- Email: [email protected]
Discrepant relationships between admission blood pressure and mortality in different stroke subtypes
Journal of the Neurological Sciences 383 (2017) 47–51
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Discrepant relationships between admission blood pressure and mortality in different stroke subtypes
MARK
Yuanyuan Liua,h, Yide Yangc, Haiqiang Jina, Chenghe Fana, Pu Lva, Wei Suna, Qing Penga, Mingming Zhaob, David K. Jind, Jiguang Wange, Lawrence K.S. Wongf, Craig S. Andersong, Lemin Zhengb,⁎, Yining Huanga,⁎, for the ChinaQUEST (Quality Evaluation of Stroke Care and Treatment) Investigators a
Department of Neurology, Peking University First Hospital, Beijing 100034, China The Institute of Cardiovascular Sciences, Institute of Systems Biomedicine, School of Basic Medical Sciences, and Key Laboratory of Molecular Cardiovascular Sciences of Ministry of Education, Peking University Health Science Center, Beijing 100191, China c Institute of Child and Adolescent Health, School of Public Health, Peking University Health Science Center, Beijing 100191, China d Weill Cornell Medical College of Cornell University, New York 10021, USA e The Shanghai Institute of Hypertension, RuiJin Hospital, Shanghai Jiaotong University, Shanghai, China f Prince of Wales Hospital, Chinese University of Hong Kong, Hong Kong, China g The George Institute for Global Health, Royal Prince Alfred Hospital, University of Sydney, Sydney, Australia h Department of Neurology, The First Affiliated Hospital of Zhengzhou University, China b
A R T I C L E I N F O
A B S T R A C T
Keywords: Stroke Mortality Systolic blood pressure Diastolic blood pressure U-curve effect
The relationship between blood pressure(BP) and clinical outcome in patients with acute stroke is still controversial. The present study aimed to elucidate the impact of admission blood pressure on mortality in patients with acute stroke of different subtypes. Data were from ChinaQUEST (QUality Evaluation of Stroke Care and Treatment), a multicenter, prospective hospital registry study in 37 cities across China. A total of 6427 patients were admitted within 24 h of onset and after following up for 12 months, 5501 were included in the final analysis. Multivariate Cox regression model were used in data analysis. A “U-curve shaped” relationship was observed between admission systolic or diastolic BP and mortality at 12 months in the overall study population. Compared to first quartile, the Hazard ratio (HR) for the systolic BP of top quartile was 1.444 (95%CI 1.854–1.636), while the HR was 0.692 (95%CI 0.802–0.930) for the second quartile. Similar associations were observed when we applied admission diastolic BP. In subgroup analysis, the U-shaped effect was remained only in patients with intracranial hemorrhage (ICH). The HR for the systolic BP of top quartile was 2.274 (95%CI 1.878–2.755), while the HR was 0.751 (95%CI 0.571–0.986) for the second quartile. Moreover, admission diastolic BP of top quartile was significantly associated with elevated risk of death for patients with ischemic stroke caused by small vessel diseases (LACI)(HR 1.470; CI 1.040–2.078). In addition, we found a heterogeneity of the admission BP distribution among different subtypes, which may explain the “U-curve” effect.
1. Introduction A transient initially increase of blood pressure (BP) occurred in three quarters of patients with acute stroke and usually declines spontaneously within the first few days after admission [1,2].Whether the post-stroke hypertension was a physiological response to maintain perfusion of ischemic penumbra because of impaired auto regulation or a sigh of the poor outcome was still controversial [3,4]. Some studies have suggested that high BP on admission was associated with increased stroke mortality [5–7], poor functional outcome and stroke
⁎
recurrence, although this was not confirmed in other studies [3,8,9]. On the contrary, other studies have reported that low systolic BP may be associated with poor outcome [7,10]. The variable proportions of each stroke subtypes of subjects among studies may resulted in the conflicting conclusions. However, there were scarcely any investigations that have focused on different effects of admission BP on different subtypes of acute stroke. In the present study, patients were classified based on pathogenic mechanisms as large vessel atherosclerotic stroke (LVA), small vessel or lacunar infarct (LACI), cardioembolic stroke(CE), intracranial hemorrhage (ICH) and stroke of other etiology (e.g. retinal
Corresponding author at: Department of Neurology, Peking University First Hospital, Beijing 100034, China E-mail addresses: [email protected] (L. Zheng), [email protected] (Y. Huang).
http://dx.doi.org/10.1016/j.jns.2017.09.032 Received 24 April 2017; Accepted 21 September 2017 Available online 22 September 2017 0022-510X/ © 2017 Elsevier B.V. All rights reserved.
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
infarct, venous infarct, carotid dissection) or undetermined according to TOAST classification [11]. On the other hand, by the Oxfordshire Community Stroke Project(OCSP) classification [12], patients were classified into following groups including total anterior circulation syndrome(TACS), partial anterior circulation syndrome(PACS), lacunar infarction syndrome(LACS), and posterior circulation syndrome(POCS). The purpose of the present post hoc analysis of the ChinaQUEST study was to elucidate the impact of admission blood pressure on poor outcomes in different subtypes of acute stroke during 12 months followup. Moreover, we compared the admission BP distribution among different stroke subtypes.
Table 1 General characteristics of the study population. Survival
No. of patients Age Female Current smoking Regular drinker Hypertension history Atrial fibrillation history Diabetes mellitus history Stroke history TIA history Antihypertensive treatment Antiplatelet treatment Lipid-lowering treatment No. of antihypertensive agent use 0 1 2 3 4 Class of antihypertensive agent use ACEI/ARB CCB Diuretic β-Blocker Glasgow Coma Scale score 3–7 8–12 13–15 Stroke subtype LVA LACI CE ICH other or undetermined etiology Stroke location TACS PACS LACS POCS Uncertain
2. Subjects and method 2.1. Subjects ChinaQUEST (QUality Evaluation of Stroke Care and Treatment) is a multicenter, prospective hospital (n = 62, 14 Level 2 and 48 Level 3 hospitals of varying size: < 500 beds [26%], 500 to 1000 beds [34%], and > 1000 beds [40%]) registry study in 37 cities in China, as previously described [13,14]. Over a 5 month period in 2006, consecutive patients in China with an acute stroke within 24 h of onset were enrolled and assessed at the time of discharge, 3 months, and 12 months post-stroke. Data were collected by trained local staff according to a standardized protocol, and then transferred to a secure website located at The George Institute for Global Health in Sydney, Australia. This study was approved by the ethics committees of Peking University First Hospital (Beijing), Ruijin Hospital (Shanghai), Prince of Wales Hospital (Hong Kong), and The University of Sydney. Good Clinical Practice guidelines in accordance with the Declaration of Helsinki were used, and the privacy of patients was strictly protected. 2.2. Explanatory and outcome variables Baseline information was predominantly obtained by face-to-face interviews with patients and proxies or through medical record review. Follow-up details were obtained by telephone interviews. Death information was ascertained from family or through medical records, police records, or other sources. Supine blood pressure was measured by trained nurses at the upper arm on admission using an automated cuff or a mercury sphygmomanometer, and BP values were defined as the average values of three measurements. Demographic data (age, gender), medical history(hypertension, atrial fibrillation, diabetes mellitus, previous stroke, transient ischemic attack, coronary heart disease, including myocardial infarction, angina and heart failure) and vascular risk factors(current smoker, regular drinker), and current medications(antiplatelets, anticoagulants, antihypertensives, antidiabetics, and lipid-lowering drug) were evaluated in this study.
Death
P value
n
%
n
%
4878 63(mean) 1868 1407 1363 3563 228 784 1281 197 3183
12(SD) 38.30 28.9 27.9 73.0 4.7 16.1 26.3 4.0 65.3
623 68(mean) 259 126 117 468 67 90 228 21 359
12(SD) 41.6 20.3 18.8 75.1 10.8 14.4 36.6 3.4 57.6
< 0.001 0.132 < 0.001 < 0.001 0.23 < 0.001 0.268 < 0.001 0.417 < 0.001
3033 1345
62.2 27.60
227 80
36.4 12.80
< 0.001 < 0.001
1695 1872 1010 271 30
34.7 38.4 20.7 5.6 0.6
264 206 107 39 7
42.4 33.1 17.2 6.3 1.1
1566 2415 439 405
32.1 48.7 9.0 8.3
167 224 100 74
26.8 36.0 16.1 11.9
151 534 4136
3.1 10.9 84.8
159 167 283
25.5 26.8 45.4
788 2786 119 1054 131
16.2 57.1 2.4 21.6 2.7
146 165 45 253 14
23.4 26.5 7.2 40.6 2.2
509 2562 902 642 263
10.4% 52.5% 18.5% 13.2% 5.4%
160 282 45 90 46
25.7% 45.3% 7.2% 14.4% 7.4%
non-significant variables (P > 0.1) was subsequently used to generate a final model in the Cox regression analysis. All statistical analyses were performed using SPSS version 20.0. 3. Results
2.3. Statistical analysis 3.1. Baseline characteristics Continuous variables were displayed as means ( ± standard deviation) and categorical variables as frequency (percentage). Baseline characteristics were compared between different subgroups using the Mann-Whitney U test and the Chi-square test for continuous and categorical variables. Mortality from all causes was analyzed at 12 months, and the time since onset of the stroke was used as the time scale. Multivariate Cox proportional hazard models were used to assess the impact of the hypertension stage on survival, with adjustment of risk factors including age, gender, history of stroke, history of coronary heart disease, history of atrial fibrillation, history of heart failure, Glasgow Coma Scale(GCS) score, anti-hypertension medication, antiplatelet medication and use of lipid lowering agents. All P values were two-sided, with P < 0.05 considered statistically significant. Candidate covariates included all variables with P < 0.05 for association with survival in bivariable analysis. Backward elimination of
A total of 6427 patients were screened for enrollment. 345 patients dropped out of the study; 578 patients were excluded because of a lack of BP information; 3 patients were excluded because they were under 18 years of age. A total of 5501 patients were finally included in our analysis and 623 patients died during follow-up. 32 died during time of hospitalization, while 387 patients and 204 patients died during 3 months and 12 months follow-up period, respectively. The detailed demographic characteristics are presented in Table 1. The main causes of death were cerebrovascular diseases (374 cases), following by injuries or other causes(143 cases), infection (62cases) and cardiovascular diseases(42 cases). The mean age of patients who died were 68 years old(SD,12),older than those who survived. Moreover, the atrial fibrillation history, stroke history, use of medications, Glasgow Coma Scale score, and stroke subtype distribution were significantly different 48
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
Fig. 1. Mortality stratified by systolic and diastolic blood pressure. It revealed a U-curve pattern between SBP or DBP on admission and mortality at 12 months in patients with acute stroke.
lower for Q2.(HR0.692, 95%CI 0.802–0.930; Table 2,). Similar associations were observed when we applied admission DBP. Compared to Q1,HR for patients with diastolic BP of Q4 was 1.586, with 95%CI ranging from 1.823 to 2.094; and HR for Q2 was 0.643, with 95%CI ranging from 0.734 to 0.837 (Table 2). A U-shaped association between admission BP and mortality was revealed in overall follow-up analysis (Fig. 1). In analysis for different subtypes, it was observed that the U-curve effect was only remained in patients with ICH (Table 3). Compared to Q1, HR for patients with systolic BP of Q4 was 2.274, with 95%CI ranging from 1.878 to 2.755; and HR for Q2 was 0.751, with 95%CI ranging from 0.571 to 0.986. In addition, patients with ICH were found to be more vulnerable to high SBP than patients with ischemic stroke. Compared to reference (SBP 121–129 mmHg), SBP above 150 mmHg had significant higher risk for death in patients with ICH, while much more higher admission BP levels were significantly associated with elevated risk of death in patients with either LACI or LVA after adjusting for potential confounding factors (Supplementary Table 1).However, patients with LACI were more tolerant to high SBP, SBP above 170 mmHg started to be in higher risk while in patients with LVA, SBP above 160 mmHg was the threshold of the risk increase (Supplementary Table 1). Moreover, higher DBP was found to be association with higher risk for death for patients of LACI, and the HR for patients with BP of Q4 was 1.470, with 95%CI ranging from 1.040 to 2.078.(Table 3). Similar U-shaped associations were observed in patients with ICH when we applied admission DBP (Table 3).
between these two groups. 3.2. Mortality distribution stratified by admission blood pressure Mortality distribution stratified by systolic and diastolic blood pressure was displayed in Fig. 1. It was observed that the lowest mortality rate occurred in patients with systolic BP ranging from 130 to 139 mmHg and diastolic BP ranging from 70 to 79 mmHg, presenting a U shaped pattern. 3.3. Associations between admission blood pressure and stroke mortality The BP were categorized into 4 groups(Q1-Q4)according to the quartile, the cut-off value were 126 mmHg, 135 mmHg, or 146 mmHg for systolic BP and 75 mmHg, 80 mmHg, 90 mmHg, respectively. Compared to Q1, the risk of death was higher for patients with admission systolic BP of Q4 (HR 1.444, 95%CI 1.854–1.636;Table 2,) and Table 2 Multivariable cox regression analyses of admission BP and relationship with mortality at 12 months. Variable
SBP
DBP
SBP
DBP
HR
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 120–129 Lowest-99 100–109 110–119 130–139 140–149 150–159 160–169 170-hignest 70–79 Lowest-59 60–69 80–89 90–99 100–109 110-highest
95%CI
P value
low
high
0.692 0.739 1.444
0.802 0.860 1.636
0.930 1.003 1.854
0.004 0.054 < 0.001
0.643 0.826 1.586
0.734 0.952 1.823
0.837 1.097 2.094
< 0.001 0.494 < 0.001
3.346 1.552 0.846 0.854 1.021 1.386 1.605 3.112
1.811 0.92 0.56 0.65 0.771 1.01 1.124 2.342
6.182 2.62 1.279 1.122 1.352 1.902 2.292 4.136
< 0.001 0.1 0.428 0.258 0.886 0.043 0.009 < 0.001
1.096 0.852 0.654 0.857 1.52 2.47
0.513 0.644 0.541 0.691 1.189 1.862
2.342 1.127 0.791 1.062 1.942 3.276
0.812 0.262 < 0.001 0.158 0.001 < 0.001
3.4. Variable admission BP distribution patterns in different stroke subtypes Furthermore, we compared the admission BP distribution among different stroke subtypes. Interestingly, it was shown that patients with ICH, accounted for the highest proportion in systolic BP above 170 mmHg or diastolic BP above 110 mmHg, whilst the percentage of patients with LVA and CE and stroke of other or undetermined mechanism was higher in systolic BP below 110 mmHg or diastolic BP below 70 mmHg (Fig. 2a, b). Patients with these stroke subtypes were prone to be with admission BP extremes in one direction. On the other hand, admission BP in patient with LACI was found to be distributed as an inverted “U-shaped curve”, with higher proportion in the proper BP range. Our results showed no difference in admission BP distribution among stroke subtypes classified by different damaged region (Fig. 2c, d). 4. Discussion The present study demonstrated that in a series of consecutive and non-selected patients, the relationships between admission BP and mortality were discrepant among different stroke subtypes. It revealed a “U-curve” relationship between admission BP and long-term mortality
CI, confidence interval; HR, hazard ratio; adjustment for risk factors including age, gender, history of stroke, history of coronary heart disease, history of atrial fibrillation, history of heart failure, Glasgow Coma Scale score, anti-hypertension medication, antiplatelet medication and use of lipid lowering agents.
49
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
Table 3 Multivariable Cox Regression Analyses of admission BP and relationship with mortality at 12 months in different subtypes of strokes. LVA HR
SBP
DBP
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
LACI 95%CI
p
low
high
0.749 1.261 1.309
0.545 0.946 0.984
1.031 1.680 1.742
0.076 0.113 0.065
0.894 1.128 1.251
0.688 0.839 0.855
1.163 1.516 1.832
0.404 0.426 0.249
HR
CE 95%CI
p
HR
low
high
0.833 0.978 1.223
0.639 0.744 0.934
1.086 1.284 1.603
0.176 0.870 0.144
0.820 1.137 1.470
0.643 0.866 1.040
1.045 1.493 2.078
0.108 0.357 0.029
ICH 95%CI
p
low
high
1.473 0.705 0.781
0.889 0.339 0.409
2.440 1.464 1.492
0.133 0.348 0.455
0.678 1.033 1.440
0.386 0.567 0.736
1.192 1.882 2.815
0.177 0.915 0.287
HR
95%CI
p
low
high
0.751 0.688 2.274
0.571 0.514 1.878
0.986 0.920 2.755
0.039 0.012 0.000
0.576 0.874 2.075
0.450 0.691 1.710
0.737 1.107 2.520
0.000 0.265 0.000
CI, confidence interval; HR, hazard ratio; adjustment for risk factors including age, gender, history of stroke, history of coronary heart disease, history of atrial fibrillation, history of heart failure, Glasgow Coma Scale score, anti-hypertension medication, anti-platelet medication and use of lipid lowering agents.
(stroke caused by CE or LVA). Few investigations have took the influence of different pathogenesis and site of stroke into account, which may resulted in the conflicting result. In the present study, patients with ICH were found to be more vulnerable to high SBP. While in patients with LVA, threshold of SBP level for higher risk was located between those of ICH and LACI. It can presumably be explained by hypothesis that raised SBP might increase the evolution of bleeding, rebleeding, hemorrhagic transformation in ischemic and edema in brain, which were the most common causes of death in patients with ICH or LVA [18]. The management of BP in the acute phase of stroke represents a long-standing debate. INTERACT2 found that early intensive lowering of blood pressure improved functional outcomes in ICH patients [19]. However, there are no published clinical trials with sufficient power to confirm the favorable effect of immediate BP reduction on clinical outcomes among patients with acute ischemic stroke [20–23]. Further studies should address on effect of BP reduction treatment among different ischemic stroke subtypes because of the inconformity of the tolerance to high SBP. In addition, our study confirm a heterogeneity of the admission BP distribution among different subtypes of acute stroke. Patients with
in the overall study population. In subgroup analysis, the U-shaped effect was remained in patients with ICH. However, only higher admission diastolic BP levels was found to be significantly associated with elevated risk of death in patients with LACI after adjusting for potential confounding factors. The influence of BP during the acute phase of ischemic stroke is still a matter of controversy. Previous studies has widely described an independent association between poor functional outcome and high systolic BP [15]. High admission BP may favor the cerebral edema, hemorrhagic transformation of the ischemic area and development of cerebral reinfarction [7]. On the other hand, some observational studies also demonstrated that the relationship between BP and prognosis was U-shaped in patients with either brain infarction or hemorrhage [7,10]. The usual explanation was that low admission BP was associated with an excess of early death from ischemic myocardial event and early cerebral reinfarction [16,17]. Moreover, the proportion of cardiovascular diseases as the primary cause of death in patients with low BP is significantly higher than those with high BP, presumably reflecting the effect of this comorbidity although after adjustment for the cardiovascular risk factors. Our results revealed that low admission BP were common in patients with severe clinical ischemic stroke subtypes
Fig. 2. The various patterns of distribution of admission BP among different stroke subtypes.
50
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
Reference
stroke caused by LVA and CE were more likely to have a lower BP while BP of patients with ICH were prone to be extreme in the higher direction. Whilst admission BP of patients with LACI who usually had the less severe neurological damage at baseline was mostly distributed between the two extreme ends. This phenomenon may explain the “Ucurve” effect of admission blood pressure on mortality rate in the overall analysis. Various proportions of stroke subtypes among different studies may lead to conflicting conclusions. However, no effect of stroke subtype classified by OCSP has been found on the admission BP value, which was different to the previous study [24]. There were several limitations in our study. Firstly, it was a post hoc analysis of a stroke registration study in China, further clinical trials were required to confirm the effect of admission hypertension or hypotension on outcomes in patients with different stroke subtypes. Secondly, although multivariable analyses had been adjusted for established predictive factors, other potential factors we did not control might be confounding variables. For instance, NIHSS score and history of carotid stenosis had not been assessed in this study. Thirdly, the BP level was only measured at single time point and prehospital use of antihypertesion agents could be other confounding factors that has not been exactly described in the present study.
[1] J.D. Wallace, L.L. Levy, Blood pressure after stroke, JAMA 246 (1981) 2177–2180. [2] M. Britton, A. Carlsson, U. de Faire, Blood pressure course in patients with acute stroke and matched controls, Stroke 17 (1986) 861–864. [3] B. Carlberg, K. Asplund, E. Hagg, The prognostic value of admission blood pressure in patients with acute stroke, Stroke 24 (1993) 1372–1375. [4] G. Harper, C.M. Castleden, J.F. Potter, Factors affecting changes in blood pressure after acute stroke, Stroke 25 (1994) 1726–1729. [5] N. Ahmed, G. Wahlgren, High initial blood pressure after acute stroke is associated with poor functional outcome, J. Intern. Med. 249 (2001) 467–473. [6] K. Irie, T. Yamaguchi, K. Minematsu, T. Omae, The J-curve phenomenon in stroke recurrence, Stroke 24 (1993) 1844–1849. [7] J. Leonardi-Bee, P.M. Bath, S.J. Phillips, P.A. Sandercock, Blood pressure and clinical outcomes in the International Stroke Trial, Stroke 33 (2002) 1315–1320. [8] S. Tuhrim, J.M. Dambrosia, T.R. Price, J.P. Mohr, P.A. Wolf, A. Heyman, et al., Prediction of intracerebral hemorrhage survival, Ann. Neurol. 24 (1988) 258–263. [9] Y. Zhang, K.H. Reilly, W. Tong, T. Xu, J. Chen, L.A. Bazzano, et al., Blood pressure and clinical outcome among patients with acute stroke in Inner Mongolia, China, J. Hypertens. 26 (2008) 1446–1452. [10] K.N. Vemmos, G. Tsivgoulis, K. Spengos, N. Zakopoulos, A. Synetos, E. Manios, et al., U-shaped relationship between mortality and admission blood pressure in patients with acute stroke, J. Intern. Med. 255 (2004) 257–265. [11] H.P. Adams Jr., B.H. Bendixen, L.J. Kappelle, J. Biller, B.B. Love, D.L. Gordon, et al., Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment, Stroke 24 (1993) 35–41. [12] J. Bamford, P. Sandercock, M. Dennis, J. Burn, C. Warlow, Classification and natural history of clinically identifiable subtypes of cerebral infarction, Lancet (London, England). 337 (1991) 1521–1526. [13] H. Jin, S. Zhu, J.W. Wei, J. Wang, M. Liu, Y. Wu, et al., Factors associated with prehospital delays in the presentation of acute stroke in urban China, Stroke 43 (2012) 362–370. [14] E. Heeley, C.S. Anderson, Y. Huang, S. Jan, Y. Li, M. Liu, et al., Role of health insurance in averting economic hardship in families after acute stroke in China, Stroke 40 (2009) 2149–2156. [15] M. Willmot, J. Leonardi-Bee, P.M. Bath, High blood pressure in acute stroke and subsequent outcome: a systematic review, Hypertension 43 (2004) 18–24. [16] H.S. Jorgensen, H. Nakayama, H.R. Christensen, H.O. Raaschou, J.P. Kampmann, T.S. Olsen, Blood pressure in acute stroke. The Copenhagen Stroke Study, Cerebrovasc. Dis. (Basel, Switzerland). 13 (2002) 204–209. [17] D.R. Lisk, J.C. Grotta, L.M. Lamki, H.D. Tran, J.W. Taylor, D.A. Molony, et al., Should hypertension be treated after acute stroke? A randomized controlled trial using single photon emission computed tomography, Arch. Neurol. 50 (1993) 855–862. [18] A.I. Qureshi, Acute hypertensive response in patients with stroke: pathophysiology and management, Circulation 118 (2008) 176–187. [19] L. Manning, Y. Hirakawa, H. Arima, X. Wang, J. Chalmers, J. Wang, et al., Blood pressure variability and outcome after acute intracerebral haemorrhage: a post-hoc analysis of INTERACT2, a randomised controlled trial, The Lancet Neurol. 13 (2014) 364–373. [20] P.M. Bath, L. Woodhouse, P. Scutt, K. Krishnan, J.M. Wardlaw, D. Bereczki, et al., Efficacy of nitric oxide, with or without continuing antihypertensive treatment, for management of high blood pressure in acute stroke (ENOS): a partial-factorial randomised controlled trial, Lancet (London, England) 385 (2015) 617–628. [21] E.C. Sandset, P.M. Bath, G. Boysen, D. Jatuzis, J. Korv, S. Luders, et al., The angiotensin-receptor blocker candesartan for treatment of acute stroke (SCAST): a randomised, placebo-controlled, double-blind trial, Lancet (London, England) 377 (2011) 741–750. [22] J. He, Y. Zhang, T. Xu, Q. Zhao, D. Wang, C.S. Chen, et al., Effects of immediate blood pressure reduction on death and major disability in patients with acute ischemic stroke: the CATIS randomized clinical trial, JAMA 311 (2014) 479–489. [23] T.G. Robinson, J.F. Potter, G.A. Ford, C.J. Bulpitt, J. Chernova, C. Jagger, et al., Effects of antihypertensive treatment after acute stroke in the continue or Stop PostStroke Antihypertensives Collaborative Study (COSSACS): a prospective, randomised, open, blinded-endpoint trial, The Lancet Neurol. 9 (2010) 767–775. [24] A. Semplicini, A. Maresca, G. Boscolo, M. Sartori, R. Rocchi, V. Giantin, et al., Hypertension in acute ischemic stroke: a compensatory mechanism or an additional damaging factor? Arch. Intern. Med. 163 (2003) 211–216.
5. Conclusion In accordance to several previous reports, our results revealed a Ucurve relationship between admission BP and poor outcome in patients with acute stroke in the overall analysis. The U-shaped effect was remained only in patients with ICH. On the contrary, only higher admission diastolic BP levels were found significantly associated with elevated risk of death for patients with LACI. In addition, our study confirm a heterogeneity of the admission BP distribution among different subtypes of acute stroke. Further randomised clinical trials addressing the relationship between antihypertensive treatment in patients with different subtypes of stroke are required. Funding The ChinaQUEST study was supported by grants from 81170101 from the National Natural Science Foundation of China; the Macquarie Bank Foundation, The George Foundation, and AstraZeneca Pharmaceutical Chinaand was supported by theNational Natural Science Foundation of China(91639108,81370235,81770272), and by theNational High Technology Research and Development Program of China(2016YFC0903000). Disclosures There is no conflict of interest. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jns.2017.09.032.
51
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Discrepant relationships between admission blood pressure and mortality in different stroke subtypes
MARK
Yuanyuan Liua,h, Yide Yangc, Haiqiang Jina, Chenghe Fana, Pu Lva, Wei Suna, Qing Penga, Mingming Zhaob, David K. Jind, Jiguang Wange, Lawrence K.S. Wongf, Craig S. Andersong, Lemin Zhengb,⁎, Yining Huanga,⁎, for the ChinaQUEST (Quality Evaluation of Stroke Care and Treatment) Investigators a
Department of Neurology, Peking University First Hospital, Beijing 100034, China The Institute of Cardiovascular Sciences, Institute of Systems Biomedicine, School of Basic Medical Sciences, and Key Laboratory of Molecular Cardiovascular Sciences of Ministry of Education, Peking University Health Science Center, Beijing 100191, China c Institute of Child and Adolescent Health, School of Public Health, Peking University Health Science Center, Beijing 100191, China d Weill Cornell Medical College of Cornell University, New York 10021, USA e The Shanghai Institute of Hypertension, RuiJin Hospital, Shanghai Jiaotong University, Shanghai, China f Prince of Wales Hospital, Chinese University of Hong Kong, Hong Kong, China g The George Institute for Global Health, Royal Prince Alfred Hospital, University of Sydney, Sydney, Australia h Department of Neurology, The First Affiliated Hospital of Zhengzhou University, China b
A R T I C L E I N F O
A B S T R A C T
Keywords: Stroke Mortality Systolic blood pressure Diastolic blood pressure U-curve effect
The relationship between blood pressure(BP) and clinical outcome in patients with acute stroke is still controversial. The present study aimed to elucidate the impact of admission blood pressure on mortality in patients with acute stroke of different subtypes. Data were from ChinaQUEST (QUality Evaluation of Stroke Care and Treatment), a multicenter, prospective hospital registry study in 37 cities across China. A total of 6427 patients were admitted within 24 h of onset and after following up for 12 months, 5501 were included in the final analysis. Multivariate Cox regression model were used in data analysis. A “U-curve shaped” relationship was observed between admission systolic or diastolic BP and mortality at 12 months in the overall study population. Compared to first quartile, the Hazard ratio (HR) for the systolic BP of top quartile was 1.444 (95%CI 1.854–1.636), while the HR was 0.692 (95%CI 0.802–0.930) for the second quartile. Similar associations were observed when we applied admission diastolic BP. In subgroup analysis, the U-shaped effect was remained only in patients with intracranial hemorrhage (ICH). The HR for the systolic BP of top quartile was 2.274 (95%CI 1.878–2.755), while the HR was 0.751 (95%CI 0.571–0.986) for the second quartile. Moreover, admission diastolic BP of top quartile was significantly associated with elevated risk of death for patients with ischemic stroke caused by small vessel diseases (LACI)(HR 1.470; CI 1.040–2.078). In addition, we found a heterogeneity of the admission BP distribution among different subtypes, which may explain the “U-curve” effect.
1. Introduction A transient initially increase of blood pressure (BP) occurred in three quarters of patients with acute stroke and usually declines spontaneously within the first few days after admission [1,2].Whether the post-stroke hypertension was a physiological response to maintain perfusion of ischemic penumbra because of impaired auto regulation or a sigh of the poor outcome was still controversial [3,4]. Some studies have suggested that high BP on admission was associated with increased stroke mortality [5–7], poor functional outcome and stroke
⁎
recurrence, although this was not confirmed in other studies [3,8,9]. On the contrary, other studies have reported that low systolic BP may be associated with poor outcome [7,10]. The variable proportions of each stroke subtypes of subjects among studies may resulted in the conflicting conclusions. However, there were scarcely any investigations that have focused on different effects of admission BP on different subtypes of acute stroke. In the present study, patients were classified based on pathogenic mechanisms as large vessel atherosclerotic stroke (LVA), small vessel or lacunar infarct (LACI), cardioembolic stroke(CE), intracranial hemorrhage (ICH) and stroke of other etiology (e.g. retinal
Corresponding author at: Department of Neurology, Peking University First Hospital, Beijing 100034, China E-mail addresses: [email protected] (L. Zheng), [email protected] (Y. Huang).
http://dx.doi.org/10.1016/j.jns.2017.09.032 Received 24 April 2017; Accepted 21 September 2017 Available online 22 September 2017 0022-510X/ © 2017 Elsevier B.V. All rights reserved.
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
infarct, venous infarct, carotid dissection) or undetermined according to TOAST classification [11]. On the other hand, by the Oxfordshire Community Stroke Project(OCSP) classification [12], patients were classified into following groups including total anterior circulation syndrome(TACS), partial anterior circulation syndrome(PACS), lacunar infarction syndrome(LACS), and posterior circulation syndrome(POCS). The purpose of the present post hoc analysis of the ChinaQUEST study was to elucidate the impact of admission blood pressure on poor outcomes in different subtypes of acute stroke during 12 months followup. Moreover, we compared the admission BP distribution among different stroke subtypes.
Table 1 General characteristics of the study population. Survival
No. of patients Age Female Current smoking Regular drinker Hypertension history Atrial fibrillation history Diabetes mellitus history Stroke history TIA history Antihypertensive treatment Antiplatelet treatment Lipid-lowering treatment No. of antihypertensive agent use 0 1 2 3 4 Class of antihypertensive agent use ACEI/ARB CCB Diuretic β-Blocker Glasgow Coma Scale score 3–7 8–12 13–15 Stroke subtype LVA LACI CE ICH other or undetermined etiology Stroke location TACS PACS LACS POCS Uncertain
2. Subjects and method 2.1. Subjects ChinaQUEST (QUality Evaluation of Stroke Care and Treatment) is a multicenter, prospective hospital (n = 62, 14 Level 2 and 48 Level 3 hospitals of varying size: < 500 beds [26%], 500 to 1000 beds [34%], and > 1000 beds [40%]) registry study in 37 cities in China, as previously described [13,14]. Over a 5 month period in 2006, consecutive patients in China with an acute stroke within 24 h of onset were enrolled and assessed at the time of discharge, 3 months, and 12 months post-stroke. Data were collected by trained local staff according to a standardized protocol, and then transferred to a secure website located at The George Institute for Global Health in Sydney, Australia. This study was approved by the ethics committees of Peking University First Hospital (Beijing), Ruijin Hospital (Shanghai), Prince of Wales Hospital (Hong Kong), and The University of Sydney. Good Clinical Practice guidelines in accordance with the Declaration of Helsinki were used, and the privacy of patients was strictly protected. 2.2. Explanatory and outcome variables Baseline information was predominantly obtained by face-to-face interviews with patients and proxies or through medical record review. Follow-up details were obtained by telephone interviews. Death information was ascertained from family or through medical records, police records, or other sources. Supine blood pressure was measured by trained nurses at the upper arm on admission using an automated cuff or a mercury sphygmomanometer, and BP values were defined as the average values of three measurements. Demographic data (age, gender), medical history(hypertension, atrial fibrillation, diabetes mellitus, previous stroke, transient ischemic attack, coronary heart disease, including myocardial infarction, angina and heart failure) and vascular risk factors(current smoker, regular drinker), and current medications(antiplatelets, anticoagulants, antihypertensives, antidiabetics, and lipid-lowering drug) were evaluated in this study.
Death
P value
n
%
n
%
4878 63(mean) 1868 1407 1363 3563 228 784 1281 197 3183
12(SD) 38.30 28.9 27.9 73.0 4.7 16.1 26.3 4.0 65.3
623 68(mean) 259 126 117 468 67 90 228 21 359
12(SD) 41.6 20.3 18.8 75.1 10.8 14.4 36.6 3.4 57.6
< 0.001 0.132 < 0.001 < 0.001 0.23 < 0.001 0.268 < 0.001 0.417 < 0.001
3033 1345
62.2 27.60
227 80
36.4 12.80
< 0.001 < 0.001
1695 1872 1010 271 30
34.7 38.4 20.7 5.6 0.6
264 206 107 39 7
42.4 33.1 17.2 6.3 1.1
1566 2415 439 405
32.1 48.7 9.0 8.3
167 224 100 74
26.8 36.0 16.1 11.9
151 534 4136
3.1 10.9 84.8
159 167 283
25.5 26.8 45.4
788 2786 119 1054 131
16.2 57.1 2.4 21.6 2.7
146 165 45 253 14
23.4 26.5 7.2 40.6 2.2
509 2562 902 642 263
10.4% 52.5% 18.5% 13.2% 5.4%
160 282 45 90 46
25.7% 45.3% 7.2% 14.4% 7.4%
non-significant variables (P > 0.1) was subsequently used to generate a final model in the Cox regression analysis. All statistical analyses were performed using SPSS version 20.0. 3. Results
2.3. Statistical analysis 3.1. Baseline characteristics Continuous variables were displayed as means ( ± standard deviation) and categorical variables as frequency (percentage). Baseline characteristics were compared between different subgroups using the Mann-Whitney U test and the Chi-square test for continuous and categorical variables. Mortality from all causes was analyzed at 12 months, and the time since onset of the stroke was used as the time scale. Multivariate Cox proportional hazard models were used to assess the impact of the hypertension stage on survival, with adjustment of risk factors including age, gender, history of stroke, history of coronary heart disease, history of atrial fibrillation, history of heart failure, Glasgow Coma Scale(GCS) score, anti-hypertension medication, antiplatelet medication and use of lipid lowering agents. All P values were two-sided, with P < 0.05 considered statistically significant. Candidate covariates included all variables with P < 0.05 for association with survival in bivariable analysis. Backward elimination of
A total of 6427 patients were screened for enrollment. 345 patients dropped out of the study; 578 patients were excluded because of a lack of BP information; 3 patients were excluded because they were under 18 years of age. A total of 5501 patients were finally included in our analysis and 623 patients died during follow-up. 32 died during time of hospitalization, while 387 patients and 204 patients died during 3 months and 12 months follow-up period, respectively. The detailed demographic characteristics are presented in Table 1. The main causes of death were cerebrovascular diseases (374 cases), following by injuries or other causes(143 cases), infection (62cases) and cardiovascular diseases(42 cases). The mean age of patients who died were 68 years old(SD,12),older than those who survived. Moreover, the atrial fibrillation history, stroke history, use of medications, Glasgow Coma Scale score, and stroke subtype distribution were significantly different 48
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
Fig. 1. Mortality stratified by systolic and diastolic blood pressure. It revealed a U-curve pattern between SBP or DBP on admission and mortality at 12 months in patients with acute stroke.
lower for Q2.(HR0.692, 95%CI 0.802–0.930; Table 2,). Similar associations were observed when we applied admission DBP. Compared to Q1,HR for patients with diastolic BP of Q4 was 1.586, with 95%CI ranging from 1.823 to 2.094; and HR for Q2 was 0.643, with 95%CI ranging from 0.734 to 0.837 (Table 2). A U-shaped association between admission BP and mortality was revealed in overall follow-up analysis (Fig. 1). In analysis for different subtypes, it was observed that the U-curve effect was only remained in patients with ICH (Table 3). Compared to Q1, HR for patients with systolic BP of Q4 was 2.274, with 95%CI ranging from 1.878 to 2.755; and HR for Q2 was 0.751, with 95%CI ranging from 0.571 to 0.986. In addition, patients with ICH were found to be more vulnerable to high SBP than patients with ischemic stroke. Compared to reference (SBP 121–129 mmHg), SBP above 150 mmHg had significant higher risk for death in patients with ICH, while much more higher admission BP levels were significantly associated with elevated risk of death in patients with either LACI or LVA after adjusting for potential confounding factors (Supplementary Table 1).However, patients with LACI were more tolerant to high SBP, SBP above 170 mmHg started to be in higher risk while in patients with LVA, SBP above 160 mmHg was the threshold of the risk increase (Supplementary Table 1). Moreover, higher DBP was found to be association with higher risk for death for patients of LACI, and the HR for patients with BP of Q4 was 1.470, with 95%CI ranging from 1.040 to 2.078.(Table 3). Similar U-shaped associations were observed in patients with ICH when we applied admission DBP (Table 3).
between these two groups. 3.2. Mortality distribution stratified by admission blood pressure Mortality distribution stratified by systolic and diastolic blood pressure was displayed in Fig. 1. It was observed that the lowest mortality rate occurred in patients with systolic BP ranging from 130 to 139 mmHg and diastolic BP ranging from 70 to 79 mmHg, presenting a U shaped pattern. 3.3. Associations between admission blood pressure and stroke mortality The BP were categorized into 4 groups(Q1-Q4)according to the quartile, the cut-off value were 126 mmHg, 135 mmHg, or 146 mmHg for systolic BP and 75 mmHg, 80 mmHg, 90 mmHg, respectively. Compared to Q1, the risk of death was higher for patients with admission systolic BP of Q4 (HR 1.444, 95%CI 1.854–1.636;Table 2,) and Table 2 Multivariable cox regression analyses of admission BP and relationship with mortality at 12 months. Variable
SBP
DBP
SBP
DBP
HR
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 120–129 Lowest-99 100–109 110–119 130–139 140–149 150–159 160–169 170-hignest 70–79 Lowest-59 60–69 80–89 90–99 100–109 110-highest
95%CI
P value
low
high
0.692 0.739 1.444
0.802 0.860 1.636
0.930 1.003 1.854
0.004 0.054 < 0.001
0.643 0.826 1.586
0.734 0.952 1.823
0.837 1.097 2.094
< 0.001 0.494 < 0.001
3.346 1.552 0.846 0.854 1.021 1.386 1.605 3.112
1.811 0.92 0.56 0.65 0.771 1.01 1.124 2.342
6.182 2.62 1.279 1.122 1.352 1.902 2.292 4.136
< 0.001 0.1 0.428 0.258 0.886 0.043 0.009 < 0.001
1.096 0.852 0.654 0.857 1.52 2.47
0.513 0.644 0.541 0.691 1.189 1.862
2.342 1.127 0.791 1.062 1.942 3.276
0.812 0.262 < 0.001 0.158 0.001 < 0.001
3.4. Variable admission BP distribution patterns in different stroke subtypes Furthermore, we compared the admission BP distribution among different stroke subtypes. Interestingly, it was shown that patients with ICH, accounted for the highest proportion in systolic BP above 170 mmHg or diastolic BP above 110 mmHg, whilst the percentage of patients with LVA and CE and stroke of other or undetermined mechanism was higher in systolic BP below 110 mmHg or diastolic BP below 70 mmHg (Fig. 2a, b). Patients with these stroke subtypes were prone to be with admission BP extremes in one direction. On the other hand, admission BP in patient with LACI was found to be distributed as an inverted “U-shaped curve”, with higher proportion in the proper BP range. Our results showed no difference in admission BP distribution among stroke subtypes classified by different damaged region (Fig. 2c, d). 4. Discussion The present study demonstrated that in a series of consecutive and non-selected patients, the relationships between admission BP and mortality were discrepant among different stroke subtypes. It revealed a “U-curve” relationship between admission BP and long-term mortality
CI, confidence interval; HR, hazard ratio; adjustment for risk factors including age, gender, history of stroke, history of coronary heart disease, history of atrial fibrillation, history of heart failure, Glasgow Coma Scale score, anti-hypertension medication, antiplatelet medication and use of lipid lowering agents.
49
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
Table 3 Multivariable Cox Regression Analyses of admission BP and relationship with mortality at 12 months in different subtypes of strokes. LVA HR
SBP
DBP
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
LACI 95%CI
p
low
high
0.749 1.261 1.309
0.545 0.946 0.984
1.031 1.680 1.742
0.076 0.113 0.065
0.894 1.128 1.251
0.688 0.839 0.855
1.163 1.516 1.832
0.404 0.426 0.249
HR
CE 95%CI
p
HR
low
high
0.833 0.978 1.223
0.639 0.744 0.934
1.086 1.284 1.603
0.176 0.870 0.144
0.820 1.137 1.470
0.643 0.866 1.040
1.045 1.493 2.078
0.108 0.357 0.029
ICH 95%CI
p
low
high
1.473 0.705 0.781
0.889 0.339 0.409
2.440 1.464 1.492
0.133 0.348 0.455
0.678 1.033 1.440
0.386 0.567 0.736
1.192 1.882 2.815
0.177 0.915 0.287
HR
95%CI
p
low
high
0.751 0.688 2.274
0.571 0.514 1.878
0.986 0.920 2.755
0.039 0.012 0.000
0.576 0.874 2.075
0.450 0.691 1.710
0.737 1.107 2.520
0.000 0.265 0.000
CI, confidence interval; HR, hazard ratio; adjustment for risk factors including age, gender, history of stroke, history of coronary heart disease, history of atrial fibrillation, history of heart failure, Glasgow Coma Scale score, anti-hypertension medication, anti-platelet medication and use of lipid lowering agents.
(stroke caused by CE or LVA). Few investigations have took the influence of different pathogenesis and site of stroke into account, which may resulted in the conflicting result. In the present study, patients with ICH were found to be more vulnerable to high SBP. While in patients with LVA, threshold of SBP level for higher risk was located between those of ICH and LACI. It can presumably be explained by hypothesis that raised SBP might increase the evolution of bleeding, rebleeding, hemorrhagic transformation in ischemic and edema in brain, which were the most common causes of death in patients with ICH or LVA [18]. The management of BP in the acute phase of stroke represents a long-standing debate. INTERACT2 found that early intensive lowering of blood pressure improved functional outcomes in ICH patients [19]. However, there are no published clinical trials with sufficient power to confirm the favorable effect of immediate BP reduction on clinical outcomes among patients with acute ischemic stroke [20–23]. Further studies should address on effect of BP reduction treatment among different ischemic stroke subtypes because of the inconformity of the tolerance to high SBP. In addition, our study confirm a heterogeneity of the admission BP distribution among different subtypes of acute stroke. Patients with
in the overall study population. In subgroup analysis, the U-shaped effect was remained in patients with ICH. However, only higher admission diastolic BP levels was found to be significantly associated with elevated risk of death in patients with LACI after adjusting for potential confounding factors. The influence of BP during the acute phase of ischemic stroke is still a matter of controversy. Previous studies has widely described an independent association between poor functional outcome and high systolic BP [15]. High admission BP may favor the cerebral edema, hemorrhagic transformation of the ischemic area and development of cerebral reinfarction [7]. On the other hand, some observational studies also demonstrated that the relationship between BP and prognosis was U-shaped in patients with either brain infarction or hemorrhage [7,10]. The usual explanation was that low admission BP was associated with an excess of early death from ischemic myocardial event and early cerebral reinfarction [16,17]. Moreover, the proportion of cardiovascular diseases as the primary cause of death in patients with low BP is significantly higher than those with high BP, presumably reflecting the effect of this comorbidity although after adjustment for the cardiovascular risk factors. Our results revealed that low admission BP were common in patients with severe clinical ischemic stroke subtypes
Fig. 2. The various patterns of distribution of admission BP among different stroke subtypes.
50
Journal of the Neurological Sciences 383 (2017) 47–51
Y. Liu et al.
Reference
stroke caused by LVA and CE were more likely to have a lower BP while BP of patients with ICH were prone to be extreme in the higher direction. Whilst admission BP of patients with LACI who usually had the less severe neurological damage at baseline was mostly distributed between the two extreme ends. This phenomenon may explain the “Ucurve” effect of admission blood pressure on mortality rate in the overall analysis. Various proportions of stroke subtypes among different studies may lead to conflicting conclusions. However, no effect of stroke subtype classified by OCSP has been found on the admission BP value, which was different to the previous study [24]. There were several limitations in our study. Firstly, it was a post hoc analysis of a stroke registration study in China, further clinical trials were required to confirm the effect of admission hypertension or hypotension on outcomes in patients with different stroke subtypes. Secondly, although multivariable analyses had been adjusted for established predictive factors, other potential factors we did not control might be confounding variables. For instance, NIHSS score and history of carotid stenosis had not been assessed in this study. Thirdly, the BP level was only measured at single time point and prehospital use of antihypertesion agents could be other confounding factors that has not been exactly described in the present study.
[1] J.D. Wallace, L.L. Levy, Blood pressure after stroke, JAMA 246 (1981) 2177–2180. [2] M. Britton, A. Carlsson, U. de Faire, Blood pressure course in patients with acute stroke and matched controls, Stroke 17 (1986) 861–864. [3] B. Carlberg, K. Asplund, E. Hagg, The prognostic value of admission blood pressure in patients with acute stroke, Stroke 24 (1993) 1372–1375. [4] G. Harper, C.M. Castleden, J.F. Potter, Factors affecting changes in blood pressure after acute stroke, Stroke 25 (1994) 1726–1729. [5] N. Ahmed, G. Wahlgren, High initial blood pressure after acute stroke is associated with poor functional outcome, J. Intern. Med. 249 (2001) 467–473. [6] K. Irie, T. Yamaguchi, K. Minematsu, T. Omae, The J-curve phenomenon in stroke recurrence, Stroke 24 (1993) 1844–1849. [7] J. Leonardi-Bee, P.M. Bath, S.J. Phillips, P.A. Sandercock, Blood pressure and clinical outcomes in the International Stroke Trial, Stroke 33 (2002) 1315–1320. [8] S. Tuhrim, J.M. Dambrosia, T.R. Price, J.P. Mohr, P.A. Wolf, A. Heyman, et al., Prediction of intracerebral hemorrhage survival, Ann. Neurol. 24 (1988) 258–263. [9] Y. Zhang, K.H. Reilly, W. Tong, T. Xu, J. Chen, L.A. Bazzano, et al., Blood pressure and clinical outcome among patients with acute stroke in Inner Mongolia, China, J. Hypertens. 26 (2008) 1446–1452. [10] K.N. Vemmos, G. Tsivgoulis, K. Spengos, N. Zakopoulos, A. Synetos, E. Manios, et al., U-shaped relationship between mortality and admission blood pressure in patients with acute stroke, J. Intern. Med. 255 (2004) 257–265. [11] H.P. Adams Jr., B.H. Bendixen, L.J. Kappelle, J. Biller, B.B. Love, D.L. Gordon, et al., Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment, Stroke 24 (1993) 35–41. [12] J. Bamford, P. Sandercock, M. Dennis, J. Burn, C. Warlow, Classification and natural history of clinically identifiable subtypes of cerebral infarction, Lancet (London, England). 337 (1991) 1521–1526. [13] H. Jin, S. Zhu, J.W. Wei, J. Wang, M. Liu, Y. Wu, et al., Factors associated with prehospital delays in the presentation of acute stroke in urban China, Stroke 43 (2012) 362–370. [14] E. Heeley, C.S. Anderson, Y. Huang, S. Jan, Y. Li, M. Liu, et al., Role of health insurance in averting economic hardship in families after acute stroke in China, Stroke 40 (2009) 2149–2156. [15] M. Willmot, J. Leonardi-Bee, P.M. Bath, High blood pressure in acute stroke and subsequent outcome: a systematic review, Hypertension 43 (2004) 18–24. [16] H.S. Jorgensen, H. Nakayama, H.R. Christensen, H.O. Raaschou, J.P. Kampmann, T.S. Olsen, Blood pressure in acute stroke. The Copenhagen Stroke Study, Cerebrovasc. Dis. (Basel, Switzerland). 13 (2002) 204–209. [17] D.R. Lisk, J.C. Grotta, L.M. Lamki, H.D. Tran, J.W. Taylor, D.A. Molony, et al., Should hypertension be treated after acute stroke? A randomized controlled trial using single photon emission computed tomography, Arch. Neurol. 50 (1993) 855–862. [18] A.I. Qureshi, Acute hypertensive response in patients with stroke: pathophysiology and management, Circulation 118 (2008) 176–187. [19] L. Manning, Y. Hirakawa, H. Arima, X. Wang, J. Chalmers, J. Wang, et al., Blood pressure variability and outcome after acute intracerebral haemorrhage: a post-hoc analysis of INTERACT2, a randomised controlled trial, The Lancet Neurol. 13 (2014) 364–373. [20] P.M. Bath, L. Woodhouse, P. Scutt, K. Krishnan, J.M. Wardlaw, D. Bereczki, et al., Efficacy of nitric oxide, with or without continuing antihypertensive treatment, for management of high blood pressure in acute stroke (ENOS): a partial-factorial randomised controlled trial, Lancet (London, England) 385 (2015) 617–628. [21] E.C. Sandset, P.M. Bath, G. Boysen, D. Jatuzis, J. Korv, S. Luders, et al., The angiotensin-receptor blocker candesartan for treatment of acute stroke (SCAST): a randomised, placebo-controlled, double-blind trial, Lancet (London, England) 377 (2011) 741–750. [22] J. He, Y. Zhang, T. Xu, Q. Zhao, D. Wang, C.S. Chen, et al., Effects of immediate blood pressure reduction on death and major disability in patients with acute ischemic stroke: the CATIS randomized clinical trial, JAMA 311 (2014) 479–489. [23] T.G. Robinson, J.F. Potter, G.A. Ford, C.J. Bulpitt, J. Chernova, C. Jagger, et al., Effects of antihypertensive treatment after acute stroke in the continue or Stop PostStroke Antihypertensives Collaborative Study (COSSACS): a prospective, randomised, open, blinded-endpoint trial, The Lancet Neurol. 9 (2010) 767–775. [24] A. Semplicini, A. Maresca, G. Boscolo, M. Sartori, R. Rocchi, V. Giantin, et al., Hypertension in acute ischemic stroke: a compensatory mechanism or an additional damaging factor? Arch. Intern. Med. 163 (2003) 211–216.
5. Conclusion In accordance to several previous reports, our results revealed a Ucurve relationship between admission BP and poor outcome in patients with acute stroke in the overall analysis. The U-shaped effect was remained only in patients with ICH. On the contrary, only higher admission diastolic BP levels were found significantly associated with elevated risk of death for patients with LACI. In addition, our study confirm a heterogeneity of the admission BP distribution among different subtypes of acute stroke. Further randomised clinical trials addressing the relationship between antihypertensive treatment in patients with different subtypes of stroke are required. Funding The ChinaQUEST study was supported by grants from 81170101 from the National Natural Science Foundation of China; the Macquarie Bank Foundation, The George Foundation, and AstraZeneca Pharmaceutical Chinaand was supported by theNational Natural Science Foundation of China(91639108,81370235,81770272), and by theNational High Technology Research and Development Program of China(2016YFC0903000). Disclosures There is no conflict of interest. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jns.2017.09.032.
51