No, n (%) Yes, n (%) TOAST classification, n (%) CE (1) LVA (2) Other (3-5) Type of occlusion, n (%) MCA M1 (1) MCA M2 (2) TICA + tandem ICA/MCA (3-4) TCD flow finding, n (%) 1 2 3 TIBI, n (%) 0 1 2 3 ASPECTS, n (%) 6 or lower Higher than 6

71.5 ± 12.4 19 of 62 (31%) 121 of 226 (54%)

65.2 ± 12.8 43 of 62 (69%) 105 of 226 (47%)

<.001 .002

76 (50.0) 72 (52.9) 18.2 ± 4.5 37 (30.1) 111 (67.3) 137.0 ± 31.8 159.1 ± 24.3 159.9 ± 78.3

76 (50.0) 64 (47.1) 14.1 ± 5.4 86 (69.9) 54 (32.7) 131.6 ± 32.8 154.0 ± 20.3 137.5 ± 69.7

45 (46.9) 87 (54.0)

51 (53.1) 74 (46.0)

.30

42 (60.0) 71 (53.0) 35 (42.2)

28 (40.0) 63 (47.0) 48 (57.8)

.08

86 (59.7) 26 (31.0) 36 (60.0)

58 (40.3) 58 (69.1) 24 (40.0)

<.001

107 (64.5) 22 (57.9) 19 (22.6)

59 (35.5) 16 (42.1) 65 (77.4)

<.001

43 (59.7) 54 (56.8) 34 (51.5) 17 (30.9)

29 (40.3) 41 (43.2) 32 (48.5) 38 (69.1)

.01

17 (81.0) 55 (44.7)

4 (19.0% 68 (55.3)

<.01

.64 <.001 <.001 .16 .05 .01

Abbreviations: ASPECTS, Alberta Stroke Program Early Computed Tomography Score; CE, Cardioembolic; HTN, hypertension; ICA, internal carotid artery; LVA, Large vessel atherosclerosis disease; MCA, middle cerebral artery; NIHSS, National Institutes of Health Stroke Scale; rt-PA, recombinant tissue plasminogen activator; SBP, systolic blood pressure; SD, standard deviation; TCD, transcranial Doppler; TIBI, thrombolysis in brain ischemia; TICA, terminal internal carotid artery; TOAST, Trial of Org 10172 in Acute Stroke Treatment.

(0-3) was developed based on the following points (1 point for each criteria): ASPECT score of 6 or lower, baseline NIHSS score of 16 or higher, and the presence of proximal occlusion on baseline TCD. The combined score was calculated in 182 patients in whom outcome data were available. In patients with a CTA score of 0, the probabilities are 17.6% (6 of 34 patients) and 52.3% (20 of 38 patients) for poor outcome and complete recanalization, respectively, after IV thrombolysis, whereas a CTA score of 3 has probabilities of 85.7% (12 of 14 patients) and 11% (2 of 18 patients), respectively (Table 2). Based on our scoring system, the patients’ odds ratios for poor outcome were 7 (95% confidence interval: 2-23,

P = .003), 8 (3-25, P < .001), and 24 (95% confidence interval: 4-151, P = .001) for scores of 1, 2, and 3, respectively, after adjustment for stroke risk factors (age, sex, systolic blood pressure, glucose level, and onset to IV rtPA treatment). The mean time to recanalization in patients with a score of 0 is 160 ± 45 minutes from treatment onset, whereas a score of 1 is 158 ± 51 minutes, a score of 2 is 160 ± 52 minutes, and a score of 3 is 186 ± 38 minutes (P = .7).

Discussion A multimodal grading system is useful in predicting outcome in patients treated with IV rt-PA. In particular,

ARTICLE IN PRESS M. SAQQUR ET AL.

4

Table 2. CTA score probabilities and adjusted OR of poor outcome and complete recanalization in IV rt-PA-treated patients

CTA score

Probability of poor outcome (mRS score >2)

Adjusted OR of poor outcome

0 1 2 3

18% (6 of 34 patients) 52% (17 of 33 patients) 59% (37 of 63 patients) 86% (12 of 14 patients)

7 (2-23) 8 (3-26) 24 (4-151)

P value

Probability of complete recanalization

Adjusted OR of complete recanalization

P value

.001 .003 <.001 .001

53% (20 of 38 patients) 40% (19 of 47 patients) 22% (17 of 79 patients) 11% (2 of 18 patients)

.4 (.16-1.2) .16 (.06-.4) .12 (.02-.6)

.001 .113 <.001 .014

Abbreviations: CTA, computed tomography angiography; IV, intravenous; mRS, modified Rankin Scale; OR, odds ratio; rt-PA, recombinant tissue plasminogen activator.

our scoring system utilizes information available at the time of initial assessment and may be useful to guide further acute treatment decisions. We present the CTA score, which reliably and validly predicts the outcome of IV alteplase-treated ischemic stroke patients. Those with higher scores have a higher probability of poor outcome and might benefit from further interventional therapy. Future prospective studies will be helpful in validating our CTA score. The present study demonstrates that in patients with a CTA score of 0, the probabilities are 17.6% and 52.3% for poor outcome and complete recanalization, respectively, after IV thrombolysis, whereas a CTA score of 3 has probabilities of 85.7% and 11%, respectively. Specificity of the prediction at both extreme poles of the score reaches up to 100%, and the likelihood ratios are correspondingly high. Maximum possible specificity minimizes the possibility of incorrect prediction of dismal prognosis,7 which gains more importance when further interventional therapy with invasive add-on rescue strategies such as Merci catheter and Solitaire retrieval device, which are not risk-free, are being considered.4,5 From our study we can postulate that if patients with acute ischemic stroke have low CTA scores (0), there is a high probability that IV rt-PA will do the job, whereas for patients with high CTA scores (3), the best plan of action is to consider further rescue interventional treatment as the probability of clinical recovery with arterial recanalization is low. There have been several attempts in the past few years to predict long-term clinical outcome in ischemic stroke patients using several forms of clinical score.13,14 The earliest score comes from the analysis of the National Institute of Neurological Disorders and Stroke trial. This score was derived to predict outcome using baseline, 2- and 24hour, and 7- to 10-day data, and included the total baseline NIHSS score and a history of atrial fibrillation.7 Data from the placebo arms of Parts 1 and 2 of the National Institute of Neurological Disorders and Stroke rt-PA Stroke Trial were used to identify variables that could predict a poor outcome, defined as a moderately severe disability, severe disability, or death (mRS score >3) 3 months

after stroke. Frankel et al7 conclude that predicted poor outcome were an NIHSS score higher than 17 plus atrial fibrillation. Whereas, the predictor at 24 hours was an NIHSS score higher than 22 and at 7-10 days was an NIHSS score higher than 16. The second clinical scoring system, the hemorrhage after thrombolysis score, was developed as a tool to predict the risk of symptomatic intracerebral hemorrhage after IV rt-PA.8 This score consists of 3 parameters, which are the baseline glucose level (or history of diabetes), the baseline NIHSS score, and early infarct signs on the admission head computed tomography scan. Lou et al8 concluded that the hemorrhage after thrombolysis score allows reasonable risk stratification of intracerebral hemorrhage after IV rt-PA. However, the prognostic value of this scale and its use to predict the net benefit from rt-PA need to be refined and prospectively confirmed in a larger cohort of patients before they can be used in clinical decision making. Then the DRAGON score has been reported recently in literature.15 The DRAGON score (0-10 points) consists of (hyper)dense cerebral artery signs/early infarct signs on admission computed tomography scan (both = 2, either = 1, none = 0), a prestroke mRS score higher than 1 (yes = 1), age (≥80 years = 2, 65-79 years = 1, < 65 years = 0), glucose level at baseline (>8 mmol/L [>144 mg/dL] = 1), onset-to-treatment time (>90 minutes = 1), and baseline NIHSS score (>15 = 3, 10-15 = 2, 5-9 = 1, 0-4 = 0). Strbian et al15 report the proportions of patients with very poor outcome (mRS score 5-6) were 0%, 2%, 5%, 70%, and 100% for 0-1, 2, 3, 8, and 9-10 points, respectively. One of the DRAGON parameters, the prestrike mRS score, could be problematic as the application of the mRS before a stroke is inconsistent with its intended use and is not defined. To distinguish between mRS scores 0, 1, and 2, comparison of the patient’s functional abilities before and after a stroke (or potentially other brain injury) is necessary. This comparison is impossible before a stroke in most patients, forcing the raters to make subjective and undefined assumptions resulting in suboptimal reproducibility.13

ARTICLE IN PRESS COMBINE CLINICAL, TCD AND CT SCORE IN PREDICTING OUTCOME IN STROKE

Although widely used, multivariate analysis of complex datasets can only go so far in addressing independence of factors because these datasets often fail to meet the underlying assumptions necessary for their valid use. While there is considerable interest in stroke outcome predictive scores (e.g., iScore, diffusion/perfusion mismatch), many have been disappointing in predicting actual treatment response because they are typically composed of factors merely related to outcome.16 The ultimate goal for all described scores is to develop a valid and reliable score based mainly on parameters available at or shortly before administration of IV alteplase to help in the evaluation of the acute situation so the physician can deliver accurate information and the patient and relatives can make a decision.17 In the CTA score, we tried to derive the score based on imaging, blood flow, and clinical modalities. We did prove the concept that using 1 modality is not enough and adding them together will make the scoring system stronger and reliable. Further studies are needed to confirm our findings on multiple center levels. Our study has the following limitations. First, our CTA scoring tool uses TCD, which carries some interoperator variability, and some centers were not trained to use TCD in an acute setting. For this reason, we encourage all stroke centers to be trained in using TCD. Our scoring system was tried over multicenters and showed close and similar performance with nonsignificant variability. Although the scoring system takes up to 5 minutes to perform with a skilled professional, it is valid and reliable in predicting outcome and helpful in acute setting decision making. Second, our study is a retrospective one and many cases have missing follow-up data, which makes our result subject to bias and confounding effects. In conclusion, a multimodal grading system is useful in predicting outcome in patients treated with IV rt-PA. Those with higher scores might benefit from further interventional therapy.

References 1. 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. 2. Saqqur M, Uchino K, Demchuk AM, et al. Site of arterial occlusion identified by transcranial Doppler predicts the response to intravenous thrombolysis for stroke. Stroke 2007;38:948-954.

5

3. 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. 4. Diener HC, Nitschmann S. [Endovascular treatment for acute ischemic stroke: Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN)]. Internist (Berl) 2015;56:847-850. 5. Goyal M, Demchuk AM, Menon BK, et al. Randomized assessment of rapid endovascular treatment of ischemic stroke. N Engl J Med 2015;372:1019-1030. 6. 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. 7. Frankel MR, Morgenstern LB, Kwiatkowski T, et al. Predicting prognosis after stroke: a placebo group analysis from the National Institute of Neurological Disorders and Stroke rt-PA Stroke Trial. Neurology 2000;55:952-959. 8. Lou M, Safdar A, Mehdiratta M, et al. The hat score: a simple grading scale for predicting hemorrhage after thrombolysis. Neurology 2008;71:1417-1423. 9. Alexandrov AV, Molina CA, Grotta JC, et al. Ultrasoundenhanced systemic thrombolysis for acute ischemic stroke. N Engl J Med 2004;351:2170-2178. 10. Demchuk AM, Christou I, Wein TH, et al. Accuracy and criteria for localizing arterial occlusion with transcranial Doppler. J Neuroimaging 2000;10:1-12. 11. Demchuk AM, Christou I, Wein TH, et al. Specific transcranial Doppler flow findings related to the presence and site of arterial occlusion. Stroke 2000;31:140-146. 12. Burgin WS, Malkoff M, Felberg RA, et al. Transcranial Doppler ultrasound criteria for recanalization after thrombolysis for middle cerebral artery stroke. Stroke 2000;31:1128-1132. 13. Bruno A, Switzer JA. Predicting outcome of iv thrombolysis-treated ischemic stroke patients: the dragon score. Neurology 2012;79:486-487, author reply 486-487. 14. Fargen KM, Chaudry I, Turner RD, et al. A novel clinical and imaging based score for predicting outcome prior to endovascular treatment of acute ischemic stroke. J Neurointerv Surg 2013;5(Suppl 1):i38-i43. 15. Strbian D, Meretoja A, Ahlhelm FJ, et al. Predicting outcome of iv thrombolysis-treated ischemic stroke patients: the DRAGON score. Neurology 2012;78:427432. 16. Kent TA. Predicting outcome of iv thrombolysis-treated ischemic stroke patients: the dragon score. Neurology 2012;78:1368. 17. Turk AS, Magarick JA, Frei D, et al. Ct perfusion-guided patient selection for endovascular recanalization in acute ischemic stroke: a multicenter study. J Neurointerv Surg 2013;5:523-527.