Higher Fasting Glucose Next Day after Intravenous Thrombolysis Is Independently Associated with Poor Outcome in Acute Ischemic Stroke Wenjie Cao, MD, Yifeng Ling, MD, Fei Wu, MD, Lumeng Yang, MD, Xin Cheng, MD, PhD, and Qiang Dong, MD, PhD
Background: We aimed to test the outcome-predictive power of routine fasting glucose (FG) obtained at second day after onset in intravenous thrombolysis (IVT) acute ischemic stroke (AIS) patients. Methods: We identified AIS patients presenting to our institution between December 2011 and July 2013 within 4.5 hours of onset, who received admission glucose (AG) before IVT, FG, and glycated hemoglobin (HbA1c) the second day after admission, from our prospectively recorded stroke database. Multivariate logistic regression was used to assess the association of FG and 90-day modified Rankin Scale (mRS). Results: Between December 2011 and July 2013, a total of 166 AIS patients received intravenous plasminogen activator. Of those, 119 patients who have AG before IVT, FG, and HbA1c the second day were included in the study. FG independently predicted 90-day clinical unfavorable outcome (mRS, 3-6 with an odds ratio of 1.576; 95% confidence interval [CI], 1.0532.358; P 5 .027). This association was not significant in AG (P 5 .714), HbA1c (P 5.655), and history of diabetes (P 5.547). In receiver operating characteristic analysis, increased FG was associated with 90-day mRS (3-6) with an area under curve of .72, (95% CI, .65-.9; P 5.001). Conclusions: FG is a powerful predictor associated with the outcome in IVT-treated AIS patients independent of AG and HbA1c. Key Words: Admission glucose—fasting glucose—acute ischemic stroke—thrombolysis. Ó 2015 by National Stroke Association
Introduction Admission glucose (AG) level and history of diabetes mellitus (DM) are associated with poor clinical outcome after intravenous thrombolysis (IVT).1,2 However, a recent study found that IVT-treated patients with poststroke hyperglycemia (PSH) assessed with AG did not have significantly worse outcomes.3 Various physiologic From the Department of Neurology and Institute of Neurology, Huashan Hospital, Fudan University, Shanghai, China. Received June 22, 2014; revision received July 28, 2014; accepted July 30, 2014. W.C. and Y.L. contributed equally to this study. Address correspondence to Qiang Dong, MD, PhD, Department of Neurology, Huashan Hospital, No 12 Wulumuqi Zhong Rd, Shanghai 200040, China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.07.029
100
statuses, such as dietary intake or using insulin before onset, can impact AG level besides PSH and finally influence the association between PSH and outcome. Unfortunately, this issue was difficult to be controlled in clinical practice. The purpose of our study was to test the outcomepredictive strength of routine fasting glucose (FG) the second day morning after admission in IVT-treated stroke patients. We hypothesized that FG in a similar physiologic status could be more reliable to evaluate the pathologic glucose metabolism in acute ischemic stroke (AIS) patients than AG and therefore predict the outcome.
Methods Patients We identified AIS patients presenting to our institution between December 2011 and July 2013 within 4.5 hours of
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 1 (January), 2015: pp 100-103
FG PREDICTS STROKE OUTCOME
101
Table 1. Comparison of patient characteristics in those with good versus poor outcomes 90-day functional outcome
N Age, mean (SD), y Male, % Hypertension, % Diabetes, % Dyslipidemia, % Previous stroke, % Smoking, % OTT, mean (SD), min Baseline NIHSS, mean (SD) SBP, mean (SD), mm Hg DBP, mean (SD), mm Hg TOAST, % Large vessel disease Small vessel disease Cardioembolism Other and undetermined AG, mean (SD), mmol/L FG, mean (SD), mmol/L HbA1c, %
Total
mRS 0-2
mRS .2
119 66 (57-74) 65.5 (78/119) 65.5 (78/119) 31.9 (38/119) 19.3 (23/119) 14.3 (17/119) 42 (50/119) 180 (132-234) 10 (4-15) 150 (140-163) 87 (80-92)
72 63 (55-72) 69.4 (50/72) 66.7 (48/72) 23.6 (17/72) 15.3 (11/72) 9.7 (7/72) 45.8 (33/72) 184 (142-255) 6 (3-10) 147 (138-161) 86 (80-91)
47 71 (61-77) 59.6 (28/47) 63.8 (30/47) 44.7 (21/47) 25.5 (12/47) 21.3 (10/47) 36.2 (17/47) 155 (124-216) 15 (10-18) 157 (140-170) 87 (80-92)
34.7 (41/118) 27.1 (32/118) 34.7 (41/118) 3.4 (4/118) 7 (5.9-8.3) 6.1 (5.2-7.7) 6 (5.5-6.7)
38 (27/71) 35.2 (25/71) 22.5 (16/71) 4.2 (3/71) 6.4 (5.7-7.6) 5.8 (5.5-6.4) 5.8 (5.5-6.4)
29.8 (14/47) 14.9 (7/47) 53.2 (25/47) 2.1 (1/47) 7.8 (6.3-11.1) 7.6 (5.5-11.1) 6.3 (5.8-7.8)
P value
.027 .325 .844 .026 .235 .108 .345 .129 ,.001 .139 .696 .005
.001 ,.001 .002
Abbreviations: AG, admission glucose; DBP, diastolic blood pressure; FG, fasting glucose; HbA1c, glycated hemoglobin; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OTT, onset to treatment time; SD, standard deviation; SBP, systolic blood pressure; TOAST, Trial of Org 10172 in Acute Stroke Treatment.
onset, who received AG before intravenous plasminogen activator (IV-tPA), FG, and glycated hemoglobin (HbA1c) the second day after admission. We supplied the meal according to the swallow test if the patients were admitted before 8 pm on the first day. All patients were limited on bed in the intensive care unit during the first 24 hours after IV-tPA. Thus, the situation of dietary intake and bed rest level were similar among patients. Baseline National Institutes of Health Stroke Scale (NIHSS), 90-days mRS (modified Rankin Scale), blood pressure, and onset to treatment time were obtained from the database. Histories of hypertension, dyslipidemia, DM, previous stroke, and smoking were recorded at the admission. Stroke etiology was classified according to the Stop Stroke Study Trial of Org 10172 in Acute Stroke Treatment based on the following magnetic resonance imaging and computer tomography angiography after admission. The study was performed with the informed consent of the patients or their next of kin and with ethical approval from the Institutional Review Board of Huashan Hospital.
Statistics Analysis Statistical analyses were performed using SPSS, version 20 (SPSS Inc, Chicago, IL). P less than .05 was considered to indicate statistical significance. Patients were dichotomized by using the 90-day mRS score into good (mRS
score 0-2) versus poor outcome (mRS score 3-6). Differences of patients’ characteristics between outcomes were tested by the Fishers exact test for categorical and the Mann–Whiney test for continuous value. Multivariate binary logistic regression (including variables with P value ,.2) was used to assess the association of AG, HbA1c, and FG with 90-day poor outcome (mRS 3-6). Receiver operating characteristic analysis was performed between FG and poor outcome.
Results Between December 2011 and July 2013, a total of 166 AIS patients received IV-tPA. Of those, 119 patients have AG before IV-tPA, FG, and HbA1c the second day after admission. Patient characteristics (total and stratified by outcome) are listed in Table 1. As expected, increased age and higher baseline NIHSS, history of DM and atrial fibrillation, higher AG, FG, and HbA1c were significantly associated with worse functional outcome. Multivariate logistic regression analysis between variables (P value ,.2) and 90-day mRS was listed in Table 2. Previous stroke, onset to treatment time, baseline NIHSS, and FG was independently associated with outcome after adjust DM history, AG, HbA1c, systolic blood pressure, atrial fibrillation, and age. FG predicted favorable outcome with an odds ratio of 1.576 (95% confidence interval, 1.053-2.358; P 5.027). This association was
W. CAO ET AL.
102
not significant in AG (P 5.713), HbA1c (P 5.657), and history of DM (P 5 .549). In receiver operating characteristic analysis, increased FG was associated with 90-day mRS (3-6) with an area under curve of .72, (95% confidence interval, .65-.9; P 5 .001; Fig 1).
Discussion In this present study, increased FG was independently associated with poor outcome. However, no significant associations were found between history of DM, AG, or HbA1c and outcome after adjusting FG and vascular risks. Those results demonstrate FG is a powerful outcome predictor in AIS patients who received IVT. PSH is associated with poor clinical outcome and death.4 AG, commonly used to evaluate PSH, however, can be influenced by physiologic blood glucose frustration at admission. Data of dietary intake and insulin using before stroke onset were not controlled in prior PSH studies, which may affect the post-IVT clinical outcomes. Insignificant association between AG and clinical outcomes in IVT-treated patients has been reported in 1 study.3 Those opposite results need further data to reveal a correct association between hyperglycemia and outcome. Many prior AG-based PSH studies have notified the limitations of preonset data did not include the information of using insulin and dietary intake.3,5 On the other hand, AG over 8 mmol/L in predicting IVT outcome has been currently included in 2 population-based tools of DRAGON and iScore, which are widely used to predict the outcome in IVT patients.6,7 As a sort of random glucose test, AG over 8 mmol/L can also be observed in AIS patients without glucose metabolic disorder who onset within 2 hours after diet. Such phenomenon can definitely be a bias but difficult to be controlled in PSH studies. Highlight of our study was further setting FG and HbA1c to evaluate the long-term and short-term glucose metabolic status in IVT-treated patients. Previous study has confirmed HbA1c is an independent predictor for outcome, after adjusted AG.8 However, the strength of FG in predicting outcome was just analyzed in a highly selected non-DM AIS population.5 Notably, our study found increased FG was significantly associated with poor outcome independent of AG and HbA1c. FG obtained in a similar physiologic status is a more reliable predictor, advancing over AG and HbA1c, by balancing the various physiologic glucose metabolisms before onset. We therefore suggest FG the second day after admission should be included in IVT outcome prediction. Limitations of our study are because of the small sample size and derived from a single-center experience. The bed rest level and the situation of dietary intake at night of first day were not analyzed in detail, which could impact
Table 2. Multivariate logistic regression for 90-day poor clinical outcome (mRS 3-6) Predictors
OR
95% CI
P value
Age Diabetes Previous stroke OTT Baseline NIHSS SBP TOAST HbA1c AG FG
.985 .602 8.9 .991 1.228 .996 1.624 1.264 1.077 1.576
.941-1.031 .116-3.133 1.904-41.594 .983-.999 1.111-1.357 .968-1.025 .865-3.049 .452-3.532 .726-1.597 1.053-2.358
.511 .547 .005 .024 ,.001 .787 .131 .655 .714 .027
Abbreviations: AG, admission glucose; CI, confidence interval; FG, fasting glucose; HbA1c, glycosylated hemoglobin A1; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; OTT, onset to treatment time; SBP, systolic blood pressure; TOAST, Trial of Org 10172 in Acute Stroke Treatment.
the real FG level. The vessel recanalization in acute large artery occlusive patients and the size of the infarctions were also not assessed in present study.
Summary FG is a powerful predictor associated with the outcome in IVT-treated AIS patients independent of AG and HbA1c. Further studies are necessary to better define the relationship between FG and outcome in a randomized controlled trail.
Figure 1. Receiver operating characteristic (ROC) curve of fasting glucose in predicting 90-day poor outcome. Abbreviations: AUC, area under curve; CI, confidence interval.
FG PREDICTS STROKE OUTCOME
103
References 1. Ahmed N, Davalos A, Eriksson N, et al. Association of admission blood glucose and outcome in patients treated with intravenous thrombolysis: results from the Safe Implementation of Treatments in Stroke International Stroke Thrombolysis Register (SITS-ISTR). Arch Neurol 2010;67:1123-1130. 2. Bruno A, Levine SR, Frankel MR, et al. Admission glucose level and clinical outcomes in the NINDS rt-PA Stroke Trial. Neurology 2002;59:669-674. 3. Meurer WJ, Scott PA, Caveney AF, et al. Lack of association between hyperglycaemia at arrival and clinical outcomes in acute stroke patients treated with tissue plasminogen activator. Int J Stroke 2010;5:163-166. 4. Desilles JP, Meseguer E, Labreuche J, et al. Diabetes mellitus, admission glucose, and outcomes after stroke throm-
5.
6.
7.
8.
bolysis: a registry and systematic review. Stroke 2013; 44:1915-1923. Sykora M, Diedler J, Poli S, et al. Association of nondiabetic hyperglycemia with autonomic shift in acute ischaemic stroke. Eur J Neurol 2012;19:84-90. Strbian D, Seiffge DJ, Breuer L, et al. Validation of the DRAGON score in 12 stroke centers in anterior and posterior circulation. Stroke 2013;44:2718-2721. Saposnik G, Reeves MJ, Johnston SC, et al. Predicting clinical outcomes after thrombolysis using the iScore: results from the Virtual International Stroke Trials Archive. Stroke 2013;44:2755-2759. Rocco A, Heuschmann PU, Schellinger PD, et al. Glycosylated hemoglobin a1 predicts risk for symptomatic hemorrhage after thrombolysis for acute stroke. Stroke 2013; 44:2134-2138.
Background: We aimed to test the outcome-predictive power of routine fasting glucose (FG) obtained at second day after onset in intravenous thrombolysis (IVT) acute ischemic stroke (AIS) patients. Methods: We identified AIS patients presenting to our institution between December 2011 and July 2013 within 4.5 hours of onset, who received admission glucose (AG) before IVT, FG, and glycated hemoglobin (HbA1c) the second day after admission, from our prospectively recorded stroke database. Multivariate logistic regression was used to assess the association of FG and 90-day modified Rankin Scale (mRS). Results: Between December 2011 and July 2013, a total of 166 AIS patients received intravenous plasminogen activator. Of those, 119 patients who have AG before IVT, FG, and HbA1c the second day were included in the study. FG independently predicted 90-day clinical unfavorable outcome (mRS, 3-6 with an odds ratio of 1.576; 95% confidence interval [CI], 1.0532.358; P 5 .027). This association was not significant in AG (P 5 .714), HbA1c (P 5.655), and history of diabetes (P 5.547). In receiver operating characteristic analysis, increased FG was associated with 90-day mRS (3-6) with an area under curve of .72, (95% CI, .65-.9; P 5.001). Conclusions: FG is a powerful predictor associated with the outcome in IVT-treated AIS patients independent of AG and HbA1c. Key Words: Admission glucose—fasting glucose—acute ischemic stroke—thrombolysis. Ó 2015 by National Stroke Association
Introduction Admission glucose (AG) level and history of diabetes mellitus (DM) are associated with poor clinical outcome after intravenous thrombolysis (IVT).1,2 However, a recent study found that IVT-treated patients with poststroke hyperglycemia (PSH) assessed with AG did not have significantly worse outcomes.3 Various physiologic From the Department of Neurology and Institute of Neurology, Huashan Hospital, Fudan University, Shanghai, China. Received June 22, 2014; revision received July 28, 2014; accepted July 30, 2014. W.C. and Y.L. contributed equally to this study. Address correspondence to Qiang Dong, MD, PhD, Department of Neurology, Huashan Hospital, No 12 Wulumuqi Zhong Rd, Shanghai 200040, China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.07.029
100
statuses, such as dietary intake or using insulin before onset, can impact AG level besides PSH and finally influence the association between PSH and outcome. Unfortunately, this issue was difficult to be controlled in clinical practice. The purpose of our study was to test the outcomepredictive strength of routine fasting glucose (FG) the second day morning after admission in IVT-treated stroke patients. We hypothesized that FG in a similar physiologic status could be more reliable to evaluate the pathologic glucose metabolism in acute ischemic stroke (AIS) patients than AG and therefore predict the outcome.
Methods Patients We identified AIS patients presenting to our institution between December 2011 and July 2013 within 4.5 hours of
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 1 (January), 2015: pp 100-103
FG PREDICTS STROKE OUTCOME
101
Table 1. Comparison of patient characteristics in those with good versus poor outcomes 90-day functional outcome
N Age, mean (SD), y Male, % Hypertension, % Diabetes, % Dyslipidemia, % Previous stroke, % Smoking, % OTT, mean (SD), min Baseline NIHSS, mean (SD) SBP, mean (SD), mm Hg DBP, mean (SD), mm Hg TOAST, % Large vessel disease Small vessel disease Cardioembolism Other and undetermined AG, mean (SD), mmol/L FG, mean (SD), mmol/L HbA1c, %
Total
mRS 0-2
mRS .2
119 66 (57-74) 65.5 (78/119) 65.5 (78/119) 31.9 (38/119) 19.3 (23/119) 14.3 (17/119) 42 (50/119) 180 (132-234) 10 (4-15) 150 (140-163) 87 (80-92)
72 63 (55-72) 69.4 (50/72) 66.7 (48/72) 23.6 (17/72) 15.3 (11/72) 9.7 (7/72) 45.8 (33/72) 184 (142-255) 6 (3-10) 147 (138-161) 86 (80-91)
47 71 (61-77) 59.6 (28/47) 63.8 (30/47) 44.7 (21/47) 25.5 (12/47) 21.3 (10/47) 36.2 (17/47) 155 (124-216) 15 (10-18) 157 (140-170) 87 (80-92)
34.7 (41/118) 27.1 (32/118) 34.7 (41/118) 3.4 (4/118) 7 (5.9-8.3) 6.1 (5.2-7.7) 6 (5.5-6.7)
38 (27/71) 35.2 (25/71) 22.5 (16/71) 4.2 (3/71) 6.4 (5.7-7.6) 5.8 (5.5-6.4) 5.8 (5.5-6.4)
29.8 (14/47) 14.9 (7/47) 53.2 (25/47) 2.1 (1/47) 7.8 (6.3-11.1) 7.6 (5.5-11.1) 6.3 (5.8-7.8)
P value
.027 .325 .844 .026 .235 .108 .345 .129 ,.001 .139 .696 .005
.001 ,.001 .002
Abbreviations: AG, admission glucose; DBP, diastolic blood pressure; FG, fasting glucose; HbA1c, glycated hemoglobin; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OTT, onset to treatment time; SD, standard deviation; SBP, systolic blood pressure; TOAST, Trial of Org 10172 in Acute Stroke Treatment.
onset, who received AG before intravenous plasminogen activator (IV-tPA), FG, and glycated hemoglobin (HbA1c) the second day after admission. We supplied the meal according to the swallow test if the patients were admitted before 8 pm on the first day. All patients were limited on bed in the intensive care unit during the first 24 hours after IV-tPA. Thus, the situation of dietary intake and bed rest level were similar among patients. Baseline National Institutes of Health Stroke Scale (NIHSS), 90-days mRS (modified Rankin Scale), blood pressure, and onset to treatment time were obtained from the database. Histories of hypertension, dyslipidemia, DM, previous stroke, and smoking were recorded at the admission. Stroke etiology was classified according to the Stop Stroke Study Trial of Org 10172 in Acute Stroke Treatment based on the following magnetic resonance imaging and computer tomography angiography after admission. The study was performed with the informed consent of the patients or their next of kin and with ethical approval from the Institutional Review Board of Huashan Hospital.
Statistics Analysis Statistical analyses were performed using SPSS, version 20 (SPSS Inc, Chicago, IL). P less than .05 was considered to indicate statistical significance. Patients were dichotomized by using the 90-day mRS score into good (mRS
score 0-2) versus poor outcome (mRS score 3-6). Differences of patients’ characteristics between outcomes were tested by the Fishers exact test for categorical and the Mann–Whiney test for continuous value. Multivariate binary logistic regression (including variables with P value ,.2) was used to assess the association of AG, HbA1c, and FG with 90-day poor outcome (mRS 3-6). Receiver operating characteristic analysis was performed between FG and poor outcome.
Results Between December 2011 and July 2013, a total of 166 AIS patients received IV-tPA. Of those, 119 patients have AG before IV-tPA, FG, and HbA1c the second day after admission. Patient characteristics (total and stratified by outcome) are listed in Table 1. As expected, increased age and higher baseline NIHSS, history of DM and atrial fibrillation, higher AG, FG, and HbA1c were significantly associated with worse functional outcome. Multivariate logistic regression analysis between variables (P value ,.2) and 90-day mRS was listed in Table 2. Previous stroke, onset to treatment time, baseline NIHSS, and FG was independently associated with outcome after adjust DM history, AG, HbA1c, systolic blood pressure, atrial fibrillation, and age. FG predicted favorable outcome with an odds ratio of 1.576 (95% confidence interval, 1.053-2.358; P 5.027). This association was
W. CAO ET AL.
102
not significant in AG (P 5.713), HbA1c (P 5.657), and history of DM (P 5 .549). In receiver operating characteristic analysis, increased FG was associated with 90-day mRS (3-6) with an area under curve of .72, (95% confidence interval, .65-.9; P 5 .001; Fig 1).
Discussion In this present study, increased FG was independently associated with poor outcome. However, no significant associations were found between history of DM, AG, or HbA1c and outcome after adjusting FG and vascular risks. Those results demonstrate FG is a powerful outcome predictor in AIS patients who received IVT. PSH is associated with poor clinical outcome and death.4 AG, commonly used to evaluate PSH, however, can be influenced by physiologic blood glucose frustration at admission. Data of dietary intake and insulin using before stroke onset were not controlled in prior PSH studies, which may affect the post-IVT clinical outcomes. Insignificant association between AG and clinical outcomes in IVT-treated patients has been reported in 1 study.3 Those opposite results need further data to reveal a correct association between hyperglycemia and outcome. Many prior AG-based PSH studies have notified the limitations of preonset data did not include the information of using insulin and dietary intake.3,5 On the other hand, AG over 8 mmol/L in predicting IVT outcome has been currently included in 2 population-based tools of DRAGON and iScore, which are widely used to predict the outcome in IVT patients.6,7 As a sort of random glucose test, AG over 8 mmol/L can also be observed in AIS patients without glucose metabolic disorder who onset within 2 hours after diet. Such phenomenon can definitely be a bias but difficult to be controlled in PSH studies. Highlight of our study was further setting FG and HbA1c to evaluate the long-term and short-term glucose metabolic status in IVT-treated patients. Previous study has confirmed HbA1c is an independent predictor for outcome, after adjusted AG.8 However, the strength of FG in predicting outcome was just analyzed in a highly selected non-DM AIS population.5 Notably, our study found increased FG was significantly associated with poor outcome independent of AG and HbA1c. FG obtained in a similar physiologic status is a more reliable predictor, advancing over AG and HbA1c, by balancing the various physiologic glucose metabolisms before onset. We therefore suggest FG the second day after admission should be included in IVT outcome prediction. Limitations of our study are because of the small sample size and derived from a single-center experience. The bed rest level and the situation of dietary intake at night of first day were not analyzed in detail, which could impact
Table 2. Multivariate logistic regression for 90-day poor clinical outcome (mRS 3-6) Predictors
OR
95% CI
P value
Age Diabetes Previous stroke OTT Baseline NIHSS SBP TOAST HbA1c AG FG
.985 .602 8.9 .991 1.228 .996 1.624 1.264 1.077 1.576
.941-1.031 .116-3.133 1.904-41.594 .983-.999 1.111-1.357 .968-1.025 .865-3.049 .452-3.532 .726-1.597 1.053-2.358
.511 .547 .005 .024 ,.001 .787 .131 .655 .714 .027
Abbreviations: AG, admission glucose; CI, confidence interval; FG, fasting glucose; HbA1c, glycosylated hemoglobin A1; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; OTT, onset to treatment time; SBP, systolic blood pressure; TOAST, Trial of Org 10172 in Acute Stroke Treatment.
the real FG level. The vessel recanalization in acute large artery occlusive patients and the size of the infarctions were also not assessed in present study.
Summary FG is a powerful predictor associated with the outcome in IVT-treated AIS patients independent of AG and HbA1c. Further studies are necessary to better define the relationship between FG and outcome in a randomized controlled trail.
Figure 1. Receiver operating characteristic (ROC) curve of fasting glucose in predicting 90-day poor outcome. Abbreviations: AUC, area under curve; CI, confidence interval.
FG PREDICTS STROKE OUTCOME
103
References 1. Ahmed N, Davalos A, Eriksson N, et al. Association of admission blood glucose and outcome in patients treated with intravenous thrombolysis: results from the Safe Implementation of Treatments in Stroke International Stroke Thrombolysis Register (SITS-ISTR). Arch Neurol 2010;67:1123-1130. 2. Bruno A, Levine SR, Frankel MR, et al. Admission glucose level and clinical outcomes in the NINDS rt-PA Stroke Trial. Neurology 2002;59:669-674. 3. Meurer WJ, Scott PA, Caveney AF, et al. Lack of association between hyperglycaemia at arrival and clinical outcomes in acute stroke patients treated with tissue plasminogen activator. Int J Stroke 2010;5:163-166. 4. Desilles JP, Meseguer E, Labreuche J, et al. Diabetes mellitus, admission glucose, and outcomes after stroke throm-
5.
6.
7.
8.
bolysis: a registry and systematic review. Stroke 2013; 44:1915-1923. Sykora M, Diedler J, Poli S, et al. Association of nondiabetic hyperglycemia with autonomic shift in acute ischaemic stroke. Eur J Neurol 2012;19:84-90. Strbian D, Seiffge DJ, Breuer L, et al. Validation of the DRAGON score in 12 stroke centers in anterior and posterior circulation. Stroke 2013;44:2718-2721. Saposnik G, Reeves MJ, Johnston SC, et al. Predicting clinical outcomes after thrombolysis using the iScore: results from the Virtual International Stroke Trials Archive. Stroke 2013;44:2755-2759. Rocco A, Heuschmann PU, Schellinger PD, et al. Glycosylated hemoglobin a1 predicts risk for symptomatic hemorrhage after thrombolysis for acute stroke. Stroke 2013; 44:2134-2138.
