Stress Hyperglycemia and Prognosis of Minor Ischemic Stroke and Transient Ischemic Attack The CHANCE Study (Clopidogrel in High-Risk Patients With Acute Nondisabling Cerebrovascular Events)

Yuesong Pan, MD*; Xueli Cai, MD*; Jing Jing, MD, PhD; , MD, PhD; Hao Li, PhD; Yongjun Wang, MD; Xingquan Zhao, MD, PhD; Liping Liu, MD, PhD; David Wang, DO; S. Claiborne Johnston, MD, PhD; Tiemin Wei, MD; Yilong Wang, MD, PhD; on behalf of the CHANCE Investigators

Background and Purpose—We aimed to determine the association between stress hyperglycemia and risk of new stroke in patients with a minor ischemic stroke or transient ischemic attack.

Downloaded from Methods—A subgroup of 3026 consecutive patients from 73 prespecified sites of the CHANCE trial (Clopidogrel in High- Risk Patients With Acute Nondisabling Cerebrovascular Events) were analyzed. Stress hyperglycemia was measured by glucose/glycated albumin (GA) ratio. Glucose/GA ratio was calculated by fasting plasma glucose divided by GA and categorized into 4 even groups according to the quartiles. The primary outcome was a new stroke (ischemic or hemorrhagic) at 90 days. We assessed the association between glucose/GA ratio and risk of stroke by multivariable Cox

http://stroke.ahajournals.org/ regression models adjusted for potential covariates. Results—Among 3026 patients included, a total of 299 (9.9%) new stroke occurred at 3 months. Compared with patients with the lowest quartile, patients with the highest quartile of glucose/GA ratio was associated with an increased risk of stroke at 3 months after adjusted for potential covariates (12.0% versus 9.2%; adjusted hazard ratio, 1.46; 95% confidence interval, 1.06–2.01). Similar results were observed after further adjusted for fasting plasma glucose. We also observed that higher level of glucose/GA ratio was associated with an increased risk of stroke with a threshold of 0.29 using a Cox regression model with restricted cubic spline. Conclusions—Stress hyperglycemia, measured by glucose/GA ratio, was associated with an increased risk of stroke in

by guest on October 23, 2017 patients with a minor ischemic stroke or transient ischemic attack. Clinical Trial Registration—URL: http://www.clinicaltrials.gov. Unique identifier: NCT00979589. (Stroke. 2017;48:3006-3011. DOI: 10.1161/STROKEAHA.117.019081.) Key Words: clopidogrel ◼ diabetes mellitus ◼ hyperglycemia ◼ risk ◼ stroke

tress hyperglycemia is the relative transient increase ischemic attack (TIA). Furthermore, most previous studies Sin glucose secondary to neurohormonal derangements defined stress hyperglycemia according to fasting or random and the inflammatory that occur during a critical illness like glucose without considering the background glucose level, stroke.1,2 Previous study showed that acute stress hypergly- although they were analyzed by diabetic status.3,4 Previous cemia predicts increased risks of in-hospital mortality and study showed that relative hyperglycemia, defined as admis- poor functional recovery after ischemic stroke.3 However, sion glucose divided by estimated average glucose derived few study investigated the association of stress hyperglycemia from glycosylated hemoglobin (HbA1c), was a better predic- with the risk of new stroke after ischemic stroke or transient tive biomarker of critical illness than absolute hyperglycemia.2

Received August 13, 2017; final revision received September 4, 2017; accepted September 14, 2017. From the Department of Neurology, Tiantan Hospital (Y.P., J.J., X.M., H.L., Yongjun Wang, X.Z., L.L., Yilong Wang) and Department of Epidemiology and Health Statistics, School of Public Health (Y.P.), Capital Medical University, Beijing, ; China National Clinical Research Center for Neurological Diseases, Beijing (Y.P., J.J., X.M., H.L., Yongjun Wang, X.Z., L.L., Yilong Wang); Center of Stroke, Beijing Institute for Brain Disorders, China (Y.P., J.J., X.M., H.L., Yongjun Wang, X.Z., L.L., Yilong Wang); Beijing Key Laboratory of Translational Medicine for Cerebrovascular Disease, China (Y.P., J.J., X.M., H.L., Yongjun Wang, X.Z., L.L., Yilong Wang); Beijing Municipal Key Laboratory of Clinical Epidemiology, China (Y.P.); Department of Neurology (X.C.) and Department of Cardiology (T.W.), Lishui Hospital of Zhejiang University (the Central Hospital of Lishui), China; INI Stroke Network, OSF Healthcare System, University of Illinois College of Medicine, Peoria (D.W.); and Dell Medical School, University of Texas at Austin (S.C.J.). *Drs Pan and Cai contributed equally. The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA. 117.019081/-/DC1. Correspondence to Yilong Wang, MD, PhD, No. 6 Tiantanxili, Dongcheng District, Beijing 100050, China. E-mail [email protected] © 2017 American Heart Association, Inc. Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.117.019081 3006 Pan et al Stress Hyperglycemia and Prognosis of Stroke 3007

Glycated albumin (GA) is a well-validated measure of the by quartiles of glucose/GA ratio (Q1–Q4) for further comparisons. The mean glucose concentration over ≈2 to 4 weeks.5,6 GA can novel index of glucose/GA ratio quantifies the extent of acute elevation reflect the background glucose level before the event when in plasma glucose compared with the background plasma glucose levels. tested within 2 days after the onset of stroke or TIA. In addi- tion, GA is not affected by chronic kidney disease and ane- Outcome Assessment mia and can be more accurately reflecting the actual status of The primary outcome in the current analyses was a new stroke (isch- glycemic control compared with HbA1c.7,8 Therefore, it can emic or hemorrhagic) during the 90-day follow-up period, which was same as that in the trial.10 Secondary outcomes included a new be postulated that the ratio of fasting plasma glucose (FPG) composite vascular event (ischemic stroke, hemorrhagic stroke, myo- to GA could reflect the stress hyperglycemia considering the cardial infarction, or vascular death) and ischemic stroke alone. All background glucose level before the onset of event. reported events were verified by a central adjudication committee that In this study, we investigated the association of stress was blinded to the study group assignments. hyperglycemia, measured by glucose/GA ratio, and outcomes of patients with a minor ischemic stroke or TIA, using the data Statistical Analysis derived from the CHANCE trial (Clopidogrel in High-Risk Continuous variables were presented as median with interquartile and Patients With Acute Nondisabling Cerebrovascular Events). categorical variables as percentages. The baseline characteristics of the patients included in and excluded from the analysis were com- pared using Wilcoxon rank-sum test for continuous variables and χ2 Methods test for categorical variables. The baseline variables among different Downloaded from Study Participants quartiles of glucose/GA ratio were compared by Kruskal–Wallis test for continuous variables and χ2 test for categorical variables. The CHANCE trial is a randomized, double-blind, controlled trial con- We tested the statistical significance of group of glucose/GA ducted at 114 hospitals in China between October 2009 and July 2012. ratio×study group assignment in a multivariable Cox model to exam- Details on the rational, design, and major results of the CHANCE trial 9–11 ine the interaction effect of category of stress hyperglycemia by study have been published previously. In brief, 5170 patients with a minor group assignment. Adjusted hazard ratios and their 95% confidence ischemic stroke (National Institutes of Health Stroke Scale score of http://stroke.ahajournals.org/ intervals were calculated using multivariable Cox regression mod- 3) or high-risk TIA (ABCD2 score [age, blood pressure, clinical fea- ≤ els to assess the associations between quartiles of glucose/GA ratio tures, duration of TIA, and diabetes] of ≥4) within 24 hours after onset and prognosis of minor stroke or TIA. In the multivariable regression of the events were randomized to the group of clopidogrel plus aspirin analyses, 2 models were performed. In the first model, we adjusted (day 1: clopidogrel 300 mg and open-label aspirin 75–300 mg; days for all the potential covariates including history of diabetes melli- 2–21: clopidogrel 75 mg and aspirin 75 mg/d; and days 22–90: clopi- tus. In the second model, we further adjusted for FPG to assess the dogrel 75 mg/d) or aspirin alone (day 1: open-label aspirin 75–300 mg; and days 2–90: aspirin 75 mg/d) in the CHANCE trial. The pro- tocol of the CHANCE trial was approved by ethics committee at each Table 1. Baseline Characteristics of the Patients Included study center. Written informed consent was obtained from all partici- and Excluded

by guest on October 23, 2017 pants or their representatives before being enrolled into the study. Data are available to researchers on request for purposes of reproducing Included Excluded P the results or replicating the procedure by directly contacting the cor- Characteristic (n=3026) (n=2144) Value responding author ([email protected]). Age (y), median (IQR) 62.2 (54.7–71.2) 62.3 (54.6–71.4) 0.77 Among 114 clinical sites included in the CHANCE trial, 73 (64%) sites with 3044 patients had prior experience collecting samples and Female, n (%) 1015 (33.5) 735 (34.3) 0.58 agreed to participate in the prespecified biomarker substudy. All patients recruited by these 73 sites participated this substudy, and a Medical history, n (%) separate written consent form was obtained from the patient. Ischemic stroke 578 (19.1) 455 (21.2) 0.06 TIA 94 (3.1) 80 (3.7) 0.22 Baseline Data Collection Myocardial infarction 54 (1.8) 42 (2.0) 0.65 Baseline data on demographics and cardiovascular risk factors were col- lected through face-to-face interviews by trained research coordinators Angina 94 (3.1) 90 (4.2) 0.04 from participating hospitals. Cardiovascular risk factors included his- tory of ischemic stroke, TIA, myocardial infarction, angina, congestive Congestive heart failure 54 (1.8) 26 (1.2) 0.10 heart failure, atrial fibrillation or flutter, valvular heart disease, hyperten- Known atrial fibrillation 57 (1.9) 39 (1.8) 0.87 sion, diabetes mellitus, hypercholesterolemia, and smoking status. or flutter Valvular heart disease 10 (0.3) 4 (0.2) 0.33 Assessment of Stress Hyperglycemia Venous blood was drawn from patients after overnight fasting within Hypertension 1972 (65.2) 1427 (66.6) 0.30 the first 2 days after randomization. FPG was then measured. The Diabetes mellitus 611 (20.2) 482 (22.5) 0.047 serum specimens were collected and shipped on ice by overnight cou- rier from each site to Beijing Tiantan Hospital, where all serum speci- Hypercholesterolemia 317 (10.5) 256 (11.9) 0.10 mens were stored at −80°C. Serum GA levels were assayed with a kit Previous or current smoking, 1295 (42.8) 926 (43.2) 0.78 (catalog number 4085-717; Ruiyuan Bio-Technique Co Ltd, Ningbo, n (%) China) using a Roche Modular P800 system in the clinical laboratory of Beijing Tiantan Hospital.12 The GA level was expressed as a per- Index event, n (%) 0.03 centage of total serum albumin. All measurements were performed Minor stroke 2215 (73.2) 1510 (70.4) by laboratory personnel blinded to the study samples, study group assignments, and outcomes. TIA 811 (26.8) 634 (29.6) Stress hyperglycemia was measured by glucose/GA ratio, which was calculated using the following formula: glucose/GA ratio=FPG New stroke at 3 mo 299 (9.9) 216 (10.1) 0.82 (mmol/L)/GA (%). The patients were then categorized into 4 groups IQR indicates interquartile range; and TIA, transient ischemic attack. 3008 Stroke November 2017

independent association of stress hyperglycemia with the outcome among which 3026 (99.4%) patients had tested both FPG and controlled for absolute hyperglycemia. GA. Table 1 shows the baseline characteristics of the patients We further evaluated the pattern and magnitude of associations included in this analysis. The patient included in and those between glucose/GA ratio and risk of stroke using a Cox regression model with restricted cubic splines for glucose/GA ratio adjusting for excluded from this analysis were well balanced except that covariates. The 25th percentile of the glucose/GA ratio (0.29) was those included had a lower proportion of history of angina treated as the reference, and the fifth, 25th, 50th, 75th, and 95th per- and diabetes mellitus, and TIA as the index event. Among centiles of glucose/GA ratio were treated as the 5 knots for spline. the 3026 patients included, the median age was 62.2 (range, All analyses were conducted with SAS 9.4 (SAS Institute Inc, 37–94), and 1015 (33.5%) patients were female. Cary, NC), and a 2-sided P value <0.05 was considered to be statisti- cally significant. The median glucose/GA ratio was 0.33 (interquartile, 0.29–0.39). Table 2 shows the baseline characteristics of the Results patients by quartile of glucose/GA ratio. Patients with higher glucose/GA ratio were younger, less likely to be female and Study Participants and Characteristics have a history of known atrial fibrillation or flutter, valvular A total of 3044 consecutive patients with a minor ischemic heart disease, and diabetes mellitus, but more likely to be stroke or TIA were enrolled in the 73 sites for this substudy, current smoker and have a history of hypercholesterolemia.

Table 2. Baseline Characteristics of the Patients by Quartile of Glucose/GA Ratio Downloaded from Glucose/GA Ratio Characteristic Q1 (n=763)* Q2 (n=755) Q3 (n=756) Q4 (n=752) P Value Age (y), median (IQR) 67.0 (59.3–74.1) 63.5 (56.6–72.2) 60.3 (53.5–68.8) 58.0 (51.5–66.5) <0.001 Female, n (%) 266 (34.9) 266 (35.2) 264 (34.9) 219 (29.1) 0.03 http://stroke.ahajournals.org/ Medical history, n (%) Ischemic stroke 159 (20.8) 155 (20.5) 137 (18.1) 127 (16.9) 0.15 TIA 22 (2.9) 30 (4.0) 19 (2.5) 23 (3.1) 0.41 Myocardial infarction 19 (2.5) 15 (2.0) 10 (1.3) 10 (1.3) 0.25 Angina 32 (4.2) 21 (2.8) 15 (2.0) 26 (3.5) 0.08 Congestive heart failure 12 (1.6) 15 (2.0) 14 (1.9) 13 (1.7) 0.94

by guest on October 23, 2017 Known atrial fibrillation or flutter 24 (3.1) 15 (2.0) 8 (1.1) 10 (1.3) 0.01 Valvular heart disease 6 (0.8) 3 (0.4) 0 (0.0) 1 (0.1) 0.04 Hypertension 487 (63.8) 484 (64.1) 490 (64.8) 511 (68.0) 0.31 Diabetes mellitus 207 (27.1) 138 (18.3) 117 (15.5) 149 (19.8) <0.001 Hypercholesterolemia 67 (8.8) 89 (11.8) 65 (8.6) 96 (12.8) 0.01 Smoking status, n (%) <0.001 Never smoking 469 (61.5) 450 (59.6) 421 (55.7) 391 (52.0) Previous smoker 83 (10.9) 78 (10.3) 70 (9.3) 68 (9.0) Current smoker 211 (27.7) 227 (30.1) 265 (35.1) 293 (39.0) Index event, n (%) 0.43 Minor stroke 563 (73.8) 560 (74.2) 536 (70.9) 556 (73.9) TIA 200 (26.2) 195 (25.8) 220 (29.1) 196 (26.1) NIHSS score on admission, median (IQR) 2 (0–2) 2 (0–2) 1 (0–2) 2 (0–2) 0.08 Time to randomization (h), median (IQR) 12.0 (7.0–20.0) 12.0 (6.5–19.0) 11.5 (6.2–19.3) 12.0 (6.2–19.0) 0.15 Antiplatelet therapy, n (%) 0.30 Aspirin only 374 (49.0) 367 (48.6) 401 (53.0) 378 (50.3) Clopidogrel+aspirin 389 (51.0) 388 (51.4) 355 (47.0) 374 (49.7) FPG (mmol/L), median (IQR) 4.9 (4.5–5.7) 5.2 (4.8–5.8) 5.6 (5.1–6.4) 6.2 (5.6–7.6) <0.001 GA (%), median (IQR) 18.9 (17.2–23.4) 16.7 (15.4–18.7) 15.5 (14.2–18.1) 14.2 (12.8–16.9) <0.001 Glucose/GA ratio, median (IQR) 0.26 (0.24–0.28) 0.31 (0.30–0.32) 0.36 (0.34–0.37) 0.43 (0.41–0.47) FPG indicates fasting plasma glucose; GA, glycated albumin; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; and TIA, transient ischemic attack. *Quartiles of glucose/GA ratio, Q1<0.290, Q2=0.290–0.333, Q3=0.334–0.385, Q4>0.385. Pan et al Stress Hyperglycemia and Prognosis of Stroke 3009

Furthermore, patients with higher glucose/GA ratio had higher interaction=0.52; highest versus lowest quartiles of glucose/ FPG (median FPG, 4.9, 5.2, 5.6, and 6.2 mmol/L in patients GA ratio: adjusted hazard ratio, 1.70; 95% confidence inter- with the first, second, third, and fourth quartiles of glucose/ val, 0.96–3.00; P=0.07 for patients with history of diabetes GA ratio, respectively) but lower GA (median GA, 18.9%, mellitus, and adjusted hazard ratio, 1.40; 95% confidence 16.7%, 15.5%, and 14.2% in patients with the first, second, interval, 0.94–2.07; P=0.10 for patients without history of dia- third, and fourth quartiles of glucose/GA ratio, respectively). betes mellitus). Similar results were found for the end points of composite vascular events and ischemic stroke. No signifi- Association of Stress Hyperglycemia With Outcome cant association was observed between quartiles of glucose/ Table 3 shows the 3-month prognosis after a minor stroke or GA ratio and risk of bleeding. In the sensitivity analyses fur- TIA across quartiles of glucose/GA ratio. There were 299 ther adjusted for FPG, similar results were observed for each (9.9%) new stroke occurrence at 3 months, among which end point (Table 3, model 2). 293 (98.0%) were ischemic stroke and 6 (2.0%) were hemor- Using a Cox regression model with restricted cubic spline, rhagic stroke. No interaction effect of study group assignment we found that higher level of glucose/GA ratio was associ- by categories of glucose/GA ratio was observed for the risk ated with an increased risk of stroke with a threshold of 0.29 of stroke (P for interaction=0.41). Compared with patients (Figure). with the lowest quartile, patients with the highest quartile of glucose/GA ratio had ≈1.5-fold of risk of stroke at 3 months Discussion Downloaded from after fully adjusted for potential covariates (12.0% versus In this study, we found that the stress hyperglycemia on admis- 9.2%; adjusted hazard ratio, 1.46; 95% confidence interval, sion was associated with an increased risk of new stroke in 1.06–2.01). No significant difference was observed between patients with a minor ischemic stroke or TIA. This association patients with and without history of diabetes mellitus (P for was independent of absolute hyperglycemia on admission. http://stroke.ahajournals.org/ Table 3. Risk of Stroke at 3 Months After a Minor Stroke or Transient Ischemic Attack by Stress Hyperglycemia Status

Model 1* Model 2† Outcomes Glucose/GA Ratio n Events, n (%) Adjusted HR (95% CI) P Value Adjusted HR (95% CI) P Value Stroke Q1 (<0.290) 763 70 (9.2) 1 1 Q2 (0.290–0.333) 755 66 (8.7) 1.00 (0.71–1.41) 0.98 1.01 (0.72–1.42) 0.96

by guest on October 23, 2017 Q3 (0.334–0.385) 756 73 (9.7) 1.16 (0.83–1.62) 0.39 1.17 (0.83–1.65) 0.37 Q4 (>0.385) 752 90 (12.0) 1.46 (1.06–2.01) 0.02 1.48 (1.05–2.09) 0.03 P for trend 0.01 0.02 Composite Q1 (<0.290) 763 70 (9.2) 1 1 events‡ Q2 (0.290–0.333) 755 66 (8.7) 1.00 (0.72–1.41) 0.98 1.01 (0.72–1.42) 0.95 Q3 (0.334–0.385) 756 74 (9.8) 1.18 (0.84–1.64) 0.34 1.19 (0.84–1.68) 0.32 Q4 (>0.385) 752 91 (12.1) 1.47 (1.07–2.03) 0.02 1.50 (1.07–2.12) 0.02 P for trend 0.01 0.01 Ischemic stroke Q1 (<0.290) 763 69 (9.0) 1 1 Q2 (0.290–0.333) 755 65 (8.6) 1.00 (0.71–1.40) 0.99 1.00 (0.71–1.41) 0.99 Q3 (0.334–0.385) 756 70 (9.3) 1.11 (0.79–1.57) 0.53 1.12 (0.79–1.58) 0.54 Q4 (>0.385) 752 89 (11.8) 1.44 (1.04–2.00) 0.03 1.45 (1.02–2.04) 0.04 p for trend 0.02 0.03 Bleeding Q1 (<0.290) 763 20 (2.6) 1 1 Q2 (0.290–0.333) 755 19 (2.5) 1.02 (0.54–1.92) 0.95 1.09 (0.57–2.07) 0.80 Q3 (0.334–0.385) 756 10 (1.3) 0.53 (0.24–1.19) 0.12 0.60 (0.26–1.36) 0.22 Q4 (>0.385) 752 10 (1.3) 0.68 (0.31–1.47) 0.32 0.81 (0.35–1.88) 0.62 p for trend 0.14 0.35 CI indicates confidence Interval; GA, glycated albumin; and HR, hazard ratio. *Model 1: adjusted for age, sex, history of ischemic stroke, transient ischemic attack, myocardial infarction, angina, congestive heart failure, known atrial fibrillation or flutter, valvular heart disease, hypertension, diabetes mellitus, hypercholesterolemia, smoking status, time to randomization, index event, and National Institutes of Health Stroke Scale on admission, antiplatelet therapy. †Model 2: adjusted for model 1+fasting plasma glucose. ‡Composite events: stroke, myocardial infarction, or death from cardiovascular causes. 3010 Stroke November 2017

events.15,16 Third, hyperglycemia may also induce platelet aggregation. Previous study showed that hyperglycemia has a deleterious effect on platelet function.17 These pathological changes tend to promote atherosclerosis and subsequent stroke. Stress hyperglycemia was usually diagnosed according to absolute hyperglycemia without previous diabetes mellitus or preexisting diabetes mellitus with deterioration of premorbid glycemic control.1,3 However, this definition failed to distin- guish stress hyperglycemia from newly diagnosed diabetes mellitus or unknown diabetes mellitus and did not consider the background glucose level. In contrast, relative measures of hyperglycemia, such as glucose/GA ratio, represent a quanti- tative measurement of the relative acute rises in plasma glu- cose concentration compared with premorbid glucose status. Previous study showed that relative measures controlling for Figure. Adjusted hazard ratios of stroke according to glucose/ background glycemia were a better biomarker of critical ill- 2 glycated albumin (GA) ratio. The solid line indicates the estimated ness than absolute hyperglycemia. Furthermore, GA level

Downloaded from hazard ratio and the dashed lines the 95% confidence intervals. reflects mean glycemia over≈ 2 to 4 weeks. Measures of stress The vertical dashed lines indicate the first, second, and third hyperglycemia based on GA are not affected by chronic kid- quartiles of the glucose/GA ratio. Reference is the 25th percentile 7,8 of the glucose/GA ratio (0.29). Data were fitted using a multivari- ney disease, anemia, or hemoglobinopathies, unlike those able Cox regression model of restricted cubic spline with 5 knots measures based on HbA1c, such as the stress hyperglycemia (the fifth, 25th, 50th, 75th, and 95th percentiles) for glucose/GA ratio (defined as admission glucose divided by estimated aver- ratio. The lowest 5% and highest 5% of participants were not 2 http://stroke.ahajournals.org/ age glucose derived from HbA1c) and plasma glucose con- shown in the figures for small sample sizes. centration/HbA1c ratio18 used in the previous studies. It was Stress hyperglycemia is common in patients with acute reported that >40% of stroke patients had reduced estimated 19 severe diseases, such as myocardial infarction and stroke. glomerular filtration rate. In addition, a separate specimen It was reported that the prevalence of stress hyperglycemia of whole blood is required to measure HbA1c, but GA can ranged from 8% to 63% in nondiabetic patients with acute be measured from serum or plasma with the blood glucose 12 stroke.3 Differences in study population and definition of stress level. On the other hand, measurement of GA was affected hyperglycemia may contribute to the heterogeneity of these by alteration in serum albumin metabolism. In addition, there

by guest on October 23, 2017 is no consistent agreement of cutoff levels of GA suggesting estimates. Previous studies showed that acute stress hyper- that diabetes mellitus is present.7 Results from the Kyushu and glycemia was associated with an increased risk of in-hospital Okinawa Population Study showed that FPG concentrations mortality in diabetic and nondiabetic patients with myocar- of 5.56, 6.11, and 7.00 mmol/L corresponded to HbA1c levels dial infarction13 and in-hospital mortality and poor functional of 5.6%, 5.9%,and 6.3% and GA levels of 15.0%, 15.7%, and recovery in nondiabetic patients with ischemic stroke.3 Recent 16.9%, respectively.5 Nevertheless, our study has suggested a study also showed that hyperglycemia was associated with practical and economical method to precisely detect and quan- 3-month mortality in nondiabetic and good previous glucose tify stress hyperglycemia, which provides prognostic informa- controlled diabetic patients with acute ischemic stroke, but tion in patients with acute minor ischemic stroke or TIA. not in those with poor previous glucose controlled diabetics.14 This study has several limitations. First, we could not esti- However, another study from Greece did not demonstrate sig- mate the association between stress hyperglycemia and stroke nificant direct association between stress hyperglycemia and prognosis by stroke cause because the baseline data of cause outcome of acute ischemic stroke.4 Previous studies mostly classification for enrolled patients were not collected in the focused on the association of stress hyperglycemia and risk of CHANCE trial. Second, only Chinese patients were enrolled 3,14 mortality and poor functional outcome after stroke ; how- in the trial. Chinese patients with ischemic stroke or TIA had a ever, few study investigated the relationship between stress higher prevalence of large artery atherosclerosis.20,21 This lim- hyperglycemia and the risk of stroke recurrence. Our study its the generalizability of the findings to a Western population further added the evidence that stress hyperglycemia was with a different disease pattern of ischemic stroke. associated with an increased risk of new stroke in patients with acute ischemic stroke or TIA. Conclusions Although the mechanism is not fully understood, several Our results demonstrated that stress hyperglycemia, measured explanations may account for the observed association between by glucose/GA ratio, was associated with an increased risk of stress hyperglycemia and an increased risk of new stroke after new stroke in patients with a minor ischemic stroke or TIA. minor stroke or TIA. First, higher relative rises of glycemia represent a greater inflammatory and neurohormonal response to the acute events and release of inflammatory and vasocon- Sources of Funding 2 This study is supported by grants from the Ministry of Science and strictive factors. Second, glucose fluctuations may aggra- Technology of the People’s Republic of China (2015BAI12B04, vate endothelial dysfunction and oxidative stress, which are 2015BAI12B02, 2016YFC0901000, 2016YFC0901001, 2 key players in favoring cardiovascular and cerebrovascular 2016YFC0901002), grants from Beijing Municipal Science and Pan et al Stress Hyperglycemia and Prognosis of Stroke 3011

Technology Commission (Z15110200390000, Z151100003915117), 10. Wang Y, Wang Y, Zhao X, Liu L, Wang D, Wang C, et al; CHANCE and grants from Beijing Municipal Commission of Health and Family Investigators. Clopidogrel with aspirin in acute minor stroke or tran- Planning (No. 2016-1-2041, SML20150502). sient ischemic attack. N Engl J Med. 2013;369:11–19. doi: 10.1056/ NEJMoa1215340. 11. Wang Y, Pan Y, Zhao X, Li H, Wang D, Johnston SC, et al; CHANCE Disclosures Investigators. Clopidogrel with aspirin in acute minor stroke or tran- Dr Johnston is the principal investigator of the POINT trial (The sient ischemic attack (CHANCE) trial: one-year outcomes. Circulation. Platelet-Oriented Inhibition in New TIA and Minor Ischemic Stroke), 2015;132:40–46. doi: 10.1161/CIRCULATIONAHA.114.014791. a National Institutes of Health-sponsored trial with clopidogrel and 12. Li J, Wang Y, Wang D, Lin J, Wang A, Zhao X, et al; CHANCE placebo donated by Sanofi. Dr Johnston also has a research grant Investigators. Glycated albumin predicts the effect of dual and single donated by AstraZeneca and have a consultant or advisory relationship antiplatelet therapy on recurrent stroke. Neurology. 2015;84:1330–1336. with AstraZeneca and Biogen. The other authors report no conflicts. doi: 10.1212/WNL.0000000000001421. 13. Capes SE, Hunt D, Malmberg K, Gerstein HC. Stress hyperglycaemia and increased risk of death after myocardial infarction in patients with References and without diabetes: a systematic overview. Lancet. 2000;355:773–778. 1. Dungan KM, Braithwaite SS, Preiser JC. Stress hyperglycaemia. Lancet. doi: 10.1016/S0140-6736(99)08415-9. 2009;373:1798–1807. doi: 10.1016/S0140-6736(09)60553-5. 14. Roquer J, Giralt-Steinhauer E, Cerdà G, Rodríguez-Campello A, 2. Roberts GW, Quinn SJ, Valentine N, Alhawassi T, O’Dea H, Stranks SN, Cuadrado-Godia E, Jiménez-Conde J, et al. Glycated hemoglobin value et al. Relative hyperglycemia, a marker of critical illness: introducing the combined with initial glucose levels for evaluating mortality risk in stress hyperglycemia ratio. J Clin Endocrinol Metab. 2015;100:4490– patients with ischemic stroke. Cerebrovasc Dis. 2015;40:244–250. doi: 4497. doi: 10.1210/jc.2015-2660. 10.1159/000440735. 3. Capes SE, Hunt D, Malmberg K, Pathak P, Gerstein HC. Stress hyper- 15. Ceriello A, Esposito K, Piconi L, Ihnat MA, Thorpe JE, Testa R, et al. Downloaded from glycemia and prognosis of stroke in nondiabetic and diabetic patients: a Oscillating glucose is more deleterious to endothelial function and oxi- systematic overview. Stroke. 2001;32:2426–2432. dative stress than mean glucose in normal and type 2 diabetic patients. 4. Tziomalos K, Dimitriou P, Bouziana SD, Spanou M, Kostaki S, Diabetes. 2008;57:1349–1354. doi: 10.2337/db08-0063. Angelopoulou SM, et al. Stress hyperglycemia and acute ischemic stroke 16. Monnier L, Mas E, Ginet C, Michel F, Villon L, Cristol JP, et al. in-hospital outcome. Metabolism. 2017;67:99–105. doi: 10.1016/j. Activation of oxidative stress by acute glucose fluctuations compared metabol.2016.11.011. with sustained chronic hyperglycemia in patients with type 2 diabetes.

http://stroke.ahajournals.org/ 5. Furusyo N, Koga T, Ai M, Otokozawa S, Kohzuma T, Ikezaki H, et JAMA. 2006;295:1681–1687. doi: 10.1001/jama.295.14.1681. al. Utility of glycated albumin for the diagnosis of diabetes mellitus 17. Worthley MI, Holmes AS, Willoughby SR, Kucia AM, Heresztyn T, in a Japanese population study: results from the Kyushu and Okinawa Stewart S, et al. The deleterious effects of hyperglycemia on platelet Population Study (KOPS). Diabetologia. 2011;54:3028–3036. doi: function in diabetic patients with acute coronary syndromes mediation by 10.1007/s00125-011-2310-6. superoxide production, resolution with intensive insulin administration. 6. Selvin E, Rawlings AM, Lutsey PL, Maruthur N, Pankow JS, Steffes M, J Am Coll Cardiol. 2007;49:304–310. doi: 10.1016/j.jacc.2006.08.053. et al. Fructosamine and glycated albumin and the risk of cardiovascu- 18. Su YW, Hsu CY, Guo YW, Chen HS. Usefulness of the plasma glu- lar outcomes and death. Circulation. 2015;132:269–277. doi: 10.1161/ cose concentration-to-HbA1c ratio in predicting clinical outcomes CIRCULATIONAHA.115.015415. during acute illness with extreme hyperglycaemia. Diabetes Metab. 7. Yazdanpanah S, Rabiee M, Tahriri M, Abdolrahim M, Rajab A, Jazayeri 2017;43:40–47. doi: 10.1016/j.diabet.2016.07.036. HE, et al. Evaluation of glycated albumin (GA) and GA/HbA1c ratio for 19. Luo Y, Wang X, Matsushita K, Wang C, Zhao X, Hu B, et al; CNSR by guest on October 23, 2017 diagnosis of diabetes and glycemic control: a comprehensive review. Crit Investigators. Associations between estimated glomerular filtration rate Rev Clin Lab Sci. 2017;54:1–14. and stroke outcomes in diabetic versus nondiabetic patients. Stroke. 8. Selvin E, Rawlings AM, Grams M, Klein R, Sharrett AR, Steffes M, et 2014;45:2887–2893. doi: 10.1161/STROKEAHA.114.005380. al. Fructosamine and glycated albumin for risk stratification and predic- 20. Suri MF, Qiao Y, Ma X, Guallar E, Zhou J, Zhang Y, et al. Prevalence tion of incident diabetes and microvascular complications: a prospec- of intracranial atherosclerotic stenosis using high-resolution magnetic tive cohort analysis of the Atherosclerosis Risk in Communities (ARIC) resonance angiography in the general population: the Atherosclerosis study. Lancet Diabetes Endocrinol. 2014;2:279–288. doi: 10.1016/ Risk in Communities Study. Stroke. 2016;47:1187–1193. doi: 10.1161/ S2213-8587(13)70199-2. STROKEAHA.115.011292. 9. Wang Y, Johnston SC; CHANCE Investigators. Rationale and design of a 21. Wang Y, Zhao X, Liu L, Soo YO, Pu Y, Pan Y, et al; CICAS Study randomized, double-blind trial comparing the effects of a 3-month clop- Group. Prevalence and outcomes of symptomatic intracranial large idogrel-aspirin regimen versus aspirin alone for the treatment of high- artery stenoses and occlusions in China: the Chinese Intracranial risk patients with acute nondisabling cerebrovascular event. Am Heart J. Atherosclerosis (CICAS) Study. Stroke. 2014;45:663–669. doi: 10.1161/ 2010;160:380.e1–386.e1. doi: 10.1016/j.ahj.2010.05.017. STROKEAHA.113.003508. Stress Hyperglycemia and Prognosis of Minor Ischemic Stroke and Transient Ischemic Attack: The CHANCE Study (Clopidogrel in High-Risk Patients With Acute Nondisabling Cerebrovascular Events) Yuesong Pan, Xueli Cai, Jing Jing, Xia Meng, Hao Li, Yongjun Wang, Xingquan Zhao, Liping

Downloaded from Liu, David Wang, S. Claiborne Johnston, Tiemin Wei and Yilong Wang on behalf of the CHANCE Investigators

Stroke. 2017;48:3006-3011; originally published online October 19, 2017; http://stroke.ahajournals.org/ doi: 10.1161/STROKEAHA.117.019081 Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2017 American Heart Association, Inc. All rights reserved. Print ISSN: 0039-2499. Online ISSN: 1524-4628

by guest on October 23, 2017 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://stroke.ahajournals.org/content/48/11/3006

Data Supplement (unedited) at: http://stroke.ahajournals.org/content/suppl/2017/10/18/STROKEAHA.117.019081.DC1

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at: http://www.lww.com/reprints

Subscriptions: Information about subscribing to Stroke is online at: http://stroke.ahajournals.org//subscriptions/

SUPPLEMENTAL MATERIALS

The CHANCE investigators: Yongjun Wang (BeijingTiantan Hospital, Principal Investigator ); S.Claiborne Johnston (Dell Medical School, University of Texas at Austin, Austin, USA, Co-Principal Investigator ); Yilong Wang (BeijingTiantan Hospital, Executive Committee); Xingquan Zhao (BeijingTiantan Hospital, Site Investigator ); Zhimin Wang ( Taizhou First People’s Hospital, Site Investigator); Haiqin Xia ( Taiyuan Iron And Steel (Group) Co., Ltd., General Hospital, Site Investigator) ; Bin Li ( Dagang Oilfield Gengeal Hospital, Site Investigator) ; Guiru Zhang ( Penglai People’s Hospital, Site Investigator) ; Xudong Ren ( The Third People’s Hospital Of Datong, Site Investigator); Chunling Ji ( The Fourth Central Hospital Of Tianjin, Site Investigator); Guohua Zhang ( The Second Hospital Of Hebei Medical University, Site Investigator); Jianhua Li ( The First Hospital Of Fangshan District, Beijing, Site Investigator); Bohua Lu ( Beijing Puren Hospital, Site Investigator); Liping Wang ( Tianjin Ninghe District Hospital, Site Investigator); Shutao Feng ( The People’s Hospital Of Zhengzhou, Site Investigator); Dali Wang ( Affiliated Hospital Of North China Coal Medical College, Site Investigator); WeiguoTang ( Zhejiang Zhoushan Hospital, Site Investigator); Juntao Li ( Han Dan Central Hospital, Site Investigator); Hongtian Zhang ( Zhecheng People’s Hospital, Site Investigator); Guanglai Li (Shanxi Medical University Second Hospital, Site Investigator); Baojun Wang ( Baotou Central Hospital, Site Investigator); Yuhua Chen (The General Hospital Of Changjiang River Shipping, Site Investigator); Ying Lian ( Dalian Economic And Technological Development Zone Hospital, Site Investigator); Bin Liu ( First Neurology Department, Affiliated Hospital Of North China Coal Medical College, Site Investigator); Junfang Teng ( The First Affiliated Hospital Of Zhengzhou University, Site Investigator); Rubo Sui ( First Affiliated Hospital Of Liaoning Medical, Site Investigator); Lejun Li ( Lianyungang Municipal Hospital Of TCM, Site Investigator); Zhiling Yuan ( Central Hospital In Qiu County, Site Investigator); Dawei Zang (Tianjin First Center Hospital, Site Investigator); Zuneng Lu ( Renmin Hospital Of Wuhan University, Site Investigator); Li Sun ( Qingdao Central Hospital, Site Investigator); Dong Wang ( Baogang Hospital, Site Investigator); Liying Hou ( Changzhi City People’s Hospital Of Shanxi Province, Site Investigator); Dongcai Yuan ( HaLixun International Peace Hospital, Site Investigator); Yongliang Cao ( People’s Hospital Of Linzi District, Zibo, Site Investigator); Hui Li (Yantai City Yantai Mountain Hospital, Site Investigator); Xiuge Tan ( Beijing Pinggu District Hospital, Site Investigator); Huicong Wang ( Taiyuan Central Hospital, Site Investigator); Haisong Du ( Chengde Central Hospital, Site Investigator); Mingyi Liu (Shijiazhuang Central Hospital, Site Investigator); Suping Wang ( First Neurology Department, Dalian Municipal Central Hospital, Site Investigator); Qiuwu Liu (Xian 141 Hospital, Site Investigator); Zhong Zhang ( Chengdu Third Municipal People’s Hospital, Site Investigator); Qifu Cui ( Affiliated Hospital Of Chifeng University, Site Investigator); Runqing Wang ( Zhengzhou Central Hospital, Site Investigator); Jialin Zhao ( Ningbo City, Zhejiang Province Lihuili Hospital Medical Center, Site Investigator); Jiewen Zhang ( Henan Provincial People’s Hospital, Site Investigator); Jianping Zhao ( Jinzhong City Second Hospital, Site Investigator); Qi Bi ( Beijing Anzhen Hospital, Capital Medical University, Site Investigator); Xiyou Qi ( Beijing Huairou District Chinese Medicine Hospital, Site Investigator); Junyan Liu ( Hebei Medical University Third Hospital, Site Investigator); Changxin Li ( First Affiliated Hospital Shanxi Medical Unversity, Site Investigator); Ling Li ( Hebei Provincial People’s Hospital, Site Investigator); Xiaoping Pan ( First Municipal Peoples Hospital, Site Investigator); Junling Zhang ( Central Hospital In Cangzhou, Site Investigator); Derang Jiao ( The Chinese People’s Armed Police Force Medical School Affiliated Hospital, Site Investigator); Zhao Han ( Zhejiang Wenzhou Medical College First Affiliated Hospital, Site Investigator); Dawei Qian ( Jilin Central Hospital, Site Investigator); Jin Xiao ( Anhui Maanshan Central Hospital, Site Investigator); Yan Xing ( Beijing Aviation Industry Central Hospital, Site Investigator); Huishan Du ( Luhe Hospital, Tongzhou District, Beijing, Site Investigator); Guang Huang ( Beijing Fuxing Hospital, Capital Medical University, Site Investigator); Yongqiang Cui ( The 306th Hospital Of P.L.A, Site Investigator); Yan Li ( The First Affiliated Hospital Of Tianjin University Of Chinese Medicine, Site Investigator); Lianyuan Feng ( Baiqiuen International Peace Hospital Of People’s Liberation Army, Site Investigator); Lianbo Gao ( Fourth Affiliated Hospital Of China Medical University, Site Investigator); Bo Xiao ( Xiangya Hospital Central‐South University, Site Investigator); Yibin Cao ( Tangshan Worker’s Hospital, Site Investigator); Yiping Wu ( The 1st Hospital In Handan, Site Investigator); Jinfeng Liu ( Yangquan Coal (Group) Co., Ltd. General Hospital, Site Investigator); Zhiming Zhang ( Tianjin Tianhe Hospital, Site Investigator); Zhengxie Dong ( Nantong First People’s Hospital, Site Investigator); Limin Wang ( The 1st Hospital Of Zhangjiakou City, Site Investigator); Li He ( West China Hospital, Sichuan University, Site Investigator); Xinchen Wang ( The Second Affiliated Hospital Of Shandong University Of TCM, Site Investigator); Xueying Guo ( Fenyang Hospital Of Shanxi Province, Site Investigator); Ming Wang ( Zhejiang Zhoushan Putuo District People’s Hospital, Site Investigator); Xiaosha Wang ( Xiyuan Hospital Of China Academy Of Chinese Traditional Medicine, Site Investigator); Jiandong Jiang ( No.2 People’s Hospital East In Lianyungang City, Site Investigator); Renliang Zhao ( Affiliated Hospital Of Qingdao University Medical College, Site Investigator); Shengnian Zhou ( Qilu Hospital Of Shandong University, Site Investigator); HaoHu ( Zibo Hospital Of Traditional Chinese Medicine, Site Investigator); Maolin He ( Beijing Shijitan Hospital, Site Investigator); Fengchun Yu ( Beijing Haidian Hospital, Site Investigator); Quping Ouyang ( Beijing Shunyi District Hospital, Site Investigator); Jingbo Zhang ( Dalian Third Municipal Hospital, Site Investigator); Anding Xu ( The First Affliated Hospital Of Jinan University, Site Investigator); Xiaokun Qi ( Navy Genaral Hospital Of P.L.A, Site Investigator); Lei Wang ( Beijing Second Artillery General Hospital, Site Investigator); Fuming Shi ( Beijing Daxing District Hospital, Site Investigator); Fuqiang Guo ( Sichuan Province People’s Hospital, Site Investigator); Jianfeng Wang ( Dalian Municipal Central Hospital, Site Investigator); Fengli Zhao ( The Second Hospital In Baoding, Site Investigator); Ronghua Dou ( The Hospital Combine Traditional Chinese And Western Medicine In Cang zhou, Site Investigator); Dongning Wei ( The 309th Hospital Of P.L.A, Site Investigator); Qingwei Meng ( Liangxiang Hospital Of Fangshan District, Beijing, Site Investigator); Yilu Xia ( HuaXin Hospital First Hospital Of Tsinghua University, Site Investigator); ShiminWang ( TianjinHuanhu Hospital, Site Investigator); Zhangcang Xue ( Shijiazhuang First Hospital, Site Investigator); Yuming Xu ( The First Affiliated Hospital Of Zhengzhou University, Site Investigator); Liping Ma ( Xinzhou City People’s Hospital, Site Investigator); Chun Wang ( Sichuan Province People’s Hospital Of Deyang City, Site Investigator); Jiang Wu ( First Hospital, Jilin University, Site Investigator); Yifeng Du ( Shandong Provincial Hospital, Site Investigator); Yinzhou Wang ( Fujian Province Hospital, Site Investigator); Lijun Xiao ( Liaoyang City Third People’s Hospital, Site Investigator); Fucong Song ( Handan City Center Hospital, Site Investigator); Wenli Hu ( Beijing Chaoyang Hospital, Capital Medical University, Site Investigator); Zhigang Chen ( Beijing University Of Chinese Medicine East Hospital, Site Investigator); Qingrui Liu ( Hebei Medical University Fourth Hospital, Site Investigator); Jiemin Zhang ( The Fourth Affiliated Hospital Of Soochow University, Site Investigator); Mei Chen ( Zhejiang University Of Chinese Medicine Affiliated First Hospital, Site Investigator); Xiaodong Yuan ( Affiliated Hospital Of Kailuan Company Ltd, Site Investigator); Zhihui Liu ( Affiliated Hospital Of Weifang Medical University, Site Investigator); Guozhong Li ( The First Hospital Of Harbin Medical University, Site Investigator); Xiaohong Li (Dalian Friendship Hospital, Site Investigator); Tingchen Tian (Tianjin Dagang Hospital, Site Investigator).