Factors influencing acute blood pressure values in stroke subtypes

Transcription

1 (2004) 18, & 2004 Nature Publishing Group All rights reserved /04 $ ORIGINAL ARTICLE Factors influencing acute blood pressure values in stroke subtypes KN Vemmos 1, K Spengos 2, G Tsivgoulis 2, N Zakopoulos 1, E Manios 1, V Kotsis 1, M Daffertshofer 3 and D Vassilopoulos 2 1 Department of Clinical Therapeutics, Alexandra Hospital, University of Athens, Athens, Greece; 2 Department of Neurology, Eginition Hospital, University of Athens, Athens, Greece; 3 Department of Neurology, University of Heidelberg, Klinikum Mannheim, Mannheim, Germany The aim of this prospective observational study was to determine the association of acute blood pressure values with independent factors (demographic, clinical characteristics, early complications) in stroke subgroups of different aetiology. We evaluated data of 346 first-ever acute (o24 h) stroke patients treated in our stroke unit. Casual and 24-h blood pressure (BP) values were measured. Stroke risk factors and stroke severity on admission were documented. Strokes were divided into subgroups of different aetiopathogenic mechanism. Patients were imaged with CT-scan on admission and 5 days later to determine the presence of brain oedema and haemorrhagic transformation. The relationship of different factors to 24-h BP values (24-h BP) was evaluated separately in each stroke subgroup. In large artery atherosclerotic stroke (n ¼ 59), history of hypertension and stroke severity correlated with higher 24-h BP respectively. In cardioembolic stroke (n ¼ 87), history of hypertension, stroke severity, haemorrhagic transformation and brain oedema were associated with higher 24-h BP, while heart failure with lower 24-h BP. History of hypertension and coronary artery disease was related to higher and lower 24-h BP, respectively, in lacunar stroke (n ¼ 75). In patients with infarct of undetermined (n ¼ 57) cause 24-h BP were mainly influenced by stroke severity and history of hypertension. An independent association between higher 24-h BP and history of hypertension and cerebral oedema was documented in intracerebral haemorrhage (n ¼ 68). In conclusion, different factors influence acute BP values in stroke subtypes of different aetiology. If the clinical significance of these observations is verified, a differentiated approach in acute BP management based on stroke aetiology may be considered. (2004) 18, doi: /sj.jhh Keywords: acute stroke; blood pressure; ischaemic stroke; intracerebral haemorrhage Introduction Despite the high prevalence of a transient elevation of blood pressure (BP) values following acute stroke, 1,2 the optimal management of poststroke hypertension has not been established and remains a matter of debate. 3 5 Arguments in favour of lowering BP include reducing the formation of brain oedema, 5 7 lessening the risk of haemorrhagic transformation 5,8 and forestalling early recurrent stroke. 5,6 However, aggressive use of antihypertensive medication in acute stroke could be detrimental, because of secondary reduction of perfusion and cerebral blood flow (CBF) in the area of ischaemia, which could expand the size of the infarction. 9,10 Correspondence: Dr KN Vemmos, Acute Stroke Unit, Department of Clinical Therapeutics, Alexandra Hospital, Vasilissis Sofias 80, Athens, Greece. vemmosk@ath.forthnet.gr Received 21 July 2003; revised 9 October 2003; accepted 9 October 2003 Although recommendations for BP management in stroke have been published, the appropriate treatment of BP in the setting of acute stroke remains a controversial issue and demands further investigation. 13 In particular, interest has focused recently on the therapeutic control of BP before and during the administration of thrombolytic agents. 14,15 Stroke is a group of heterogeneous cerebrovascular diseases concerning aetiology, and BP in the acute stage may depend on stroke type, stroke severity, comorbidity and other characteristics of the individual stroke patients. A detailed knowledge of the factors influencing BP changes in patients with different stroke subtypes, reflecting different underlying vascular mechanisms, could therefore provide us with crucial data concerning BP management during the first hours after ictus, when clinicians are confronted with treatment decisions. Previous hypertension is well documented as a determinant of high BP levels in acute stroke, 1,2,16,17 but limited data are available concerning the potential impact of different demographic characteristics and stroke risk

2 254 factors (heart failure, coronary artery disease, atrial fibrillation, diabetes mellitus) on BP values in the acute stage of stroke. 6,16,17 Furthermore, it has not been determined as to which factors are associated with acute BP values in stroke subgroups of different aetiology. Noninvasive 24-h BP monitoring reduces the pressor response to hospital admission, diminishes the measurement variability and observer bias and correlates more closely with target organ damage, when compared to casually recorded BP. 18,19 Additionally, it provides high number of readings and abolishes digit preference, allowing a more precise documentation of BP values in acute stroke. Therefore, the aim of the present study was to investigate in stroke subtypes of different aetiopathogenic mechanisms the relation of BP values during the first hours of ictus to demographic characteristics, stroke risk factors, concomitant diseases, stroke severity and acute stroke complications (brain oedema and haemorrhagic transformation), by means of 24-h BP monitoring. Subjects and methods Data collection Between January 1998 and December 2001, 404 consecutive first-ever stroke patients were admitted from the emergency room in the stroke-unit of our hospital within 24 h after symptom onset. Nearly 90% of the stroke patients hospitalised in our stroke unit had used the emergency services and were therefore admitted in the acute stage of stroke to our hospital, which is a university institution providing tertiary care services to the urban population of the city of Athens. All patients were included in The Athens Stroke Registry, a prospective observational stroke data bank. Details of this study and exclusion criteria (transient ischaemic attacks, age below 18, recurrent stroke and subarachnoid haemorrhage) have been published elsewhere. 20 On admission, an internist specialised in stroke or a neurologist examined all patients, and assessed stroke severity using the Scandinavian Stroke Scale (SSS). 21 This scale evaluates the level of consciousness, eye movement, motor power in the hand, arm and leg, orientation and aphasia, facial paresis and gait. The total score ranges from 0 (minimum) to 58 (maximum). The score decreases with the severity of stroke. All subjects underwent an initial brain CT scan as well as a 12-lead electrocardiogram on admission and Doppler ultrasonography of the cervical arteries during the first 24 h of hospitalisation. A second CT- or MRI scan was performed during hospitalisation in our study population and selected patients (60%) underwent transthoracic or transoesophageal echocardiography. According to the Trial of Org in Acute Stroke Treatment (TOAST) criteria, 22 ischaemic strokes were classified based on aetiopathogenic mechanisms into the following groups: large artery atherosclerosis (atherosclerotic stroke), cardioembolic stroke, small artery occlusion (lacunar stroke) and infarct of undetermined cause. The diagnosis of intracerebral haemorrhage (ICH) was made when hyperdense lesions were identified on admission CT scan in one of five following territories: lobar, lenticulocapsular, thalamic, cerebellar or brainstem. 20 Risk factors, such as history of hypertension, atrial fibrillation, coronary artery disease, heart failure, cigarette smoking, diabetes mellitus and hypercholesterolaemia were documented. 20 Patients were diagnosed as having a history of hypertension, if they had evidence of systolic blood pressure (SBP) above 140 mmhg or diastolic blood pressure (DBP) above 90 mmhg at any time before or 4 weeks after stroke onset or if they were receiving antihypertensive medication before the index event. 23 A history of coronary heart disease (myocardial infarction, angioplasty, angina pectoris and coronary artery bypass surgery) and of heart failure was assessed by questionnaire and medical confirmation. Alert patients were always precisely questioned about the exact time of symptom onset. Their answers were double-checked by asking the patients relatives the same questions. In cases of patients with aphasic disorders and decreased level of consciousness, the exact time of onset was elicited by the same questioning of relatives in order to evaluate more accurately the elapsed time from stroke onset to hospital admission. All patients underwent an admission CT scan in order to distinguish brain infarction from intracerebral haemorrhage as the cause of stroke. A second CT scan was performed 3 5 days later with the aim of visualising the ischaemic lesion, examining oedema formation and investigating the presence of haemorrhagic infarct transformation. Cerebral oedema was defined as the presence of midline shift, sulcal effacement or ventricular compression in acute ischaemic stroke patients, 24 and as a hypodense area around the hyperdense haematoma accompanied by mass effect in patients with ICH. 25 Haemorrhagic transformation was defined as the presence of petechial haemorrhages or confluent purpuric patterns scattered throughout the infarct (haemorrhagic infarct), and as a localised homogeneous collection of blood within the region of the infarction (intrainfarct haematoma). 26 CT films were evaluated by two independent neuroradiologists, blinded to clinical details and BP measurements of the patients. BP monitoring Casual supine BP was measured by the same internist on hospital admission, in both arms at three occasions every 15 min using a standard mercury sphygmomanometer. The mean value was taken after ensuring that there was no BP difference

3 greater than 10 mmhg between the two arms. Oscillometric Space Labs equipment (Space- Labs Inc., Redmond, WA, USA) was used for 24-h BP monitoring. The cuff was fixed to the nonhemiplegic arm with tape and three BP readings were taken concomitant with sphygmomanometer readings to ensure that the average of the two sets of values did not differ by more than 5 mmhg. The device was then set to obtain automatic BP readings at 15 min intervals. BP recording was initiated within 24 h from stroke onset and during the monitoring, the patients were not receiving antihypertensive medication. A total of between 72 and 96 pairs (three or four readings per hour) of systolic and diastolic BP values (SBP and DBP, respectively) per patient per 24 h and times were fed into our statistical analysis program. The accuracy of the monitoring devices was verified monthly by obtaining 10 automatic and 10 auscultatory BP readings simultaneously from the same arm via a Y-tube. In all instances, values did not differ more than 5 mmhg. Patients with a casual BP difference greater than 10 mmhg between the two arms were excluded (n ¼ 5). Patients with missing admission casual BP values were also excluded from any further evaluation (n ¼ 8). A total of less than 72 pairs (three readings per hour) of systolic and diastolic BP values per 24 h forced the exclusion of further 22 patients. Further 12 patients were also excluded because they received antihypertensive medication during BP monitoring. A second CT scan was not performed in 11 patients, whose data were therefore not further evaluated. The study population consisted of 346 acute stroke patients fulfilling all the above-mentioned criteria. Statistical analysis Statistical analyses were performed to assess those factors that were independently associated with 24-h BP values in the acute phase of stroke. Multivariate linear regression models were used to determine independent relations to 24-h BP values. All variables of interest including stroke risk factors, demographic, clinical and radiographic characteristics were entered as covariates in the model. Separate analyses were performed for patients with different stroke subgroups. The analyses were performed using the backward procedure. Data are presented as mean 7 s.d. or as percentages and 95% confidence intervals (95% CI) are provided when indicated. A two-tailed value of Po0.05 was considered to be statistically significant. The Statistical Package for Social Science (SPSS Inc., version 10.0 for Windows) was used for statistical analyses. Results The patients population (n ¼ 346, mean age years) consisted of 229 male and 117 female persons. The mean delay from symptom onset to hospital admission was h. Ischaemic stroke was diagnosed in 278 patients (80.3%), while 68 patients (19.7%) presented with stroke symptoms due to ICH. The distribution of ischaemic stroke subtypes was as follows: atherosclerotic stroke 21.2% (n ¼ 59), cardioembolic stroke 31.3% (n ¼ 87), lacunar stroke 27.0% (n ¼ 75) and infarct of undetermined cause 20.5% (n ¼ 57). The detailed demographic, clinical, radiographic characteristics and BP recording variables of patients with different stroke subtypes are presented in Table 1. Based on the findings of the second CT scan, brain oedema and haemorrhagic transformation was identified in 106 and 34 patients, respectively. Cerebral herniation was evident in nine cases from the initial CT scan, while the presence of petechial haemorrhages throughout the infarct was visualized in two subjects on admission and was verified by the second CT scan. In the ischaemic stroke population, atrial fibrillation, heart failure, brain oedema and haemorrhagic transformation were mostly observed among patients with cardioembolic stroke, while diabetes mellitus and hypertension had a higher prevalence among patients with lacunar or atherosclerotic stroke. Subjects suffering from intracerebral haemorrhage had higher admission and 24-h systolic and diastolic blood pressure values (SBP and DBP, respectively) than patients with ischaemic stroke. A history of hypertension and brain oedema was present in nearly 80% of these patients. Multivariate linear regression models were used to determine independent relations to 24-h SBP and 24-h DBP values separately for patients with different stroke subgroups. The linear regression analyses included the following variables: age, gender, time interval between stroke onset and hospital admission, history of hypertension, smoking, hypercholesterolaemia, diabetes mellitus, atrial fibrillation, coronary artery disease, heart failure, stroke severity on admission assessed by the SSS score, cerebral oedema and haemorrhagic transformation. The regression coefficients with 95% CI for the significant variables in the regression analyses, performed separately for every stroke subgroup using the backward procedure, are shown in Table 2 (24-h SBP) and Table 3 (24-h DBP). The regression coefficient is an estimate of the specific association of a given factor with 24-h SBP or DBP. In the atherosclerotic stroke population, a onepoint increase in the SSS correlated to a 0.53 mmhg and a 0.21 mmhg decrease in 24-h SBP (95% CI: ) and 24-h DBP (95% CI: ), respectively, indicating that the patients with milder atherosclerotic stroke and higher SSS score on admission sustained lower acute BP values. In addition, acute 24-h DBP values increased by 10.1 mmhg (95% CI: ) in patients with known hypertension. In subjects suffering from cardioembolic stroke, higher BP values were independently associated 255

4 256 Table 1 Demographics, risk factors, BP recordings and clinical and radiographic characteristics of 346 first-ever acute stroke patients Variables LAA (n=59) CE (n=87) LAC (n=75) IUC (n=57) ICH (n=68) Age (years) Sex (male) Elapsed time from stroke onset (h) Hypertension Diabetes Hypercholesterolaemia Smoking Coronary artery disease Atrial fibrillation Heart failure Stroke severity (SSS) Brain oedema F Haemorrhagic infarct transformation F 17.5 F Admission BP Systolic (mmhg) Diastolic (mmhg) BP monitoring Systolic (mmhg) Diastolic (mmhg) Continuous data are presented as mean 7 s.d., noncontinuous data as percentages. LAA=large artery atherosclerosis; CE=cardioembolic stroke; LAC=lacunar stroke; IUC=infarct of undetermined cause; ICH=intracerebral haemorrhage. Table 2 Multivariate linear regression analysis of factors associated with 24-h SBP values in stroke subtypes of different aetiology Variables in the equation Linear regression coefficient 95% CI P-value Atherosclerotic stroke Stroke severity (SSS) to Cardioembolic stroke History of hypertension to 30.9 o0.001 Stroke severity (SSS) to Brain oedema to 25.8 o0.001 Lacunar stroke Coronary artery disease to History of hypertension to Infarct of undetermined cause Stroke severity (SSS) to 0.24 o0.001 History of hypertension to Intracerebral haemorrhage History of hypertension to Brain oedema to Impact on 24-h SBP is shown as increase or decrease (95% CI) for each significant variable in the multivariate linear regression model. Stroke severity assessed by the Scandinavian Stroke Scale (SSS) is given per one-point increase. with a history of hypertension and brain oedema formation. Increased stroke severity documented by decreased SSS score and concomitant heart failure correlated with higher 24-h SBP and lower 24-h DBP values, respectively. Haemorrhagic transformation of the cardioembolic infarct was independently associated with an increase in 24-h DBP by 9.6 mmhg (95% CI: ). Table 3 Multivariate linear regression analysis of factors associated with 24-h DBP pressure values in stroke subtypes of different aetiology Variables in the equation Linear regression coefficient 95% CI P-value Atherosclerotic stroke Stroke severity (SSS) to History of hypertension to Cardioembolic stroke History of hypertension to Haemorrhagic transformation to Heart failure to Brain oedema to Lacunar stroke History of hypertension to Age to Coronary artery disease to Infarct of undetermined cause Age to 2.4 o0.001 Stroke severity (SSS) 0.22 to 0.34 to History of hypertension to Intracerebral haemorrhage History of hypertension to Brain oedema to Age (years) to Impact on 24-h DBP values is shown as increase or decrease (95% CI) for each significant variable in the multivariate linear regression model. Stroke severity assessed by the Scandinavian Stroke Scale (SSS) is given per one-point increase. Age is given per 10-year increase. In lacunar stroke history of hypertension and of coronary artery disease were associated with higher and lower 24-h BP values, respectively. Moreover, a

5 10-year increase in age correlated to a 3.8 mmhg decrease in 24h-DBP (95% CI: ). In subjects with infarct of undetermined cause, 24-h BP values increased in patients with hypertension and decreased in patients with milder stroke on admission (higher SSS score). Older age was independently associated with lower 24-h DBP values. Finally, in patients with intracerebral haemorrhage, a history of hypertension and oedema formation were independently related to higher 24-h BP values. A 10-year increase in age correlated to a 4 mmhg decrease in 24-h DBP (95% CI: ). The same independent associations for every stroke subtype were revealed after performing the analyses using a forward linear regression model. Discussion To our knowledge, this is the first study to describe separately for the different stroke subtype groups the association of BP values in the acute stroke stage with stroke severity, demographic characteristics, risk factors and early stroke complications by means of 24-h BP monitoring. We provide data regarding factors influencing BP changes during the first hours of ictus, when clinicians are confronted with the question of how to restore blood supply and maximise cerebral blood flow to the ischaemic penumbra and simultaneously minimise the risk of brain haemorrhage and cerebral oedema formation. Certain studies have previously explored the impact of different factors on casual BP in the overall ischaemic 6,17 and haemorrhagic stroke population. 17 In another report, BP monitoring was used to determine the effect of stroke category on BP in a smaller number of patients, without investigating the relation of different factors to acute BP values in different stroke subtypes. 16 Our results suggest that BP values in acute stroke are associated with different factors in stroke subtypes of different pathogenic aetiology. In patients with atherosclerotic stroke, history of hypertension and stroke severity were the major determinants of BP values, while in cardioembolic stroke BP values were additionally associated with heart failure, haemorrhagic transformation and brain oedema. In lacunar stroke, hypertension and coronary artery disease correlated with higher and lower 24-h BP values, respectively. Older age was associated with decreased 24-h DBP values. In patients with infarct of undetermined cause, 24-h BP values were mainly influenced by age, stroke severity and hypertension. Finally, in the intracerebral haemorrhage population, an independent association of brain oedema and hypertension with higher BP values was documented, while age correlated with lower DBP values. The demonstrated relationship between history of hypertension and high BP values in all stroke subgroups is well supported by previous re- ports. 1,2,16,17,27,28 The high acute BP values in patients with known hypertension could be explained by insufficient antihypertensive treatment. 17 They could also be attributed to an abnormal response of chronic hypertensives to the cerebrovascular event, 2,17 since autoregulation occurs at higher BP levels in these patients due to the right shift in their BP/CBF curve. 5,9 In favour of the latter explanation is the regression of the initially elevated BP levels towards normal values in the days following stroke, which was reported in previous studies. 16,28,29 According to our results, older age correlated with lower 24-h DBP values in certain stroke subtypes. Data from the Framingham Heart Study indicate that SBP increases continuously across all age groups, whereas DBP increases until up to approximately the sixth decade of life and then begins to decrease steadily, resulting in a steep rise in pulse pressure. 30 Both changes reflect increased aortic stiffness and pulse wave velocity. 31 The correlation between increasing age and lower 24-h DBP values in different stroke subtypes of our study population (mean age 68.5 years) could be explained on the basis of these previous reports. An association was documented between heart failure or coronary artery disease and lower 24-h BP values in patients with cardioembolic or lacunar stroke, respectively. A similar relationship between low admission SBP and ischaemic heart disease has been described by other studies 6,17 and can be attributed to the poor left ventricular function and low cardiac output of patients suffering from this ominous comorbidity. 32 Stroke severity correlated with higher 24-h BP values in all ischaemic subgroups with the exception of lacunar stroke. Certain reports have addressed the association of high admission BP values with impaired consciousness 27 and clinical deterioration, 33 while others suggested that the absence of mean BP decrease during the first hours and days of stroke onset was related to stroke severity on admission 28 and clinical deterioration, respectively. 24 It should be mentioned that other investigators, who explored the association of admission BP to stroke severity challenge these results. 17,34 We assume that the severe neurological impairment on admission and the sustained high BP values during the first hours of ictus in certain stroke subtypes (CE, IUC and LAA) may reflect physiological reactions 28 and be related to subsequent oedema formation. 35 The presence of cerebral oedema in patients with CE and ICH was related to higher 24-h BP values. These findings are in accordance with previous observational studies indicating that poststroke hypertension could be associated with brain herniation in both ischaemic 6,24 and haemorrhagic stroke. 36 A possible explanation is that elevated BP values during the first hours following stroke onset cause blood brain barrier disruption, increase cerebral perfusion pressure and facilitate oedema forma- 257

6 258 tion. 4,37 Additionally, extremely high systemic blood pressure values are associated with impaired cerebral autoregulation and forced vasodilatation with leakage of fluid in the perivascular tissue, leading to cerebral oedema as well as hypertensive encephalopathy. 4,9,38 Finally, our results indicate a correlation between high 24-h DBP values in patients with cardioembolic stroke and subsequent haemorrhagic transformation. Since the diagnosis of haemorrhagic transformation was made from the CT scan performed 3-5 days after stroke onset in 32 out of 34 cases, a cause effect relationship between elevated DBP values and haemorrhagic transformation could be assumed. Both experimental 39 and observational 8 studies have stressed the role of high blood pressure in the onset of an infarct, predisposing towards haemorrhagic transformation. Furthermore, excessively high BP values were associated with parenchymal haemorrhage in patients participating in clinical trials of recombinant tissue plasminogen activator. 40,41 In contrast, in the International Stroke Trial no correlation between haemorrhagic infarcts and admission SBP was documented, 6 but this association was not investigated in different ischaemic stroke subgroups and the relationship of DBP to haemorrhagic transformation was not evaluated. Certain limitations of the present report need to be addressed. The study was conducted in a single centre and certain risk factors like smoking were defined as binary variables, although duration and severity are important. In addition, the sample sizes for each stroke subtype are relatively small, allowing a possibility of false-negative associations (Type II errors). The absence of an association between SBP values and haemorrhagic transformation in this study in contrast with other reports may thus be due to insufficient power. On the other hand, the data set has a number of strengths. Rigorous exclusion criteria were used, CT evaluation was blinded and all patients were studied during the first hours of ictus and not days later after the index event, when the acute BP changes following stroke are resolved. 5,9,13 Moreover, BP levels were documented by means of 24-h BP monitoring, which has advantages over the casual recordings and requires fewer subjects in clinical research studies. 19 To our knowledge, this is the largest reported series of patients, who underwent 24-h BP monitoring during the acute stage of stroke. In conclusion, different factors are associated with BP values during the first hours of ictus in stroke subtypes, which represent a heterogeneous group of cerebrovascular diseases concerning pathogenic mechanisms. In certain subgroups (lacunar stroke, atherosclerotic stroke, infarct of undetermined cause), stroke severity, age and comorbidity (hypertension, coronary artery disease) are the major determinants of BP values, while in others (cardioembolic stroke, intracerebral haemorrhage) high acute BP values are additionally associated with detrimental complications in the first days after stroke onset, such as cerebral oedema and haemorrhagic transformation. Altering BP levels may have different effects in patients with stroke subtypes of different pathogenic mechanisms (ischaemia vs haemorrhage, cortical vs lacunar lesion). Thus, a differentiated therapeutic approach in matters of BP regulation in patients with acute stroke of different aetiology might be considered. Should the same rules of BP management be applied to a subject with a small motor deficit due to a lacunar stroke and an unconscious patient suffering from a massive hemispheric cardioembolic infarction? Prospective clinical studies of sufficient sizes and with sufficient power to include representative patient groups of different age, stroke risk factors (hypertension, coronary heart disease, heart failure), stroke severity and especially stroke aetiology are required to clarify the issue of optimal BP management in acute stroke and can be a crucial part of the answer to this question. The 24-h BP monitoring could be a useful tool in accurately documenting acute BP changes in these studies. Acknowledgements We thank Miss Mary Batsara, Mrs Athina Peppa and Miss Angela Konstantinopoulou for their valuable help in data collection. References 1 Wallace JD, Levy LL. Blood pressure after stroke. JAMA 1981; 246: Britton M, Carlsson A, de Faire U. Blood pressure course in patients with acute stroke and matched controls. Stroke 1986; 17: Yatsou FM, Zivin J. Hypertension in acute ischaemic strokes: not to treat. Arch Neurol 1985; 42: Spence JD, Del Maestro RF. Hypertension in acute ischaemic strokes: treat. Arch Neurol 1985; 42: Bath FJ, Bath PMW. What is the correct management of blood pressure in acute stroke? The blood pressure in Acute Stroke Collaboration. Cerebrovasc Dis 1997; 7: Leonardi-Bee J, Bath PM, Phillips SJ, Sandercock PA. Blood pressure and clinical outcomes in the International Stoke Trial. Stroke 2002; 33: Lees KR, Dyker AG. Blood pressure control after acute stroke. J Hypertens Suppl 1996; 14: S35 S38. 8 Ogata J et al. Hemorrhagic infarct of the brain without a reopening of the occluded arteries in cardioembolic stroke. Stroke 1989; 20: Powers WJ. Acute hypertension after acute stroke: the scientific basis for treatment decisions. Neurology 1993; 43: Fischberg GM et al. Stroke precipitated by moderate blood pressure reduction. J Emerg Med 2000; 19:

7 11 Hacke W et al. Acute treatment of ischemic stroke. European Stroke Initiative (EUSI). Cerebrovasc Dis 2000; 10(Suppl 3): Adams Jr HP et al. Stroke Council of the American Stroke Association. Guidelines for the early management of patients with ischemic stroke: a scientific statement from the Stroke Council of the American Stroke Association. Stroke 2003; 34: Bath P et al. International Society of Hypertension Writing Group. International Society of Hypertension (ISH): statement on the management of blood pressure in acute stroke. J Hypertens 2003; 21: Bath P. Alteplase not yet proven for acute ischaemic stroke. Lancet 1998; 352: Slyter H. Ethical challenges in stroke research. Stroke 1998; 29: Morfis L, Schwartz RS, Poulos R, Howes LG. Blood pressure changes in acute cerebral infarction and hemorrhage. Stroke 1997; 28: Jorgensen HS et al. Blood pressure in acute stroke. The Copenhagen Stroke Study. Cerebrovasc Dis 2002; 13: Harper G et al. The changes in blood pressure after acute stroke: abolishing the white coat effect with 24-h ambulatory monitoring. J Intern Med 1994; 235: Coats AJ. Reproducibility or variability of casual and ambulatory blood pressure data: implications for clinical trials. J Hypertens Suppl 1990; 8: S17 S Vemmos K et al. The Athens stroke registry: results of a five-year hospital-based study. Cerebrovasc Dis 2000; 10: Lindestrom E et al. Reliability of Scandinavian Neurological Stroke Scale. Cerebrovasc Dis 1991; 1: Adams Jr HP et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org in Acute Stroke Treatment. Stroke 1993; 24: World Health Organization-International Society of Hypertension Guidelines for the Management of Hypertension. Guidelines Subcommittee. J Hypertens 1999; 17: Chamorro A et al. Blood pressure and functional recovery in acute ischemic stroke. Stroke 1998; 29: Dul K, Drayer B. CT and MR imaging of intracerebral hemorrhage. In: Kase CS, Caplan LR (eds). Intracerebral Hemorrhage. Butterworth-Heinemann: Boston: 1994, pp Moulin T, Crepin-Leblond T, Chopard JL, Bogousslavsky J. Hemorrhagic infarcts. Eur Neurol 1994; 34: Britton M, Carlsson A. Very high blood pressure in acute stroke. J Intern Med 1990; 228: Christensen H, Meden P, Overgaard K, Boysen G. The course of blood pressure in acute stroke is related to the severity of the neurological deficits. Acta Neurol Scand 2002; 106: Harper G, Castleden CM, Potter JF. Factors affecting changes in blood pressure after acute stroke. Stroke 1994; 25: Franklin SS et al. Hemodynamic patterns of agerelated changes in blood pressure. The Framingham Heart Study. Circulation 1997; 96: Glasser SP. On arterial physiology, pathophysiology of vascular compliance, and cardiovascular disease. Heart Dis 2000; 2: Staessen JA. Potential adverse effects of blood pressure loweringfj-curve revisited. Lancet 1996; 348: Davalos A et al. Deteriorating ischemic stroke: risk factors and prognosis. Neurology 1990; 40: Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Effect of blood pressure and diabetes on stroke in progression. Lancet 1994; 344: Krieger DW et al. Early clinical and radiological predictors of fatal brain swelling in ischemic stroke. Stroke 1999; 30: Broderick JP, Brott TG, Tomsick T, Huster G. Ultra early evaluation of intracerebral haemorrhage. J Neurosurg 1990; 72: Hatashita S, Hoff JT, Ishii S. Focal brain oedema associated with acute arterial hypertension. J Neurosurg 1986; 64: Straandgaard S, Paulson OB. Cerebral blood flow and its pathophysiology in hypertension. Am J Hypertens 1989; 2: Saku Y et al. Hemorrhagic infarct induced by arterial hypertension in cat brain following middle cerebral artery occlusion. Stroke 1990; 21: The NINDS t-pa Stroke Study Group. Intracerebral hemorrhage after intravenous t-pa therapy for ischemic stroke. Stroke 1997; 28: Larrue V, von Kummer R, del Zoppo G, Bluhmki E. Hemorrhagic transformation in acute ischemic stroke. Potential contributing factors in the European Cooperative Acute Stroke Study. Stroke 1997; 28: