Stroke

Poster Session II Thursday 31 May 2007

Cerebrovasc Dis 2007;23(suppl 2):1–147

Risk factors of stroke

1 Risk factors of stroke

LUNG FUNCTION AND LONG-TERM FATAL STROKEU. Goldbourt, D. TanneTel Aviv University Department of Epidemiolgy and Preventive Medicine, TelAviv, Israel

Background: Research on lung function and incidence of stroke has yieldedconflicting results. Participants and methods: For 10,232 men of diverse countriesof origin, civil servants and municipal employees, who participated in the IsraeliIschemic Heart Disease (IHD) study in 1963, extensive demographic, biochemical,socioeconomic and clinical information was collected in 1963, 1965 and 1968. Ofthese men, 4330 underwent an assessment of forced vital capacity (VC) and 1.0 secforced expiratory volume (FEV).Results: Over a 23-yr follow up, 1297 men died, among whom for 136 therecorded underlying cause of death was stroke. For quartiles of VC, as % of ageand height predicted level, there was a modest gradient of fatal stroke (18,16,14and 13 per 10,000 person-years) which was erased upon age-adjustment. Thecorresponding rates declined from to 23 to 18, 10 and 8 per 10,000 person yearsfor FEV, or 19,15,14 and 10 for FEV as percent of FVC, in the 1st,2nd,3rd, and4th quartiles, respectively (P for trend=0.02 for the former and 0.55 for the latterby Mantel-Cox trend test after age-adjustment and exclusion of baseline IHD andcancer). Further adjustment for height which was markedly, inversely related tostroke mortality eliminated both associations. There was no appreciable interactionbetween smoking habits and lung function with respect to long-term fatal stoke.Conclusions: In this cohort, low FEV, whether in absolute terms or in relation toFVC, did not appear to be an independent marker of risk for fatal stroke beyondage and height.


CLOPIDOGREL RESISTANCE: ROLE OF BODY MASS AND CONCOMITANTMEDICATIONSG. Feher, K. Koltai, B. Alkonyi, L. Szapary, G. Kesmarky, S. Komoly, K. TothUniversity of Pecs, Medical School, Pecs, Hungary

Introduction: Platelets have a central role in the development of arterial thrombosisand subsequent cardiovascular events. An appreciation of this has made antiplatelettherapy the cornerstone of cardiovascular disease management. Recent studies havedescribed the phenomenon of clopidogrel resistance but the possible mechanismsare still unclear.Patients and methods: The aim of this study was to compare the characteristics(risk profile, previous diseases, medications, hemorheological variables and plasmavon Willebrand factor and soluble P-selectin levels) of patients in whom clopi-dogrel provided effective platelet inhibition with those in whom clopidogrel wasnot effective in providing platelet inhibition. 157 patients with chronic cardio- andcerebrovascular diseases (83 males, mean age 61±11 yrs, 74 females, 63±13 yrs)taking 75 mg clopidogrel daily (not combined with aspirin) were included in thestudy.Results: Compared with clopidogrel-resistant patients (35 patients (22%), patientswho demonstrated effective clopidogrel inhibition had a significantly lower BMI(26.1 vs. 28.8 kg/m2, p<0.05). Patients with ineffective platelet aggregation weresignificantly more likely to be taking benzodiazepines (25% vs. 10%) and selectiveserotonin reuptake inhibitors (28% vs. 12%) (p<0.05). After an adjustment tothe risk factors and medications BMI (OR 2.62; 95% CI: 1.71 to 3.6; p<0.01),benzodiazepines (OR 5.83; 95% CI: 2.53 to 7.1; p<0.05) and SSRIs (OR 5.22; 95%CI: 2.46 to 6.83; p<0.05) remained independently associated with CLP resistance.There was no significant difference in the rheological parameters and in the plasmalevels of adhesive molecules between the two examined groups.Conclusion: The background of ineffective clopidogrel medication is complex.Drug interactions may play a role on clopidogrel bioavailability, on the other hand,

the significant difference in BMI between the two examined groups suggests thatclopidogrel therapy should be weight-adjusted.


ASSOCIATION OF METABOLIC SYNDROME WITH ISCHEMIC STROKE INPATIENTS WITH INTRACRANIAL ATHEROSCLEROSISJ.H. ParkMyongji Hospital, Kwandong University, College of Medicine, Goyang-si,South Korea

Background and purpose: Metabolic syndrome (MetS) is associated with intracra-nial atherosclerosis. Patients with more severe MetS components were reportedto be more likely to have intracranial atherosclerosis. To elucidate the associationbetween MetS and ischemic stroke, we attempted to demonstrate the association ofMetS and its individual components with frequency of ischemic stroke lesions andinvestigated the independent associations between them in acute ischemic strokepatients.Methods: We evaluated 370 acute ischemic stroke patients who underwent brainmagnetic resonance (MR) imaging and MR angiography. The stroke subgroupswere categorized as intracranial large artery atherosclerosis (IC-LAA, n=151),extracranial large artery atherosclerosis (EC-LAA, n=29), and nonatherosclerosis(NA, n=190). MetS was defined using the criteria of the National CholesterolEducation Program Adult Treatment Panel III.Results: Patients with IC-LAA group showed a higher rate of previous ischemiclesions and MetS than those with EC-LAA and NA (all P<0.001). The number ofprevious ischemic lesions showed a tendency to increase as the number of MetScomponents increased in the IC-LAA group (P=0.002). In the IC-LAA group, MetSwas independently associated with previous ischemic lesions (OR, 3.80 P<0.001)which was prominent with more severe MetS components after adjustment for riskfactors (P<0.001). Among the component conditions, high blood pressure, im-paired fasting glucose, and abdominal obesity were predominantly associated withprevious ischemic lesions (all P<0.001). Conclusions: MetS was associated withischemic stroke with IC-LAA. Controlling the MetS components is mandatory withthe aim of preventing from advanced intracranial atherosclerotic vascular damageand ischemic stroke. Further studies of different ethnics need to be performed toconfirm whether MetS is more associated with those with IC-LAA.


DEATH AND DEPENDENCE ONE YEAR AFTER THE FIRST TRANSIENTISCHAEMIC ATTACK: A POPULATION-BASED STUDY IN RURAL ANDURBAN NORTHERN PORTUGALM. Correia, M.R. Silva, E. Moreira, R. Magalhães, M.C. SilvaInstitute for Molecular and Cell Biology, University of Porto, Porto, Portugal

Background and purpose: Information about prognosis of transient ischaemic at-tacks (TIA) is scarce, particularly in population-based studies. Recent data showeda high early risk of stroke after a TIA, reaching 12.7% at seven days. In this studyprognosis of a first TIA is evaluated in terms of risk of death and dependence,taking into account the occurrence of stroke after the index event.Methods: The 141 patients with a first–ever-in-a-lifetime TIA occurred betweenOctober 1998 and September 2000 in a rural population of 18677 and an urbanpopulation of 86023 were entered a registry. A neurologist observed these patientssoon after the episode and also at three and twelve months after the TIA. Previousdependence and dependence after the episode was determined by the modifiedRankin scale (score 3 or more).Results: One year after the TIA, 10 out of 105 patients (9.5%) in the urban areawere death, compared to 6 out of 36 (16.7%) in the rural area. Amongst thesurvivors 23.7% were dependent one year after the episode, 25.8% in the urbanarea and 17.2% in rural area (excluding seven that were lost to follow-up). In theurban area this proportion reduces to 22.8% taking into account those previouslydependent. Comparing the Rankin score before and one year after the episode, 77

Poster SessionRisk factors of stroke

Cerebrovasc Dis 2007;23(suppl 2):1–147 81

(57.5%) become more dependent or died, 53 (39.6%) kept their score and onlyfour patients improved their score (Wilcoxon test, z=7.5, p<0,001). As expectedthe occurrence of stroke during the follow-up period worsened the prognosis at oneyear (Mann-Whitney, z=2.7, p=0.007).Discussion: The occurrence of a first TIA carries out not only an early risk ofstroke, but most of the patients showed an increased degree of dependence one yearafter the episode. The occurrence of stroke after a TIA is partially responsible forthis prognosis.Study supported by: FCT/FEDER project POCI/SAU-ESP/59885/2004


AGE PREDICTS DELAYED TRANSIENT ISCHAEMIC ATTACKINVESTIGATION AND MANAGEMENTT.J. Quinn, J. Dawson, K.R. Lees, M.R. WaltersGardiner Institute of Cardiovascular and Medical Sciences, University ofGlasgow, Glasgow, United Kingdom

Introduction: Cerebrovascular events increase in frequency and severity with age.Recent studies suggest that elderly UK patients diagnosed with transient ischaemicattack (TIA) are inappropriately denied evidence based intervention. Similar ex-amples of “ageism” have been reported in many other areas of hospital medicine.Due to the frequency of events immediately following TIA, urgent assessment andinitiation of treatment is essential. We examined if age influenced referral to a“fast-track” TIA clinic.Methods: Our TIA clinics assess all suspected cerebrovascular events referred.Patient details are prospectively recorded in a comprehensive database. Data werecollated for patients seen between August 1992 and January 2005. Patients werecategorised according to age: 0-40; 41-65; 66-75; 76+. Associations between ageand: mode of referral (letter; phone call; other); time to referral; time from referralto appointment and initiation of treatment prior to clinic were analysed using theKruskal-Wallis test.Results: Full data were available for 3495 of 3596 patients assessed during thestudy period, median age 67 (range 16-95). There was no significant associationbetween increasing age and delay from symptom onset to referral time (P=0.014);mode of referral (P=0.131) or time to appointment (P=0.652). The youngest patientswere less likely to be prescribed antiplatelet (P<0.001).

Age (years) Median Referral Median Appointment Number of Number onTime (days) Time (days) Phoned Referral Anti-platelet

0–40 (n=188) 6 8 28 (14.9%) 35 (18.6%)41–65 (n=1431) 8 9 191 (13.3%) 511 (35.7%)66–75 (n=1035) 7 9 164 (15.8) 337 (32.6%)Over 75 (n=805) 6 9 120 (14.9) 271 (33.7%)

Discussion: We have found no evidence of ageism in access to TIA services.However, delays to clinic assessment remain substantial. Public education strategiesto promote early presentation with TIA symptoms are required.


SOLUBLE RECEPTOR ACTIVATOR OF NUCLEAR FACTOR-KB LIGAND(RANKL) AND RISK FOR CARDIOVASCULAR DISEASES. Kiechl, G. Schett, J. Schwaiger, K. Seppi, P. Eder, G. Egger, P. Santer,A. Mayr, Q. Xu, J. WilleitInnsbruck Medical University, Innsbruck, Innsbruck, Austria

Background: Overexpression of RANKL is a prominent feature of vulnerableatherosclerotic lesions prone to rupture and was suggested to contribute to thetransition from a stable to an unstable plaque phenotype in both human andmurine atherosclerosis because of its ability to promote matrix degradation,monocyte/macrophage chemotaxsis and vascular calcification.Methods and results: The Bruneck Study is a prospective population-based surveyof men and women 40-79 years old at the 1990 baseline examination. Levels ofsoluble RANKL and other variables were assessed in 909 subjects (1990) andup-dated every five years. All cases of cardiovascular disease (CVD) were carefullyrecorded between 1990 and 2005.During follow-up, CVD (defined as ischemic stroke and TIA, myocardial infarctionand vascular death) manifested in 124 subjects. The level of soluble RANKLemerged as a highly significant predictor of vascular risk (adjusted hazard ratio[95%CI] 1.27 [1.16-1.40]; P<0.001). Predictive significance was independent ofthat afforded by classic vascular risk factors, C-reactive protein and osteoprotegerinconcentration and severity of carotid atherosclerosis. Findings were internally

consistent and robust in a variety of sensitivity analyses. Notably, soluble RANKLwas not associated with carotid or femoral artery atherosclerosis assessed andmonitored by high-resolution ultrasound.Conclusions: Our study lends large-scale epidemiological support to a role ofRANKL in CVD. In the absence of a significant association between RANKL andatherosclerosis the view that RANKL promotes plaque destabilization and ruptureis a highly appealing concept.


INCREASED PREVALENCE OF VASCULAR RISK FACTORS BUT EQUALACCESS TO HOSPITAL SERVICES IN A DEPRIVED TRANSIENTISCHAEMIC ATTACK POPULATIONT.J. Quinn, J. Dawson, K.R. Lees, M.R. WaltersGardiner Institute of Cardiovascular and Medical Sciences, University ofGlasgow, Glasgow, United Kingdom

Introduction: Cerebrovascular disease is over represented in socially deprivedcommunities but traditional vascular risk factors do not account for all the vari-ation. Unequal access to and uptake of specialist management could contribute.We analysed data on TIA clinic attendance, seeking effects of socioeconomicdeprivation.Methods: We serve an urban population including the most affluent and deprivedareas of the city. We prospectively record structured details of patients seen at clinic.Socio-economic deprivation was measured from postcodes and the 2001 census,using the Depcat ordinal hierarchical scale, ranging from 1=least deprived to7=most deprived. We analysed the association between social deprivation, referraland vascular risk factors by ANOVA. We interpreted delay from symptom onset toclinic referral as a marker of access to specialist care.Results: We assessed 3462 patients between August 1992 to January 2005. Themost deprived areas were over represented, with 1711 patients in Depcat 6 - 7.There was no association between clinic referral or attendance and Depcat. Therewere significant associations between Depcat and “lifestyle” factors of smokingand alcohol excess (each p<0.005).

Depcat1 Depcat2 Depcat3 Depcat4 Depcat5 Depcat6 Depcat7

Delay to clinic (days) 16.5 35.2 28.9 33.9 37.5 89.6 55.3Smoking % 6.1 13.9 24.7 34.6 24.8 40.1 46.3Hypertension % 39 27.9 44.1 42.3 46.3 44.4 44.1Diabetes % 5.1 3.7 7.8 18.6 7.5 9.3 7.7Dyslipidemia % 16.9 8.9 16.3 18.7 14.2 19 17.9Alcohol % 2.3 1.2 4.8 3.7 3.8 4.3 5.2

Discussion: The effect of deprivation on cerebrovascular health inequality is partlyexplained by certain vascular risk factors but not by access to hospital services.


DO CONVENTIONAL VASCULAR RISK FACTORS INFLUENCE BRAINARTERIOVENOUS MALFORMATIONS? PROSPECTIVE,POPULATION-BASED COHORT AND CASE-CONTROL STUDIEST.M. Brock, R. Al-Shahi SalmanDivision of Clinical Neurosciences, University of Edinburgh, Edinburgh, UnitedKingdom

Background: Conventional risk factors for cerebral infarction and haemorrhageare thought to play little – if any – role in either causing brain arteriovenousmalformations (AVMs), or influencing their behaviour.Methods: We extracted data on vascular risk factors from the medical recordsof 229 adults newly-diagnosed with a brain AVM, who were enrolled between1999-2003 in a prospective, population-based cohort study in Scotland (SIVMS).A sample of 36 adults with brain AVMs were sex-matched and age-matched (towithin 5 years) with 36 controls.Results: In the case-control study, adults with brain AVMs were more likely thancontrols to have smoked at some stage in their life (75% versus 50%, p=0.028;odds ratio [OR] 3.0, 95% confidence interval [CI] 1.1 to 8.1), but this relationshipdid not hold for current smokers, nor was it found for hypertension, ischaemicheart disease, alcohol consumption, hyperlipidaemia, or diabetes mellitus. In thewhole cohort, 116 (72%) had smoked sometime in their life, and 29 (14%) hadhypertension prior to the first presentation of their brain AVM. There was nosignificant difference between adults who did (n=114) and did not (n=115) presentwith intracranial haemorrhage in pre-presentation hypertension, smoking, or anyother vascular risk factors.

82 Cerebrovasc Dis 2007;23(suppl 2):1–147 16th European Stroke Conference

Discussion: Smoking appears to be associated with the occurrence of brain AVMs,but this finding needs confirmation in other cohorts, and larger studies. Vascularrisk factors do not appear to influence whether a brain AVM first presents witha haemorrhage or not, but their influence on the occurrence of haemorrhage afterAVM diagnosis requires further investigation.


NO INCREASED RISK OF RECURRENT STROKE IN PFO CARRIERSH. Poppert, A. Bockelbrink, M. Morschhaeuser, J. Schwarze, P. Heider,L. Esposito, D. SanderMunich University of Technique, Munich, Germany

Background: Contrast-enhanced transcranial Doppler ultrasonography (cTCD) isa sensitive noninvasive screening method for detection of a patent foramen ovale(PFO). We aimed to investigate the relationship between a suchlike detectedright-to-left shunt (RLS) and subtypes of cerebral ischemia as well as the risk ofstroke recurrence.Methods: The records of 763 patients with definite diagnosis of cerebral ischemiaat discharge were analyzed retrospectively. All patients had undergone TCD basedRLS detection. Stroke origin was subtyped using the TOAST classification criteria.For follow-up all patients were contacted by mail. In case a patient did not answer,we tried to contact the patient or the patients relatives and the family doctor bytelephone.Results: A RLS was detected in 140 (28%) male and in 114 (42%) female patients.These patients were younger (p<0.001) and in male patients presence of RLSwas associated with stroke of unknown origin (p=0.001). In female patients thisassociation was not significant (p=0.076). After adjustment for age no significantassociation was found in either group.Complete follow-up data with a median follow-up period of 4 years could becollected in 639 patients (83.7%). 10 shunt-carriers (4.7%) and 32 patients (7.6%)without RLS (p=0.180) had suffered a recurrent stroke. Logistic regression adjust-ing for age, gender and stroke subtype confirmed the lack of a positive correlation(OR 0.7 (95%CI 0.33-1.48)).Conclusion: We found age and gender to be important confounders in the oftencited association of PFO and cryptogenic stroke. This has not been taken intoaccount in most previous studies. Furthermore, RLS did not correlate with strokerecurrence, thus weakening the thesis of a PFO generally being an important riskfactor for stroke.


TYPE 3 PHOSPHODIESTERASE INHIBITORS MAY BE PROTECTIVEAGAINST CEREBROVASCULAR EVENTS IN PATIENTS WITHCLAUDICATIONW.M. Stone, S.R. Money, R.J. FowlMayo Clinic Arizona, Phoenix, AR, USA

Objective: The risk of cerebrovascular events in patients with mild to moderateperipheral vascular disease is significant. Cilostazol is a phosphodiesterase type 3(PDE3) inhibitor that is effective in the treatment of symptoms of peripheral arterialocclusive disease. The method of action includes antithrombotic, vasodilatory, andantiproliferative effects.Methods: The CASTLE trial was a prospective randomized double blinded trialto establish the safety of this PDE3 inhibitor use in 1435 patients with mildto moderate peripheral arterial occlusive disease. A post-hoc analysis of theCASTLE trial was undertaken to evaluate Cilostazol usage on cerebrovascularevents. Blinded adjudication of all cerebrovascular events (stroke, TIA, and carotidrevascularization) in this trial was performed. Kaplan Meier analysis was used forstatistical evaluation.Results: The overall rate of cerebrovascular events was 4.6% (66 of 1435 patients)with a mean followup of 515 days. Ischemic vascular events were more common(2.5%) than hemorrhagic events (0.3%), (p<0.05). The placebo group demonstrateda greater risk for events, 5.8% (42 of 718 patients) vs. the Cilostazol treated group,2.9% (21 of 717 patients), (p<0.05). Cerebrovascular risk factors were similar inboth groups.Conclusion: The risk of cerebrovascular events in patients with mild to moderateperipheral arterial occlusive disease is 4.6% with a mean followup of 515 days.Treatment with PDE3 inhibitors may reduce this risk. Further evaluation of the useof PDE3 inhibitors for prevention of cerebrovascular events should be considered.


DEPRESSIVE SYMPTOMS AND RISK OF STROKE: THE ROTTERDAMSTUDYM.J. Bos, T. Lindén, P.J. Koudstaal, A. Hofman, I. Skoog, H.W. Tiemeier,M.M. BretelerErasmus Medical Center, Rotterdam, Rotterdam, The Netherlands

Background: Results from previous studies that assessed whether self-reporteddepressive symptoms predispose to stroke in the general elderly population arecontroversial and they did not distinguish between men and women, nor did theyperform psychiatric workups in those who reported depressive symptoms. Weexamined the association between depressive symptoms, depressive disorder, andrisk of stroke in the general population.Methods: This prospective population-based cohort study was based on 4424participants of the third Rotterdam Study survey (1997-1999), who at that timewere ≥61 years of age, free from stroke, and underwent Center for EpidemiologicalStudies Depression Scale (CESD) interview. Depressive symptoms were consideredpresent if CESD score was ≥16. Participants with depressive symptoms underwentdiagnostic workup for depressive disorder. Follow-up for incident stroke wascomplete until January 1, 2005. Data were analyzed with Cox proportional hazardsmodels with adjustment for relevant confounders.Results: Men with depressive symptoms were at increased risk of stroke (adjustedhazard ratio (HR) 2.15; 95% confidence interval (CI) 1.10-4.22) and ischemicstroke (adjusted HR 3.25; 95% CI 1.62-6.50). In women there was no associationbetween presence of depressive symptoms and risk of stroke. The associationsthat we found were at least partly attributable to persons who reported depressivesymptoms but who did not fulfil DSM-IV diagnostic criteria for depressive disorder.Discussion: Presence of depressive symptoms is a strong risk factor for stroke inmen but not in women.


ASPIRIN RESISTANCE: POSSIBLE ROLES OF CARDIOVASCULAR RISKFACTORS, PREVIOUS DISEASE HISTORY, CONCOMITANT MEDICATIONSAND HAEMORRHEOLOGICAL VARIABLESG. Feher, K. Koltai, B. Alkonyi, L. Szapary, G. Kesmarky, S. Komoly, K. TothUniversity of Pecs, Medical School, Pecs, Pécs, Hungary

Introduction: The aim of this study was to compare the characteristics (riskprofile, previous diseases, medications and haemorrheological variables) of patientsin whom aspirin provided effective platelet inhibition with those in whom aspirinwas not effective in providing platelet inhibition.Patients and methods: 599 patients with chronic cardio- and cerebrovasculardiseases (355 men, mean age 64 ± 11 years; 244 women, mean age 63 ± 10years) taking aspirin 100-325 mg/day were included in the study. Blood wascollected between 8:00am and 9:00am from these patients after an overnight fast.The cardiovascular risk profiles, history of previous diseases, medication historyand haemorrheological parameters of patients who responded to aspirin and thosewho did not were compared. Platelet and red blood cell (RBC) aggregation weremeasured by aggregometry, haematocrit by a microhaematocrit centrifuge, andplasma fibrinogen by Clauss’ method. Plasma and whole blood viscosities weremeasured using a capillary viscosimeter.Results: Compared with aspirin-resistant patients, aspirin sensitive patients hada significantly lower plasma fibrinogen level (3.3 g/L vs 3.8 g/L; p < 0.05) andRBC aggregation values (24.3 vs 28.2; p < 0.01). In addition, significantly morepatients with effective aspirin inhibition were hypertensive (80% vs 62%; p <

0.05). Patients who had effective platelet aggregation were significantly more likelyto be taking beta-adrenoceptor antagonists (75% vs 55%; p < 0.05) and ACEinhibitors (70% vs 50%; p < 0.05), patients with ineffective platelet aggregationwere significantly more likely to be taking HMG-CoA reductase inhibitors (statins)[52% vs 38%; p < 0.05]. Use of statins remained an independent predictor ofaspirin resistance even after adjustment for risk factors and medication use (oddsratio 5.92; 95% CI 1.83, 16.9; p < 0.001).Conclusions: Impaired hemorheological parameters are associated with aspirinresistance. It is also possible that drug interactions with statins might reduce aspirinbioavailability.




PSYCHO-SOCIAL FACTORS INFLUENECE STROKE OUTCOMEG. Szabó, L. Molnár, C. Óváry, Z. NagyNational Stroke Centre/Budapest, Balassagyarmat, Hungary

Background: Factors influencing stroke outcome are subject of interest. The aimof present study was to compare the impact of “traditional” risks factors andpsychosocial factors on modified Rankin scale.Methods: In the stroke unite of a county hospital, 646 consecutive stroke patientswere include in this study. Determining disease outcome we used the modifiedRankin scale after 28 days of stroke onset. Stroke risk factors were recorded andpsychosocial data (Beck depression scale, coping skills, educational attainment,and isolation) were collected using the standardized questionnaires. For statisticswe used a multiway contingency tables with log linear modelling.Results: A log linear model fits well (p=0.786) to our data. Patients suffering formrepeated stroke showed significantly worse outcome (p<0.001). Hypertension,diabetes mellitus, hyperlipidemia and smoking were not significant factors onstroke outcome. On the other hand close relationship could be described in the3-factors interaction among depression, isolation and stroke outcome, furthermoreamong coping skills, depression, and stroke outcome, finally among coping skills,educational attainment, and stroke outcome. The 2-factors interaction betweeneducational attainment and isolation, and between educational attainment anddepression could be also proved.Discussion: The significant effects of depression, insufficient coping skills, low ed-ucational level, and isolation on stroke outcome were documented. Better managingof these factors may improve the quality of life after stroke.


STROKE IN CHRONIC KIDNEY DISEASE: PATTERNS OF STROKE,FACTORS AFFECTING DEVELOPMENT OF STROKES. Jung, S.H. Hwang, S.B. Kwon, K.H. Kwon, I.S. Koh, K.H. Yu, B.C. LeeHallym University College of Medicine, South Korea

Background: Patients with chronic kidney disease (CKD) have increased risk forstroke. Although they share conventional stroke risk factors, not all CKD patientssuffer stroke. We analyzed the pattern of stroke and tried to elucidate the factorsaffecting development of stroke in patients with CKD.Method: We used data from Hallym Stroke Registry (HSR) for enrolling patientswith stroke and chronic renal failure (CRF) or end stage renal disease (ESRD) asCKD. Forty patients were included from total 55 patients with acute stroke within7 days of onset who were diagnosed as CKD from July 1996 through June 2005.35 patients with CKD who did not develop stroke were selected as the controlgroup. We classified patients according to the modified TOAST classificationand evaluated their neurologic manifestations. We also compared the propertiesincluding stroke risk factors and laboratory findings between two groups.Results: 29 patients had ischemic stroke and 11 patients had hemorrhagic stroke.The percentage of hemorrhagic stroke was higher than general population in HSR(5.3%). The most common ischemic stroke subtype was small artery occlusion(25.0%) and followed by large atherosclerosis (20.0%). The hypertensive nephropa-thy was most common cause of CKD (47.5%) and duration of causative diseasewas 13.19±10.24 years. As compared to the control group, stroke patients withCKD showed higher rate of hypertension, prolonged activated prothrombin time,lower level of triglyceride and lower LDL cholesterol level.Conclusion: Strokes in CKD differ from general population in their pattern andthe risk factors. There is relatively larger incidence of hemorrhages compared tonon-kidney disease patients. Low lipid profile in CKD patients may suggest theprotective effect of hyperlipidemia against the development of stoke as preventingof malnutrition which is one of the most important cause of mortalities in CKDpatients.


TRANSCRANIAL DOPPLER HEMODYNAMIC PARAMETERS AND RISK OFSTROKE: THE ROTTERDAM STUDYM.J. Bos, P.J. Koudstaal, A. Hofman, J.C. Witteman, M.M. BretelerErasmus Medical Center, Rotterdam, Rotterdam, The Netherlands

Background: We explored the association between transcranial Doppler (TCD)hemodynamic indices and risk of stroke in the general population.Methods: At baseline we assessed mean flow velocity, peak systolic flow velocity,end diastolic flow velocity, and vasomotor reactivity (VMR) with TCD in 2022 Rot-terdam Study participants of age 61 years and over in both middle cerebral arteries.

All participants, who at baseline were free from previous stroke, were subsequentlyfollowed for occurrence of stroke (average follow-up time 5.1 years). We calculatedhazard ratios (HRs) with 95% confidence intervals (CIs) for the association betweenhemodynamic parameters and risk of stroke using Cox proportional hazards modelswith adjustment for age, sex, systolic blood pressure, antihypertensive drug use,diabetes mellitus, ever smoking, current smoking, carotid intima-media thickness,and carotid distensibility.Results: Risk of stroke (n=122) and ischemic stroke (n=89) increased with increas-ing middle cerebral artery flow velocity: when comparing the tertile with highestvelocity to the tertile with lowest velocity, the HR was 1.74 (95% CI 1.09-2.77) forthe association between mean flow velocity and stroke, 1.63 (95% CI 1.03-2.58)for end diastolic flow velocity and stroke, and 1.33 (95% CI 0.86-2.08) for peaksystolic flow velocity and stroke. These estimates increased 10-26% when onlyischemic strokes were included. We found no associations between vasomotorreactivity and risk of stroke.Discussion: Risk of stroke increased strongly with increasing middle cerebralartery flow velocity as measured with TCD in the general population.


A PROSPECTIVE STUDY OF PREVALENCE AND PROGNOSIS OF PVD INPATIENTS ADMITTED WITH ISCHAEMIC STROKEV. Paranna, A. Gupta, R. Prakash, L. Bachegowda, G. Shankar, K.K. Manda,A. WatkinsWest Wales Hospital, United Kingdom

Introduction: Atherosclerosis is a multisystemic,asymtomatic disease which is di-vided into Cerebrovascular disease,Peripheral Vascular disease & Coronary Arterydisease.Patients presenting with acute stroke are likely to have PVD which may beunrecognised.Presence of PVD increases the risk of Myocardial Infarction & strokeby 2 fold.Early detection & management of PVD can avoid complications.AnkleBrachial Pressure Index is simple bedside test useful in this process.Objective: The aim of the study was to evaluate the prevalence of PVD instroke patients,to assess whether symptomatic or asymptomatic PVD,severity ofthe disease & to evaluate whether PVD is an independent prognostic factor in theoutcome of stroke.Method: Prospective comparative group study which was carried out in WestWales Hospital,Carmarthen,UK.A total of 100 patients have been included in thisongoing study.This includes 55study group and 45 control group.Patients admittedwith ischaemic stroke were included in the study group & age and sex matchednon stroke general medical acute admission were included in control group.Eticalcommittee approval obtained.Intervention:The data collected included demographics,barthel index,abbreviatedmental test,associated vascular risk factors,whether symptomatic PVD in thepast,ABPI during hospitalisation,discharge destination & follow up 3months afterdischarge.Ankle BP was measured using handheld Doppler & ABPI was calculatedas systolic ankle BP/systolic arm BP.Pvalue was determined from by unpaired t testfor continuous variables & chi square test for discrete variables.Results: Average age of patients was 80yrs & 78yrs in the study & control grouprespectively.The average length of stay varied from 40 days (control group) to59 days(trial group) with no statistical difference.Average Barthel Index thoughsimilar at admission(14),at discharge average BI was 14(trail gr) and 17(controlgr) with statistical significance(p=0.001).More number of pt had AMT<7 in studygr with statistical significance(p=0.011).Study population had higher prevalence ofHypertension(69%) & hypercholesterolaemia(18%) which was statistically signifi-cant(p=0.002 & p=0.011).67-80% of study population had mild to moderate degreeof PVD based on ABPI which is statistically significant(p=<0.001 & p=0.001)which was previously undetected and asymptomatic.Conclusions: The results of this study favoured PVD as a common asymtomaticand unrecognised condition in patients with stroke.Early detection of PVD couldprevent compilcations in high risk patients.There is also a need to educate,identify& treat associated risk factors.The study is ongoing to include larger sample sizefor multiple regressional analysis to evaluate the role of PVD as an independentprognostic factor following stroke.



MIGRAINE MEDIATES THE INFLUENCE OF C677T MTHFR GENOTYPESON ISCHEMIC STROKE RISK WITH A STROKE-SUBTYPE EFFECTA. Pezzini, M. Grassi, E. Del Zotto, A. Giossi, R. Monastero, G. Dalla Volta,S. Archetti, C. Camarda, R. Camarda, A. PadovaniDipartimento di Scienze Medico-Chirurgiche, Neurologia Vascolare, SpedaliCivili di Brescia, Brescia, Italy

Background: To investigate the role of C677T MTHFR polymorphism in migrainepathogenesis and in the migraine-ischemic stroke pathway.Methods: A first genotype-migraine association study was conducted on 100patients with migraine with aura (MA), 106 with migraine without aura (MO), and105 subjects without migraine, which provided evidence in favour of associationof the TT677 MTHFR genotype with increased risk of MA compared to bothcontrol subjects (odds ratio [OR], 2.48; 95% CI, 1.11 to 5.58) and patients withMO (OR, 2.21; 95% CI, 1.01 to 4.82). Based on these findings, mediational modelsof the genotype-migraine-stroke pathway were fitted on a group of 106 patientswith spontaneous cervical artery dissection (sCAD), 227 young patients whoseischemic stroke was unrelated to a sCAD (non-CAD), and 187 control subjects,and a genotype-migraine partial mediation model was selected.Results: Both migraine and the TT-genotype were more strongly associated tothe subgroup of patients with sCAD (OR, 4.06; 95% CI, 1.63 to 10.02 for MA;OR, 5.45; 95% CI, 3.03 to 9.79 for MO; OR, 2.87; 95% CI, 1.45 to 5.68 forTT genotype) than to the subgroup of patients with non-CAD ischemic stroke(OR, 2.22; 95% CI, 1.00 to 4.96 for MA; OR, 1.81; 95% CI, 1.02 to 3.22for TT genotype) as compared to controls. The prevalence of migraine suffererscarrying the TT677 MTHFR genotype turned out to be higher among patientswith multiple-vessel dissection (3/16; 18.8%) than among those with single-vesseldissection (12/90; 13.3%) and control subjects (5/187; 2.7%) and the log-odds trendstatistically significant (χ2 (df) for log-odds trend = 11.2 (1); P = 00008).Discussion: Migraine may act as mediator in the MTHFR-ischemic stroke pathwaywith a more prominent effect in the subgroup of patients with sCAD.


PATENT FORAMEN OVALE, THROMBOPHILIC DISORDERS ANDMIGRAINE IN YOUNG PATIENTS WITH ISCHEMIC STROKEP. Martínez-Sánchez, B. Fuentes, M.V. Cuesta, J. Domínguez, L. Idrovo,L. Gabaldón, M.A. Ortega-Casarrubios, E. Díez-TejedorLa Paz University Hospital, UAM, Madrid, Spain

Background: Patent foramen ovale (PFO) has been associated to inherited throm-bophilic disorders, stroke and migraine separately. Our goal is to asses therelationship between PFO ± atrial septal aneurism (ASA), thrombophilic disordersand previous history of migraine in young patients with cryptogenic stroke.Methods: Observational study with inclusion of consecutive patients from theStroke Unit Data Bank (January 1995-October 2005). Patients under 55 years withan acute cryptogenic cerebral infarction were selected. We analyzed: demographicdata, vascular risk factors, stroke subtype, previous migraine and the presenceof thrombophilic disorders by a battery of hematological test. The presence ofPFO ± ASA was assessed by transcranial Doppler sonography monitoring andechocardiography.Results: 235 patients, mean age 42.97 ± 9.353 years. 16.6% had a PFO, 7.7% hadprevious migraine and 7.2% were diagnosed of a thrombophilic disorder. Patientswith PFO had less traditional risk factors such as hypertension, current smokingor coronary arterial disease (P<0.05). PFO+ASA was more common in women(9.3% vs 2.9%; P=0.044) and in patients with previous migraine (22.2% vs 4.1%;P=0,011). Thrombophilic disorders were more frequent in PFO patients (15.4% vs5.6%, P=0.043; OR 3.058: 95% CI 1.058-8.839) as well as in previous migrainepatients (22.2% vs 6%; P=0.031; OR 4.484: 95% CI 1.291-15.565). The frequencyof thrombophilic disorders was the highest in migraine + PFO patients (60% vs6.1%, P= 0.003; OR 23.143: 95% CI 3.570-150.017).Conclusions: In patients under 55 years, thrombophilic disorders are diagnosedmore frequently in PFO-related cerebral infarcts, especially in migranous patients.


RISK FACTOR ANALYSIS IN PATIENTS ADMITTED WITH RECURRENTISCHEMIC STROKESP.K. Shibu, S.H. Guptha, P. Owusu-AgyeiPeterborough District Hospital, Peterborough, United Kingdom

Background: We studied patients admitted with recurrent ischaemic stroke to

assess the prevalence and incidence of risk factors for vascular disease. We alsoassessed the control of known risk factors.Methods: All adult patients admitted with recurrent stroke in the year 2005were included in the study. The Royal College of Physicians guidelines 2004 onsecondary prevention for stroke were used as the gold standard. We used our localdiabetic guidelines to define the control of Diabetes Mellitus with a haemoglobinA1C level of less than 8 as acceptable control.Results: 81(24%) of 331 stroke inpatients in 2005 had a recurrent ischemic stroke.52(64%) patients had known hypertension of which 17(33%) were well controlledwhile 35(67%) patients had uncontrolled hypertension despite being on medication.12(15%) patients had newly diagnosed hypertension. 19(23%) patients were knownto have diabetes mellitus of which 7(37%) had acceptable glycemic control and12(63%) patients had poor control. There were no new diabetics. 37(45%)patientshad hypercholesterolemia despite being on statins. 10(12%) patients had untreatedhypercholesterolemia.12(14%) patients had new hypercholesterolemia. Significantcarotid artery stenosis was known in 3 patients and diagnosed in 2 new patientsfollowing admission. 47(58%)patients had not had carotid dopplers after theirprevious stroke. 13(16%) patients were current smokers and 39(48%) were exsmokers. 12(15%) patients had known atrial fibrillation and 5 patients had newlydiagnosed AF.Conclusions: Most patients with recurrent ischemic stroke had multiple risk factorsidentified at their first stroke but a majority of these risk factors were either undertreated or untreated. There was evidence of substantial under-investigation withcarotid dopplers. A smaller number of patients had newly identified vascular riskfactors at their admission with recurrent stroke. Better identification of risk factorsafter a stroke and more aggressive control could reduce the burden of recurrentischemic stroke.


ANEMIA AND CHRONIC KIDNEY DISEASE ARE RISK FACTORS FORMORTALITY IN STROKE PATIENTSP. Del Fabbro, J.-C. Luthi, P. Michel, E. Carrera, M. Burnier, B. BurnandCHUV Lausanne, Lausanne, Switzerland

Background: Chronic kidney disease (CKD) has been linked to higher stroke risk.Anemia is a common consequence of CKD, and recent evidence suggests thatanemia is a risk factor for cerebrovascular disease. The purpose of this study wasto examine among stroke patients, the association between CKD, anemia, hospitalmortality and one year mortality.Methods: We performed a retrospective cohort study with consecutive strokepatients over a 3 year period hospitalized in the stroke unit of a single universityhospital. We recorded demographic characteristics and risk factors for stroke. Re-duced kidney function was defined as a serum creatinine ≥80 μmol/L for womenand ≥106 μmol/L for men. Glomerular filtration rate (GFR) was also calculated.Anemia was defined when hemoglobine was below 12 g/dl on admission. The mainoutcome measures were hospital mortality and one year survival.Results: 963 patients aged between 16 and 97 years were included. Fifty-five-point-five percent had CKD according to our definition, and 12.1% had anemia.Thirty-three (3.7%) patients died during their hospitalization; among those, 27(81.8%) had CKD (p=0.002). 82 (9.6%) patients died during the first year afterdischarge from hospital, among them 54 (65.9%) had CKD (p=0.035). Anemia onadmission to the hospital was associated with an increased risk of death duringhospitalization and at one year after discharge.After adjustment for other factors, hemoglobine was independently associated withan increased risk of death at the hospital and one year survival.Conclusion: Both CKD and anemia are frequent among stroke patients and areassociated with an increased risk of hospital and one year survival.


SUBCLINICAL HEMODYNAMIC ABNNORMALITIES IN SYMPTOM-FREEHYPERTENSIVE PATIENTS DURING HEAD UP TILT TABLE TESTL. Csiba, Z. Bajkó, K. Csapó, P. SoltészDebrecen University, Debrecen, Hungary

The dysregulation of the autonomic nervous system plays an important role in thedevelopment of hypertension. The heart rate variability and baroreflex sensitivityare established methodes for the evaluation of the cardiac autonomic activity andprovide the assessment of the sympathetic and vagal activity. Our purpose wasto measure the cardiac parameters of the autonomic nervous system in essentiel,non-treated hypertensive patients and later the efficacy of the antihypertensivetreatment by non-invasive technique.Heart rate variability calculated with spectral analysis and baroreflex sensitivity



obtained by the sequence technique were measured in 28 healthy persons (age:48,29±8,14 yrs, m/f rate: 1) and 28 hypertensive patients (age: 46,11±6,54 év, m/frate: 1,1) by Task Force Monitor during head-up tilt table test.The number of the baroreflex sequences was significantly higher (66,9 vs 51,1;p=0,031), the baroreflex effectiveness index was significantly lower (67,3 v 58;p=0,027) and also significantly lower heart rate variability was measured in the lowfrequency (LF-RRI) range(213 ms2 vs 468,2 ms2, p=0,018) in the hypertensivegroup.The decreased baroreflex sensitivity and heart rate variability proved the autonomicdysfunction, the lower sympathetitc activity indicated long-standing systemichypertension in the hypertensive group.The authors will also show the changes after 6 months antihypertensive therapy(finished April, 2007).


PREVALENCE OF CONVENTIONAL VASCULAR RISK FACTORS INSTROKE PATIENTS UP TO 45 YEARS OLD. A COMPARISON WITHGENERAL STROKE POPULATIONJ. Lopez-Fernandez, A. Gonzalez-Hernandez, O. Fabre-Pi, J.A. Suarez-Muñoz,M. Vazquez-Espinar, S. Diaz-Nicolas, V. Araña-Toledo, A. Cubero-GonzalezHospital de GC Dr Negrin, Spain

Introduction: Ischemic stroke is considered a heterogeneous entity that presentsdifferences in relation to aetiology, pathology and prognosis, which can varyaccording to age groups. We have analysed the presence of conventional vascularrisk factors in a series of stroke in the young patient, which has been compared toanother series including the general population in our area.Material and methods: We performed a retrospective revision of the clinical his-tory of all patients admitted in our service in the interval 01/01/2001-31/12/2005.All ischemic strokes (ISs) or transient ischemic attacks (TIAs) in patients up to45 years-old were included. We recorded parameters related to the presence ofconventional vascular risk factors — high blood pressure (HBP), diabetes mellitus(DM), dyslipemia (DLP), tobacco smoking, ischemic cardiopathy (IC), peripheralarteriopathy and atrial fibrillation (AF) — and compared them to the number oftotal strokes in 2004 and 2005.Results: We included 536 patients (60,4% men and 39,6% women) in our period ofstudy; 70 of them (51,4% men and 48,6%) were up to 45 years-old. The prevalenceof HBP was 37,7% in the group <46 years-old and 62,9% in the total of strokes;of DM, it was 7,24% vs. 34,9%; of DLP, it was 46,7% vs. 32,5%; of active tobaccosmoking, it was 59,4% vs. 42,4%; of IC, it was 8,6% vs. 14,1%; of FA, it was 1,5%vs. 20%; and of peripheral arteriopathy, it was 0 vs. 4,3%.Conclusions: In our series, the prevalence of some conventional vascular riskfactors (HBP, DM, IC, peripheral arteriopathy and AF) is evidently minor inthe up to 45 year-old stroke population, meaning that different etiopathogenicmechanisms are involved. However, some conventional vascular risk factors (DLPand tobacco smoking) show a higher prevalence in this group, which, associatedto other predisposing factors, could favour ischemic events. For this reason, bothprimary and secondary prevention is highly necessary to avoid the conjunction ofprothrombotic disorders in the patient.


ASSOCIATION BETWEEN FAMILY HISTORY OF STROKE AND ISCHEMICSTROKE SUBTYPE IN YOUNG ADULTSK. Spengos, S. Vassilopoulou, M. Papadopoulou, A. Konstantinopoulou,P.P. Zis, E. Koroboki, G. TsivgoulisEginition Hospital, University of Athens, Athens, Greece

Background: Recent data have indicated that ischemic stroke (IS) subtype maybe associated with the family history of stroke (FHOS). Both population-basedand hospital-based studies have shown that FHOS is more prevalent in IS pa-tients. However, the potential relationship between the etiopathogenic mechanismof cerebral infarction and FHOS has not been studied in the former strokesubgroup.Methods: Consecutive first-ever stroke patients, aged between 15 and 45 yearsand hospitalised in the stroke wards or referred to the stoke outpatient clinicof our tertiary care University Hospital over a 5-year period, were prospectivelyincluded in a computerized observational data bank. Demographic characteristics,stroke risk factors and FHOS among any first-degree relative were documentedin all patients. According to the TOAST criteria, ischemic stroke was classi-fied based on etiopathogenic mechanisms into the following groups: large arteryatherosclerotic stroke (LAA), cardioembolic stroke (CE), small artery occlusion orlacunar infarction (LI), infarction of other determined origin (IOE) and infarction

of undetermined cause (IUC). Statistical analyses were performed using uni- andmulti-variate logistic regression modelsResults: FHOS was identified in 17 (11.4%) out of the 149 first-ever IS patients(mean age 38±7 years, male gender 58%). The distribution of FHOS among theTOAST subgroups was as follows: LAA 21% (4/19), CE 9% (2/22), LAC 17%(4/24), IOE 13% (5/38) and IUE 4% (2/46). FHOS was more prevalent (p=0.045)in the combined group of IS of vascular etiology (LAA/LA/IOE, 16%) than in thegroup of patients with CE and IUE (6%). After adjusting for stroke risk factors anddemographic characteristics FHOS was independently (p=0.052) associated withIS due to LAA, LAC or IOE (OR:3.1; 95%CI:1.0-10.1).Discussion: Our findings indicate that young adults with IS of vascular etiologyare more likely to have a positive FHOS than patients with CE or IUE.


METABOLIC SYNDROME IN SYMPTOMATIC AND ASYMPTOMATICPATIENTS WITH SEVERE CAROTID STENOSISL. Tuskan-Mohar, I. Strenja-Linic, K. Blazina, I. Antoncic, S. Dunatov,M. Bucuk, A. JurjevicUniversity Hospital Center Rijeka, Croatia

Background: Metabolic syndrome (MS) is a constellation of interrelated abnor-malities that increase the risk for the development of cerebrovascular disease. Theaim of this study was to analyze a group of symtomatic and asymptomatic patientswith severe occlusive disease of extracranial internal carotid artery (ICA) and thepresence of MS in these patients.Methods: One hundred and forty seven patients with severe carotid stenosis,treated at Department of Neurology, Rijeka University Hospital Center, Croatia,were included into the study. Sixty two (42%) patients were asymptomatic andeighty five (58%) were symptomatic. We analized a combination of vascular riskfactors such as obesity, diabetes mellitus, dyslipidemia and hypertension, which areknown elements of MS, in these patients.Results: There were 147 patients with severe carotid stenosis, 101 (69%) men (age65.5) and 46 (31%) women (age 68.3). MS with all its elements was documentedin 20% patients. The frequency of individual components of the MS: hypertension79%; diabetes mellitus 29%; abnormal lipid profil 67%; BMI>25 in 20% (BMI>30in 12%). The combination of hypertension and abnormal lipid profil was the mostcommon risk factor (43%). The all risk factors equally occurred in both sexses. MSwas more common in symptomatic than in asymptomatic patients (p=0.001).Discussion: In our series one fifth of patients with severe carotid stenosis had MSwhich was prevailing in symptomatic patients. The combination of hypertensionand atherogenic dislipidemia was the most frequent risk factor. Focusing on themost prominent risk factors, which are modifiable, and treating them is the mosteffective way to prevent stroke.


THE MOST FREQUENT RISK FACTORS FOR ISCHEMIC STROKE INYOUNG ADULTSM. Jovicevic, I. Divjak, A. JovanovicInstitute of Neurology, Clinical Centre Novi Sad, Univeristy of Novi Sad, NoviSad, Yugoslavia

Background: The objective of the study was to investigate risk factors for ischemicstroke in young adults.Methods: The study included 100 patients with ischemic stroke of both sexesaged 15-45 years treated at the Institute of Neurology in Novi Sad. All patientswere divided into three age groups: 15-25, 26-35, 36-45 years. All study patientsmet the clinical and radiological criteria for the ischemic stroke diagnosis. Thefollowing risk factors were studied: family history, diabetes, cardiac disease, patentforamen ovale, hypertension, previous stroke(s), smoking, hyperlipidemia, obesity,alcohol abuse, migraine, oral contraception, immunological diseases, pregnancyand puerperium.Results: Cardiac disease was found in 26% of all patients, of which none was inthe youngest group, while it was most frequent in the group 26-35 years (p=0.011).Patent foramen ovale was found in 6% of all patients and it was more frequentin younger patients. Arterial hypertension was present in 51% of all patientsand the percentage of patients with this risk factor was significantly higher inolder age groups (8.3%:25%:63.9%) (p<0.001). Smoking was the most frequentrisk factor, present in 55.6% and equally distributed in all age groups (p=0.918).Hyperlipidemia was the second most frequent risk factor (53.5%). Most patientswith hyperlipidemia were in the oldest group, however there was no significantdifference (p=0.406). Oral contraception was used by 4% of patients. Five femalepatients were pregnant or in puerperium (2:1:2).


Discussion: Smoking was the most frequent risk factor, registered in 55.6% ofall study patients. Study of risk factors is essential for adequate prevention andtreatment of ischemic stroke.


HEREDITARY TROMBOPHILIAS AND NEUROLOGICAL COMPLICATIONSJ. Szilasiova, B. Benova, D. Kozakova, E. Kahancova, E. Antolova,Z. GdovinovaFaculty of Medicine P.J. Safarik University and Faculty Hospital Kosice,Slovakia

In patients with unknown etiology of stroke hereditary coagulation disorders canbe the reason of stroke. There are also other neurological disorders which can becaused by thrombophilias. Studies have shown that trombophilia testing influencesless than 25% of physician’s treatment of stroke.The authors present small study consists from 10 patients (8 women, 2 men),mean age 28.5 years, with different types of hereditary coagulation disordersaffecting cerebral circulation. In 5 patients (50%) isolated and in the next 5patients (50%) combined hereditary coagulation deffect were found. Deficit ofantithrombin (AT) III, mutation of gene for methylenetetrahydrofolate reductase(MTHFR), hyperhomocysteinemia, high factor VIII, factor V Leiden, protein S andC deficit, increased level of plasminogen activator inhibitor (PAI) webe detcted.In four patients thrombosis of cerebral arteries and thrombembolic events werepreceded by delivery or using contraceptives. In four patients coagulopathy wasfound also in family members. In 9 cases neurological complications were the firstmanifestation of thrombophilia, in one stroke was preceded by phlebothrombosisof lower extremities. In five patients brain infarcts, in two cases thrombosis ofbrain venous sinuses were found and in the last three patients magnetic resonancerevealed demyelinating lesions.After therapy two patients have mild hemiparesis, next two vascular epilepsy andlast six patients are without any consequences. In patient with AT III deficit wasin the same time present thrombembolia of pulmonal, radial, axillar and cerebralarteries with numerous brain infarcts.Hereditary thrombophilias are rare reasons of stroke, but with higher frequency inyounger patients. Their detection may provide prognostic information of the risk ofrecurrent events as well as determine the most appropriate treatment.


INFLUENCE OF VASCULAR RISK FACTORS ON OXIDATIVE STRESS,ASSESSED BY MEASUREMENT OF SERIC MDA LEVELS, IN ACUTEISCHEMIC STROKE SUBTYPESA. SimionFaculty of Medicine and Pharmacy Oradea/Clinical Hospital of Neurology andPsychiatry, Oradea, Romania

Background: Oxidative stress has been involved in the pathogenesis of severaldiseases considered as risk factors for ischemic stroke.Material and method: A consecutive series of ischemic stroke patients admittedwere evaluated clinically, with brain CT and/or MRI, Duplex sonography, electro-cardiography, transthoracic echocardiography, and biochemically (measurement oflipid fractions and fasting glucose levels).Strokes were divided into large artery dis-ease (LAD), small vessel disease (SAD) and cardioembolism (CE) according to theTOAST criteria.Risk factors were diagnosed according to the international criteria(JNC 7 for hypertension, ADA for diabetes, BMI>30 kg/m2 for obesity).Smokingand alcohol intake were recorded as admitted by the patient.We recorded the NIHSSscore on admittance and Barthel index (BI) at discharge.Malondialdehyde (MDA)levels were measured on admittance(day 1, viewed as baseline levels)and on day 3and 7.Results: Patients with fewer risk factors had lower baseline-MDA levels.Smokingraised the MDA levels in LAD significantly(p<0,001)and lowered BI(2,43 to 4,6;4,18 to 5,9; 3,13 to 5,3).In SAD only baseline-MDA level was significantly higher(1,83 to 4,5, p<0,001).Diabetes significantly increased oxidative stress at firstdetermination and worsened outcome in SAD(1,83 to 4,1, p<0,001).In CE theMDA values were not as high as in microangiopathy, but outcome was even poorer.Dislipidemia, obesity and hypertension raised the baseline-MDA levels nonsignifi-cantly, except obesity in CE in which both MDA(2,17 to 3,5, p<0,001)and outcomewere significantly altered.Alcohol intake raises (p<0,001) MDA level in LAD (2,43to 5,1; 4,18 to 5,9; 3,13 to 5,3).Conlusions: Based on our results preventive antioxidant therapy would be mostbeneficial in diabetes.


THE RATE OF STROKE ASSOCIATED WITH THE USE OF THORATECVENTRICULAR ASSIST DEVICE IN OLDER PATIENTSM.B. Vijayappa, B. Clemson, M. Alsorogi, S. Al-Hawarey, D. Joseph,A. Talkad, M. Mathews, D. WangUniversity of Illinois College of Medicine at Peoria, OSF, INI, Peoria, IL, USA

Background: Because of the shortage of heart donors, ventricular assist devices(VAD) are used to provide mechanical circulatory support in patients with heartfailure awaiting planned heart transplant or in patients who are not candidates forheart transplant. The use of VADs has been increasing due to the growth of theaging population. However, it is unclear if there is an increased risk of developingstroke relative to age. Our study was to determine the effect of age on the risk ofstroke after VAD implant.Methods: This is a retrospective study exempted by the community IRB. From1996-2006, we reviewed records of all patients who received VADs as a bridgeto transplant. Demographic data, clinical and radiographic findings were collected.Descriptive statistics were applied. The patients were categorized according to age;Group I, less than 55 and Group II, 55 or older.Results: From 1996-2006, 45 patients received VADs. The number of patients ingroup I was 24 (53%) and group II was 21 (47%). Group I had 8 (33%) strokes; 6(23%) were ischemic and 2 (8%) hemorrhagic. Group II also had 8 (30%) strokes,7 (32%) were ischemic and 1 (5%) hemorrhagic. In Group I; 4 (17%) had diabetes,13 (54%) hypertension and 7 (29%) left atrial cannulation. In Group II; 11 (52%)had diabetes, 11 (52%) hypertension and 6 (29%) left atrial cannulation. The overallmortality in group I was 11 (46%) and 10 (48%) in group II. One (4%) death ineach Group was associated with stroke.Conclusion: Regardless of patient’s age, the occurrence of stroke associated withthe implantation of a VAD was similar. There was no significant increase in strokerelated mortality. Given the known benefit of VADs support to improve the rateof successful transplantation and the absence of an increased risk of stroke inolder patients, the use of VAD support in appropriate patients should be carefullyconsidered regardless of age.


ASSOCIATION OF CAROTID INTIMA MEDIA THICKNESS AND PLAQUEWITH AORTIC ARCH CALCIFICATIONB.-S. Shin, E.S. LimChonbuk National University Hospital and Medical School, Jeonju, SouthKorea

Introduction: Atherosclerosis is a generalized process and is the major cause ofcerebrovasculardisease. The more advanced atherosclerotic lesions contain calciumdeposits. Aortic arch calcification may represent a manifestation of generalizedatherosclerosis. The carotid intima media thickness (IMT) is also recognized asindependent predictors of adverse cerebrovascular outcomes. The purpose of thisstudy is to examine the relationship between the degree of aortic arch calcificationand carotid IMT and plaque.Materials and methods: A total of 56 patients (31 men and 25 women) wereincluded. All patients had chest radiography in the posteroanterior view. Severity ofaortic arch calcification was graded as follows: grade 1 (small spots of calcificationor single thin calcification of the aortic knob), grade 2 (one or more areas ofthick calcification), grade 3 (circular calcification of the aortic knob). The degreeof carotid atherosclerosis was evaluated by measuring the maximum IMT of thecommon carotid artery, carotid bulb, and internal carotid artery by duplex carotidultrasonography.Results: The mean age of patients was 70.0 ± 9.1 years and the numbers of malepatients was 31. (55.4%). Severity of aortic arch calcification was grade 1 in 26patients (46.4%), grade 2 in 18 patients (32.1%), grade 3 in 12 patients (21.4%).Aortic arch calcification was observed more commonly in elderly patients. Therewas no significant difference between aortic arch calcification and gender. Therewas no significant relationship between severity of aortic arch calcification andcarotid IMT. But there was significance between severity of aortic arch calcificationand plaque number (p=0.012).




ISCHEMIC STROKE IN PATIENTS UNDERGOING DIALYSISS. Jung, S.H. Hwang, S.B. Kwon, K.H. Yu, I.S. Koh, B.C. LeeHallym University College of Medicine, National Medical Center, Seoul, SouthKorea

Background: When we evaluated the patterns of stroke and risk factors in patientswith chronic kidney disease, it was difficult to analyze those results because theincluded patients were not homogenous. In that study, we included ischemic andhemorrhagic stroke patients with either chronic renal disease or end stage renaldisease (ESRD) undergoing dialysis. We performed this study in patient withischemic stroke with ESRD so as to elucidate the factors that affect developingischemic stroke in chronic kidney disease.Methods: We enrolled patients with acute ischemic stroke in ESRD patientsundergoing dialysis using the data from Hallym Stroke Registry (HSR). Twentypatients were included from July 1996 through June 2006. Thirty patients withhemodialysis who did not develop stroke were selected as control group with ageand sex matching. We compared the demographic features, stroke risk factors,laboratory findings in ESRD patients with or without ischemic stroke.Results: The mean age of acute ischemic stroke patients was 60.75±10.19 andmale to female ratio was 0.82. The most common ischemic stroke subtype wassmall vessel occlusion (SVO, 9 of 20 patients) and followed by large arteryatherosclerosis (LAA, 7 of 20). Ischemic stroke patients showed longer durationof causative disease of ESRD, higher rate of hypertension history, and low serumLDL cholesterol level compared with the patients with control group.Discussion: Although low serum LDL level in stroke with CKD in our previousstudy could be affected by relatively high frequency of hemorrhagic stroke, lowserum LDL was also observed in ischemic stroke patients undergoing dialy-sis. These finding shows that serum cholesterol level can reflects the status ofmalnutrition rather than atherosclerosis at least in CKD patients.

Stroke and diabetes

1 Stroke and diabetes

THE EFFECT OF STATIN PRE-TREATMENT ON INFARCT VOLUME ANDDISCHARGE DISPOSITION IN ISCHEMIC STROKE PATIENTS WITHDIABETESJ.S. Nicholas, J.C. Thomas, Z. Rumboldt, P. Tumminello, S.J. PatelMedical University of South Carolina, Charleston, SC, USA

Background: Studies have indicated reduced rates of stroke among patients withdiabetes treated with statins. In contrast to stroke prevention, the purpose of thisanalysis was to examine post-stroke outcomes (infarct volume, discharge disposi-tion) among patients with diabetes who were taking a statin at onset of ischemicstroke relative to those who were not.Methods: Study design was a retrospective cohort analysis of all verified ischemicstroke patients admitted to our university hospital 2002-2006 with magnetic res-onance diffusion weighted imaging (DWI). Of these patients, 131 presented withdiabetes or were newly diagnosed at admission. Infarct volume was calculated fromDWI, blinded to statin status. For patients with multiple infarcts, volume recordedwas the sum of all infarcts. Discharge disposition and clinical data were abstractedfrom hospital records. Statistical comparisons between statin and no-statin groupswere made using 2-sided t-tests for continuous variables, chi-square for categor-ical, and Mann-Whitney for nonparametric. Variables pre-specified as potentialconfounders of infarct volume were time to imaging, location and type of stroke.Results: Patients with diabetes who were taking a statin on admission had a statis-tically significant 55.1% decrease in median infarct volume relative to the no-statingroup (1.02 cm3 statin versus 2.27 cm3 no-statin, Mann-Whitney p=.039, N=131,46.6% on statins). While differences in discharge disposition were not statisticallysignificant, those on statins were discharged sooner (mean 4.75 days versus 5.82)and were more likely to be discharged home (37/61=60.7% versus 38/70=54.3%).Time to imaging, location and type of stroke did not differ significantly betweengroups in univariate analysis, nor did age, gender, or race (all p values >.05).Discussion: In this study, statin-pretreated patients with diabetes experiencedsignificantly smaller median infarct volumes following ischemic stroke than thosenot pretreated. Given the potential importance to treatment practices, this findingand its implications for clinical outcome should be further investigated.


THE TEMPORAL PATTERN OF POST-STROKE HYPERGLYCAEMIA IS APREDICTOR OF LONG-TERM SURVIVALM. Koufali, R. Durairaj, R. Kumar, A.K. SharmaAintree Stroke Team for Audit and Research, Aintree University Hospitals NHSFoundation Trust, United Kingdom

Background: Recent work has shown that persistent post-stroke hyperglycaemia(PHG) affects infarct size and clinical outcome. The purpose of this studywas the detailed investigation of the association between blood glucose levels(BGL)obtained over 24 hours and survival at 12 months following a stroke.Patients & Methods: 1496 patients, admitted during 2000-2005, with a confirmeddiagnosis of ischemic stroke and not previously known to be diabetic were includedin this retrospective analysis. Only Caucasian patients with confirmed onset, ad-mission, discharge and death dates were included in the study. BGL values wereobtained upon admission and at 24 hours.Results: Median admission BGL was 6.3 mmol/L, while median BGL at 24 hourspost admission was 5.5 mmol/l. 987(66%) of patients survived at 12 weeks, 568(38%) survived at 12 months, while 284(19%) patients died in hospital.807 patients (54% of total) were hyperglycaemic upon admission (glucose >6.1mmol/L) with 256 of them (32%) remaining hyperglycaemic at 24hours post-admission.Multiple regression analysis demonstrated that 24 hour PHG is independentlyassociated with: Death in hospital (p=0.017), survival at 12 months (p=0.009) andincreased lengths of stay (p=0.005).Conclusion: Evidence is presented that patients with persistent hyperglycaemiaover 24 hours have poorer outcomes. Further research is thus warranted whichwould allow us to target this group of patients for aggressive therapy for bloodsugar control following a stroke.


POST-STROKE HYPERGLYCAEMIA IS NOT AN EPIPHENOMENON OFISCHAEMIC STROKE SEVERITYM.T. McCormick, T.A. Baird, K.W. MuirUniversity of Glasgow, Glasgow, United Kingdom

Introduction: Post stroke hyperglycaemia (PSH) is common and is associated witha worse outcome. It is postulated that hyperglycaemia reflects the stress responseof a severe stroke. We sought to establish the prevalence of PSH within 48hours ofictus and describe its association with stroke severity.Methods: Patients presenting within 24hours of a suspected stroke (April 2004-January 2006) underwent 4 hourly capillary blood glucose (CBG) monitoring for48hours. Baseline demographics, NIHSS, OCSP and time of stroke onset werecollected. Stroke severity was described as severe (NIHSS > 15); moderate (NIHSS7–15) or mild (NIHSS 0–6). Hyperglycaemia was defined as a CBG > 7mmol/l,Results: 353 patients underwent CBG profiling. Median Age 72 (IQR 62,80);Median NIHSS 6 (3,13). 17% had diabetes. At presentation 29% of patients hadPSH; Median time to CBG was 238 minutes (IQR 165,494). Over the 48hour moni-toring period 75% developing Hyperglycaemia, with 25% euglycaemic throughout.Glycosylated Haemoglobin was statistically lower in patients with euglycaemiacompared to hyperglycaemia (p = 0.001). Stroke severity was not predictiveof admission hyperglycaemia whereas glycosylated haemoglobin was (OR 2.97;95%CI 1.84-4.78; p < 0.001). There was no statistically significant differencebetween mean blood glucose on initial monitoring between groups. However at 48hours, blood glucose was significantly lower in more severe strokes (NIHSS > 15),CBG = 5.8mmol/l compared to milder strokes (NIHSS 0–6), CBG = 6.6mmol/l (p= 0.015).Discussion: Post stroke Hyperglycaemia is common. Overall prevalence for the48hour period was 75%. Stroke severity was not predictive of post stroke hy-perglycaemia with more severe strokes having a statistically lower blood glucosecompared to milder strokes at 48hours.


DIABETES MELLITUS AND THE EARLY RISK OF STROKE AFTERTRANSIENT ISCHEMIC ATTACK: A HOSPITAL-BASED CASE SERIESSTUDYG. Tsivgoulis, S. Vassilopoulou, E. Manios, P.P. Zis, K. SpengosEginition Hospital, University of Athens, Athens, Greece

Background: California- and ABCD-scores reliably predict short-term risk ofstroke after TIA. Both scores contain similar components. However, diabetes


mellitus (DM) is only included in the California Score. Aim of the present studywas to evaluate the potential relationship of DM with the early risk of stroke in acohort of hospitalised TIA patients.Methods: All patients hospitalised in our Department with definite TIA during a5-year period were identified and their medical charts as well as their EmergencyRoom records were retrospectively reviewed by two investigators blinded to follow-up. Patients with previous history of stroke and those who missed their follow-upevaluations at the outpatient clinic of our Department at 1 month after admissionwere excluded. DM was specified as fasting serum glucose ≥7.0 mmol/L, nonfast-ing serum glucose ≥11.1 mmol/L, or use of oral blood sugar-lowering drugs orinsulin. The outcome events of interest in all TIA patients were subsequent strokesduring the 1-month follow-up period. Statistical analyses were performed usingthe Kaplan-Meier product-limit method and stepwise Cox’s proportional hazardsmodel.Results: The 30-day risk of stroke in the present case series (n=226) was 9.7%(95%CI:5.8-13.6%; 22 events). The 30-day risk of stroke was higher in patients withDM (17.3%; 95%CI:7.6-27.0%) than in non-diabetic patients [(7.1%; 95%CI:3.2-11.0%); log-rank test=5.20; df=1; p=0.0225]. After adjustment for demographiccharacteristics, stroke risk factors, history and number of prior TIAs, duration andsymptoms of TIAs, as well as secondary prevention treatment strategies duringhospitalisation, DM was independently (p=0.015) associated with a three-foldgreater 30-day risk of stroke (HR:2.98; 95%CI:1.28-6.94).Discussion: DM is an independent predictor of subsequent stroke in patientspresenting with TIA. It should be taken into account by prognostic scores thatstratify the risk of early stroke in TIA patients.


PATIENTS WITH STROKE HAVE A HIGH PROBABILITY OFDYSGLYCAEMIA AS ASSESSED USING AN ORAL GLUCOSE TOLERANCETESTS.H. Naqvi1, J.D. Lee2, V. Patel3, K.M. Sharobeem1

1Sandwell General Hospital, Rowley Regis; 2Department of Cardiology, GeorgeEliot Hospital NHS Trust, Nuneaton; 3Warwick Medical School, UnitedKingdom

Background and aims: In the UK, the recently published Joint British Societies’Guidelines on the Prevention of Cardiovascular Disease recommend performing anoral glucose tolerance test (OGTT) on subjects with an acute cardiovascular event.In this study, we report our experience of performing an OGTT in those suffering arecent acute stroke in a district general stroke unit.Methods: An OGTT was performed in clinically stable consecutive patients ad-mitted to the stroke unit with a diagnosis of acute stroke. The study was conductedover a period of 3 months. Those with known diabetes mellitus were excluded.Results: Data on 71 patients were available. Mean age of subjects was 70 years(Range 24-96). 49% were male. 86% suffered a non-haemorrhagic stroke. Themean time of admission to OGTT was 8.5 days. Only 43% had a normal OGTT. Theproportions of those with impaired fasting glycaemia, impaired glucose toleranceand diabetes mellitus were 3%, 34%, and 20% respectively.Conclusions: Our data suggests that, based on the results of an OGTT, patientssuffering an acute stroke have a high probability of having dysglycaemia. Thegold standard OGTT for diagnosing dysglycaemia should be employed as there isdata supporting the prevention of diabetes mellitus in those with impaired glucosetolerance. Furthermore, early treatment of diabetes mellitus can potentially offsetthe appearance of classic diabetes complications.


ETHNIC DIFFERENCES IN COMORBIDITIES AND STROKE SUBTYPES OFDIABETIC STROKE PATIENTS: THE SOUTH LONDON STROKE REGISTER(SLSR)A. Elmarimi, T. Rashid, J. Adie, O. Wood, P.U. Heuschmann, A.M. Toschke,A.G. Rudd, C.D.A. WolfeKing’s College London, London, United Kingdom

Background: Major differences in underlying risk factors of stroke have beenidentified between ethnic subgroups, which may have implications in directingsecondary preventive strategies. Data are lacking about differences in comorbiditiesand stroke subtypes between black and white diabetic stroke patients.Methods: Data were collected from the South London Stroke Register (SLSR),a population-based stroke register covering a multiethnic source population of271.817 inhabitants (2001). Analysis was restricted to those with known diabetesand to patients of black or white ethnic group. Demographics, major vascularrisk factors and stroke subtype were compared between black and white ethnic

groups. Detected differences were adjusted for age and sex by multivariable logisticregression.Results: Between Jan 1995 and Dec 2004, diabetes was identified in 419 patients(16.9%). The mean age of black diabetic patients was significantly lower than thatof whit patients (68.9y versus 73.4y, respectively; p<0.001); no differences werefound for sex. Among black diabetic patients, the prevalence of atrial fibrillation(OR 0.2; 95% CI 0.1-0.4) and of smoking (OR 0.5; 95% CI 0.3-0.7) was lower,and presence of hypertension higher (OR 2.7; 95% CI 1.5-4.8) compared to white;no other statistically significant differences for vascular risk factors were found.Distribution of stroke pathology and of clinical subtypes showed no statisticallysignificant variation between the two ethnic groups.Conclusions: Substantial ethnic differences were found in the prevalence of vascu-lar risk factors between black and white diabetic stroke patients. These differencesmight reflect different risk factor profiles and possibly point to 2 clusters withdifferent interactions between risk factors

Etiology of stroke

1 Etiology of stroke

CEREBRAL HEMODYNAMIC FEATURES OF ALCOHOLIC ABSTINENTSYNDROME AND OPIATE ABSTINENT SYNDROMEI. Voznyuk, N. Artemyeva, S. ScherbakMilitary Medical Academy, Russian Federation

Background: Chronic alcoholization and heroinism are characterized by varioushemodynamic disorders, that promote the development of the cerebral pathology.Methods and materials: The research involved 55 patients with the alcoholicabstinent syndrome (26-59 years old) and 40 patients with the opiate abstinentsyndrome (25-41 years old). The control group included 34 men and women(18-47 years old). The clinical data and results of the Doppler sonography wereinvestigated.Results: The similar decrease of the linear blood flow velocities in both middlecerebral arteries was registered in 60% of the patients with the alcoholic abstinentsyndrome and in 54,5% of the patients with the opiate abstinent syndrome; inboth anterior cerebral arteries – in 56,4% and 43,8% correspondingly; in bothposterior cerebral arteries – in 47,3%, in both vertebral arteries – in 32,7% of thepatients with the alcoholic abstinent syndrome. The decrease of the linear bloodflow velocities in the basilar artery was detected in 58,2% of the patients with thealcoholic abstinent syndrome and in 20,7% of the patients with the opiate abstinentsyndrome; the diffuse increase of the vascular tone – in 54,5% and 48,5%; thedecrease of the integrated index of the cerebrovascular reactance – in 85,5% and68,8% correspondingly. Restriction of the cerebrovascular reactivity was registeredin 41,8% of the patients with the alcoholic abstinent syndrome (vasodilatation)and in 28% of the patients with opiate abstinent syndrome (vasodilatation andvasoconstriction).Conclusion: The decrease of the linear blood flow velocities in the intracranialarteries, the diffuse increase of the vascular tone and the decrease of the integratedindex of the cerebrovascular reactance occurs reliably more often in the patientswith the alcoholic abstinent syndrome and opiate abstinent syndrome, than in thecontrol group. The similar bilateral decrease of the linear blood flow velocities waspredominantly registered in the middle cerebral arteries and in the anterior cerebralarteries.


INTERPRETATION OF COMPUTED TOMOGRAPHY ANDDIFFUSION-WEIGHTED IMAGING DURING ACUTE NEUROLOGICALEVENTS IN MITOCHONDRIAL RESPIRATORY CHAIN DISORDERSS. Mittal, W. Watson, M. Aribandi, J.P. HoseyGeisinger Medical Center, Danville, PA, USA

Sudden neurological events are common and mimic acute stroke in mitochondrialrespiratory chain disorders characterized by disruption of intracellular metabolicpathways and energy failure. We had the opportunity to examine four separatesets of computerized tomography (CT), diffusion-weighted imaging (DWI) andapparent diffusion coefficient (ADC) data collected during acute neurologicalevents from two patients with defects in the mitochondrial respiratory chainfunction. Both had prior histories of mental retardation and seizures but no knownmitochondria disorder. Clinically the events were characterized by acute onsethemispheric symptoms suggesting stroke. Initial evaluation with CT brain showedhypodensity involving the temporoparietal region of the symptomatic hemisphere

Poster SessionEtiology of stroke


with effacement of sulci and surrounding edema indicative of subacute stroke.DWI and ADC map were obtained within the first 48 hours; there was a delayof 168 hours in one event. All four DWI/ADC data were strikingly similar andshowed restricted diffusion mixed with increased ADC in the temporoparietalarea indicating a combination of cytotoxic and vasogenic edema. One patient hadA3243G mutation and the other a deficiency of cytochrome c oxidase activity andpartial deficiency of NADH dehydrogenase. Though presented acutely CT scansin all four events suggested subacute stroke. With newer treatment modalities forischemic stroke widely available it is important to distinguish them from acuteevents in mitochondrial encephalopathies. DWI helps distinguish between cytotoxicand vasogenic edema. Membrane ion pump failure and ingress of sodium into thecell during acute ischemia cause restriction of movement possibilities of watermolecules. Resultant uniformly reduced values of ADC are seen in infracted areaconfined to a vascular territory. In our observation restricted diffusion was notablyconfined to the cortical areas while the adjacent white matter showed increasedADC which was felt to be unique to patients with mitochondrial encephalopathypresenting as acute neurological events.


STROKE PATTERNS IN PATIENTS WITH INTERNAL CAROTID ARTERYDISSECTION – THE SIGNIFICANCE OF VESSEL PATENCYL.H. Bonati, S.G. Wetzel, J. Gandjour, R.W. Baumgartner, P.A. Lyrer,S.T. EngelterUniversity Hospital Basel, Basel, Switzerland

Background: Spontaneous dissection of the internal carotid artery (ICAD) is animportant cause of stroke in young and middle-aged patients. ICAD may lead toa complete vessel occlusion or residual flow through a stenotic artery. Diffusion-weighted imaging (DWI) has the potential to highlight differences in the pattern ofcerebral ischemia between occlusive and non-occlusive ICAD.Methods: DWI, Doppler and color duplex sonography, and angiographic studieswere assessed in 40 consecutive patients (median age 47 years, interquartile range[IQR] 39-56) with ischemic stroke caused by spontaneous ICAD, referred to twouniversity hospitals. Number, size and location of hyperintense lesions on DWIwere correlated with vessel patency.Results: Patients with non-occlusive ICAD (n=15) presented with more ischemiclesions (median 5, IQR 1-10) than patients with complete ICA occlusion (n=25)(2, 1-3; p=0.014). In contrast, ischemic lesions were larger in occlusive ICAD (62,50-99 mm) compared to non-occlusive ICAD (25, 10-50 mm; p=0.007). Strokepatterns differed significantly between the two groups (p=0.002). Non-occlusiveICAD was associated with disseminated lesions involving borderzone territories,whereas most patients with occlusive ICAD had large territorial infarcts.Conclusions: Our data suggest different stroke patterns in ICAD patients withoccluded ICA compared to those with stenotic ICA. These findings may be ofrelevance in the ongoing controversy about the acute treatment of stroke in ICAD.


STROKE IN PATIENTS WITH CANCER IN A GENERAL HOSPITAL:DIFFERENCES BETWEEN CLINICAL AND AUTOPSY SERIESG.S. Silva, D.L. Gomes, M.M. Alves, J.A. Fiorot Jr, A.R. MassaroUNIFESP- Universidade Federal de São Paulo, São Paulo, Brazil

Background: Cerebrovascular disorders (CVD) are frequent causes of neurologicalsymptoms in cancer patients. Clinical and autopsy series differ in the importanceof cancer specific conditions as causes of CVD in these patients. Our aim was todescribe the clinical features of patients with CVD and cancer admitted to a generalhospital, and to compare them to autopsied patients with the same diagnosis.Methods: A retrospective analysis of patients with the diagnosis of CVD andcancer in their discharge summaries from July 2005 to July 2006 was performed.The necropsies executed in our hospital from January 2004 to June 2005 werereviewed. Data collected included: demographic, stroke and cancer features.Results: Eleven patients (mean age 54 ± 13 years, 7 women) had a diagnosisof CVD and cancer. Intracerebral hemorrhages (ICH) (55%- 9% subdural and46% intraparenchimal) were more frequent than ischemic strokes. Primary cancerswere: leukemia (36.4%), lung (18.2%), primary intracerebral cancer (18.2%),colon, prostate and genitourinary tract (9.1% each). TOAST classification wascardioembolic in 50% of the patients and undetermined in 50%. Oxfordshireclassification was partial anterior circulation in 66.7% and posterior circulationin 33.3% of the patients. 66.3% of the patients died, and in 57.1% death wasstroke related. Echocardiogram was performed in 18.1% of the patients, magneticresonance imaging in 9% and Doppler in 9%. From 350 necropsy studies, fourpatients had a diagnosis of CVD and cancer (mean age 48 ± 63, 3 women). One

patient had an ICH into a germinative cell tumor metastases and the other threehad ischemic strokes- two had identified intracardiac emboli source (a fungi and anonbacterian thrombotic endocarditis).Discussion: Patients with CVD and cancer in our hospital seem to receive a lessthorough investigation of stroke etiology. Autopsy data seem to differ from clinicalones, maybe due to the severity of patients evaluated. A better investigation of thesepatients could help identify preventable causes of stroke recurrence.


EMBOLIC LESION PATTERN ON DIFFUSION WEIGHTED BRAIN IMAGINGAND AETIOLOGY OF STROKEJ.N.E. Redgrave, A. Chandratheva, D. Briley, P.M. RothwellStroke Prevention Research Unit, Department of Clinical Neurology, OxfordUniversity, Oxford, United Kingdom

Background: Diffusion-weighted MR-imaging (DWI) is highly sensitive to acutecerebral ischaemia and may help to determine the likely underlying aetiology.For example, several studies have reported an association between multiple acuteischaemic lesions and ipsilateral 50% carotid stenosis, although none has quantifiedthe predictive value.Methods: Consecutive patients referred to a specialist clinic with TIA or minorstroke had DWI and carotid MR-angiography. Ipsilateral 50% carotid bifurcationstenosis was related to the presence of solitary and multiple acute ischaemic lesionson DWI.Results: 500 patients (278 men) were studied. DWI showed acute ischaemiclesion(s) in 179/280 (63.9%) patients with minor stroke vs. 33/220 (15%) withTIA (p<0.0001). 52 (10.4%) patients had symptomatic 50% carotid stenosis. Asolitary DWI lesion was not associated with symptomatic carotid stenosis (OR1.00, 0.51-1.99, p=0.99) but multiple acute DWI lesions in the ipsilateral carotidterritory was strongly associated (OR 5.87, 2.68-12.86, p<0.001) with carotidstenosis, particularly the presence of >3 lesions (OR 8.87, 3.26-24.15, p<0.001).However, the corresponding sensitivities for prediction of the presence of 50%ipsilateral carotid stenosis were nevertheless low (25% and 15% respectively) andthe imaging appearance was also associated with cardioembolic aetiology.Conclusion: Multiple acute ischaemic lesions on DWI are strongly associatedsymptomatic 50% carotid bifurcation stenosis, but the appearance is likely to be oflimited use in aetiological classification.


PREVALENCE OF FABRY’S DISEASE IN YOUNG MALE PATIENTS WITHSTROKE OR TRANSIENT ISCHEMIC ATTACKA. Viguier, T. Levade, V. LarrueCHU de Toulouse, Toulouse, France

Background: Fabry’s disease is a rare inherited, X-linked storage disease that maycause stroke or transient ischemic attack (TIA) in young subjects. We sought todetermine the prevalence of this disease in an unselected population of young malepatients with stroke or TIA.Methods: Peripheral blood leukocyte alpha-galactosidase A activity was deter-mined in male patients aged 18-55 years, consecutively admitted to a tertiary strokeunit for acute stroke or TIA. Causes of stroke or TIA were classified using theTOAST classification.Results: 108 patients were recruited from January 2004 to June 2005. The meanage ± SD was 45 years ± 9. There were 71 ischemic strokes, 27 TIAs, 8 cerebralhemorrhages and 2 venous cerebral thrombosis. Causes of stroke or TIA were un-determined in 51/98 (52%) patients. No patient had a deficient alpha-galactosidaseA activity.Discussion: These findings confirm the rarity of Fabry’s disease among unselectedyoung male patients with stroke and TIA.


WHITE MATTER LESION LOAD IDENTIFIED BY MAGNETIC RESONANCEIMAGING IS NOT RELATED TO IPSILATERAL CAROTID ARTERYSTENOSISF.N. Doubal, J.M. Wardlaw, M.S. DennisUniversity of Edinburgh, Edinburgh, United Kingdom

Introduction: Cerebral white matter hyperintensities (WMH) on T2-weightedmagnetic resonance (MR) are common but their aetiology is unknown. They areassociated with hypertension, diabetes and atherosclerosis, and may be caused by


small vessel disease or microemboli (e.g. from the heart, aortic arch or carotid arterystenosis). Studies have linked the severity of any carotid stenosis to the severityof whole brain WMH load but have not specifically examined brain ipsilateral toa stenosis. We hypothesised that if microemboli cause WMH, e.g. from carotidstenoses, more WMHs would be found in the cerebral hemisphere ipsilateral to astenosis than in the hemisphere distal to a non-stenosed carotid.Methods: We prospectively recruited patients with lacunar or mild corticalstroke from a tertiary hospital. Patients were imaged with a 1.5T MR scanner(T2/DWI/GRE/FLAIR) and carotid doppler ultrasound. We dichotomised carotidstenosis as >or <50% NASCET. MRI scans were scored for deep and periventricu-lar WMHs in each hemisphere using the Fazekas method, blind to carotid stenosis.We compared hemispheric Fazekas scores between patients with or without uni orbilateral carotid stenosis.Results: Of 79 patients, 14 had asymmetrical carotid stenosis (one > and one<50%), 1 had bilateral >50% stenoses and 64 had bilateral <50% stenoses. In the14 with asymmetrical stenoses there was no difference in deep or periventricularWMHs between the ipsilateral (mean deep Fazekas 1.78; mean periventricularFazekas 1.42) or contralateral hemisphere (1.78;1.36) to the stenosis. In the 64 withno stenosis bilaterally there was no difference in mean deep and periventricularFazekas scores between the left (1.64;1.03) and right (1.62;1.08). The patient withbilateral stenoses had identical scores in each hemisphere.Conclusions: We found no link between cerebral hemisphere WMH score andipsilateral carotid artery stenosis, suggesting that microemboli, at least from carotidstenoses, are unlikely to cause most WMHs.


VASCULAR RISK FACTORS IN TRANSIENT ISCHAEMIC ATTACKS ANDLATE-ONSET EPILEPSYM.O. McCarron, A. Hunter, M. WattAltnagelvin Hospital, Derry, United Kingdom

Background: There is emerging evidence that late onset seizures may be the firstmanifestation of occult cerebrovascular disease. We sought to compare the vascularrisk profiles in patients with transient ischaemic attacks and patients with late-onsetepilepsy for which no other cause was identified.Methods: Consecutive patients attending a neurovascular clinic with a diagnosisof transient ischaemic attack (TIA) or epilepsy were eligible for study. Vascularrisk factors were prospectively recorded and compared in each group with the chisquare test.Results: One hundred patients were enrolled in each group. The distribution ofvascular risk factors was as follows: atrial fibrillation (2 in seizure group versus 0in TIA group, NS), hypertension (59 versus 52, NS), diabetes (9 versus 10, NS),cigarette smoking (52 versus 45, NS), heart failure (1 versus 0, NS), hyperlipi-daemia (35 versus 33, NS), ischaemic heart disease (18 versus 23, NS), previouscerebrovascular disease (33 in seizure group versus 7 in TIA group, p<0.001),excess alcohol (19 versus 16, NS), and family history of stroke (52 versus 50, NS).Discussion: Vascular risk profiles are very similar among TIA and late-onsetepilepsy patients. This study also supports existing evidence that previous cere-brovascular disease is a risk factor for late-onset epilepsy.


PATTERN OF INTRACRANIAL VERSUS EXTRACRANIALATHEROSCLEROTIC CEREBROVASCULAR DISEASE IN INDIAN PATIENTSWITH STROKE- AN ANGIOGRAPHY STUDYS. Husain, S. Sukumaran, A. Vajpayee, S.U. Khan, K.M. Rahman,S. Chaturvedi Gopal, S. SharmaSir Ganga Ram Hospital, New Delhi, New Delhi, India

Objective: To evaluate the intracranial atherosclerosis among patients of atheroscle-rotic stenos-occlusive extracranial carotid artery disease undergoing cerebral DSAfor the evaluation of the cerebral haemodynamic.Material & methods: Between May 1999 to March 2005, 224 cerebral DSA wereperformed in symptomatic patients referred to us with evidence of extracranial largevessel disease, either on Doppler, CTA or MRA. The percentage diameter stenosisfor extracranial vessel was calculated by NASCET method and for intracranial ves-sels as per WASID method. They were categorized as nonsignificant stenosis (0%to 49%), significant stenosis (50% to 99%), and total occlusion. The intracranialvessels were involved when a lesion was distal to the ophthalmic artery. For thevertebral artery, the distinction was made at the point where the artery pierced thedura at the level of foramen magnum. The intracranial extent of the stenosis wasincluded in this study up to the M2 and A2 segments in the anterior circulation andthe P1-P2 segments of the posterior cerebral artery.

Results: A total 404 lesions were present in 224 patients. Among these patients,single lesions were found in 64 (15.8%) and multiple lesions in 160 (84.2%).Of the single lesions 24(37.5%) were intracranial stenosis and 50(62.5%) wereextracranial. Lesions were located in the anterior circulation in 42 patients (65.62%)and in the posterior circulation in 22 (34.37%).Among the 340 stenoses in the 160 patients with multiple lesions, 38 (23.75%)patients had 88 lesion in the intracranial, and 92 patients (57.5%) had 186 lesionsin the extracranial vessels solely and both intra and extracranial in 30(18.75%)patients harboring 66 lesions.Overall 142(35.1%) lesions were intracranial and 262(64.9%) were extracranial.Conclusion: Intracranial atherosclerosis is common in Indian patients withatherosclerotic stenosis involving extracranial carotid artery disease.


HOMOCYSTEINE IN STROKE SUBTYPES IN THE SOUTH LONDON BLACKPOPULATIONU. Khan, C. Crossley, P. Collinson, H. MarkusSt. George’s University of London, London, United Kingdom

Introduction: Cerebral small vessel disease (SVD) is increased in black popula-tions but the underlying mechanisms are poorly understood. In Caucasians elevatedserum homocysteine (Hcy), and genetic variants causing increased Hcy levels,have been associated with SVD particularly in cases which have accompanyingleukoaraiosis where pathological studies suggest an underlying small vessel arteri-opathy and endothelial dysfunction is a proposed disease mechanism. We comparedHcy levels from black strokes enrolled in the in the prospective South LondonEthnicity and Stroke Study with community controls.Methods: Hcy, B12, folate and creatinine were measured in black strokes (N=483)from three South London hospitals and controls recruited by random sampling offamily practices (N=276). All patients were subtyped by one rater using modifiedTOAST criteria based on investigations. In SVD patients, leukoaraiosis was gradedaccording to severity (modified Fazekas scale) and patients divided into two groups:isolated lacunar infarction (ILI) and lacunar infarction in the presence of confluentleukoaraiosis (ischaemic leukoaraiosis (ILA)).Results: Hcy (μmol/L) was increased in black stroke patients (14.22 (8.80)) vs.controls (11.13 (5.34)) (OR: 4.63 (2.50-8.58), P<0.001) after adjusting for age,gender, vascular risk factors, B12, folate and creatinine. Hcy levels were signif-icantly raised in the following groups: SVD (OR: 7.16 (3.32-15.44), P<0.001),intracerebral haemorrhage (OR: 5.53 (2.07-14.72), P=0.001), cardioembolism(OR: 7.85 (2.87-21.42), P<0.001) and unknown aetiology (OR: 2.28 (1.15-4.52),P=0.018) with highest levels seen in SVD (16.19 (11.48)). Within the SVD groupHcy was higher in the ILA subgroup (19.63 (14.61)) vs. ILI (13.41 (7.06)) (OR:4.75 (1.53-14.69), P=0.007) and correlated with radiological leukoaraiosis severity(R=0.265, P=0.001).Conclusions: Hcy is a risk factor for cerebral SVD in blacks, especially in confluentleukoaraiosis. This is consistent with a role for endothelial dysfunction in SVDpathogenesis in this ethnic group.


DETECTION OF RIGHT-TO-LEFT SHUNTS IN PATIENTS WITHHEPATOPULMONARY SYNDROME IS POSSIBLE USING TRANSCRANIALDOPPLERG.S. Silva, D.L. Gomes, M.G. Vasconcelos, J.A. Fiorot Jr, M.M. Alves,C.H. Fischer, A.R. MassaroUNIFESP - Universidade Federal de São Paulo, São Paulo, Brazil

Background: Patients with liver failure can present with a large spectrum of neu-rologic symptoms. Cerebrovascular disease has been described as one of the mostfrequent neuropathology findings after liver transplantation. Right-to-left shunts(RLS) were described in patients with hepatopulmonary syndrome, related to in-trapulmonary vascular dilatations. We assessed the hypothesis that intrapulmonaryRLS in patients with hepatopulmonary syndrome can be detected by transcranialDoppler (TCD).Methods: Patients with liver failure and hepatopulmonary syndrome, selectedfrom the gastroenterology outpatient clinic of our hospital, and with a confirmedintrapulmonary RLS on transesophagealechocardiography (TEE) were evaluated.A group of patients with intracardiac RLS on TEE was also studied. All patientswere submitted to middle cerebral artery TCD monitoring by transtemporal ap-proach after the injection of saline solution (9ml) and air agitated, in the rightantecubital vein. Monitoring was performed during normal breathing and afterValsalva maneuver. The presence of microembolic signs (MES) was evaluated bythree examiners, blinded to the patient‘s diagnosis.

Poster SessionEtiology of stroke


Results: Eight patients with hepatopulmonary syndrome (mean age 53,2 ± 12,3years, 6 men) and 20 patients with intracardiac shunts (mean age 47 ± 16,8 years,8 men) were evaluated. MES were detected by TCD with a sensitivity of 87,5%in patients with intrapulmonary shunts and 90% in those with intracardiac shunts.Time delay to first MES detection was higher in patients with intrapulmonaryshunts (15s X 10 s in patients with intracardiac shunts) (p=0,04).Discussion: TCD can detect intrapulmonary shunts in patients with hepatopul-monary syndrome, with a sensitivity slightly lower than for the detection ofintracardiac shunts. Time delay for intrapulmonary shunts detection is higher whencompared to intracardiac shunts. Paradoxical emboli should be investigated as apotential cause of neurologic symptoms in patients with liver failure and TCD canbe a useful screening method for this evaluation.


RELATIONSHIP OF CAROTID ATHEROSCLEROSIS WITHLEUKOARAIOSIS IN ISCHEMIC STROKE PATIENTSE. Ben-Assayag, M. Mijajlovic, S. Shenhar-Tsarfaty, I. Bova, L. Shopin,S. Berliner, I. Shapira, N.M. BornsteinTel Aviv Sourasky Medical Center, Tel Aviv, Israel

Background: Previous studies have shown that white matter lesions are associatedwith increasing age, hypertension, diabetes and history of stroke. Although severallines of evidence suggest a role of atherosclerotic processes in atherothromboticvascular events, their involvement in leukoaraiosis (LA) remains to be determine.Our study examines the association between atherosclerosis, reflected as intima-media thickness (IMT) and carotid plaques and LA in a group of ischemic strokepatients.Methods: One hundred sixty four consecutive ischemic stroke patients wereincluded (mean age 66.7± 3.4 years, 61% males). All patients underwent braincomputed tomography (CT) and carotid dupplex with measurements of IMT in thecommon carotid artery. The extent and number of white matter lesions (WML)were recorded by 2 independent readers.Results: Seventy two patients (44%) were found to have 1 or more WML on CTimages located in frontal, parietal or occipital region. Of whom, 30% were recordedto have advanced LA. Mean IMT was significantly higher in stroke patients withLA (p=0.004) compared to those without it. Also, carotid plaque occurrence wasassociated with LA (χ2=6.154, p=0.013) and advanced LA (χ2=7.673, p=0.006).In logistic regression analysis, including age, gender, body mass index, and allvascular risk factors, LA was found to be associated with age and IMT (O.R.1.041, 95% CI 1.011-1.072, p=0.007; O.R. 2.365, 95% CI 1.129-4.954, p=0.022;respectively). White matter lesion severity was also found to be associated with ageand IMT (O.R.1.064, 95% CI 1.028-1.1, p=0.001; O.R. 2.84, 95% CI 1.248-6.462,p=0.013; respectively).Discussion: Stroke patients with LA present strong relationship with increasedcarotid IMT and plaque occurrence. Association was independent of gender, bodymass index and all vascular risk factors. This suggests that advanced atheroscleroticprocess in LA.


FIBRO MUSCULAR DYSPLASIA AND CERVICAL ARTERY DISSECTIONS:VALUE OF RENAL ARTERY ANGIOGRAPHYJ.M. de Bray, A. Pasco, F. Dubas, B. Vielle, J.F. SubraUniversity Hospital, France

Fibro muscular dysplasia (FMD) is a potential cause of cervical artery dis-section(CAD).Moderate forms of FMD are undiagnosed by magnetic resonanceangiography.The use of renal intra-arterial digital subtraction angiography(DSA)inidentifying FMD in CAD has not yet been validated. An ancillary study from aprospective study was performed to determine the prevalence of the associationof cervical and renal artery FMD in CAD assessed by DSA and to define thediagnostic impact of renal DSA in these patients.Methods: A prospective study on symptomatic recurrence of CAD was performedfrom 1994 to 2004 and is in press in Cerebro.Vasc.Dis.103 patients were con-secutively included for a CAD diagnosed by cervical MRI or suggested by 2concordant cervical imaging methods.FMD(17 patients)was defined as a string ofbeads image located in a non dissected vessel.The design of the ancillary studyconsisted in including patients with CAD investigated by both cervical and renalartery DSA.Population:54 patients fulfilled our subgroup criteria were 31 womenand 23 men.The Chi square test and Fisher’s exact test were used for assessing theassociation between renal FMD and vascular risk factors.Results: According to the presence of FMD,4 sub-groups of patients were iden-tified.I (n:4),showed renal FMD but no cervical FMD;II(n:5), had cervical and

renal FMD;III (n:12), isolated cervical FMD;IV (n:33) without FMD. Renal FMD,bilateral in half of the cases,was significantly associated with an age over 40years.Two out of patients with renal FMD had arterial hypertension.Conclusion: The prevalence of combined cervical and renal artery FMD is 9% inpatients with CAD.More FMD cases(7%)are detected by cervical and renal DSAthan by cervical DSA alone.The diagnosis of renal FMD could help in identify-ing patients at risk of renal artery dissection and renovascular hypertension.Thepresence of an arterial hypertension is not predictive for renal FMD.


CAUSES HETEROGENEITY OF LACUNAR STROKEO. Vinogradov, A. KuznetsovNational Pirogov Centre of Therapy and Surgery, Moscow, Russian Federation

Background: Cerebral microangiopathy owing to hypertension or diabetes mellitusis considered to be main cause of lacunar stroke (LS). But other causes of LS areknown too. The purpose of this study was to determine etiologies of LS other thansmall-artery disease.Material and methods: We have examined 105 patients with acute LS. We used:diffusion-weighted MRI (Giroscan INTERA NOVA, Holland), transthoracic ortransesophageal echocardiography, carotid duplex sonography (VIVID 7, USA),transcranial Doppler sonography with microemboli detection (Sonomed-300, Rus-sia); scale evaluation was performed according to NIHSS.Results: Patients with LS according to diffusion MRI were divided in 3 groups:group I – single small (less than 15 mm) lacunar focus – 59 patients (56,5%); groupII – large focus (more than 15 mm) or multiple small foci at the same vascularterritory – 24 patients (22,9%), group III – multiple foci in different vascularterritory – 22 patients (20,9%). LS caused by small-artery disease was revealedin 69 patients (65,7%). Stroke mechanism different from small-artery disease wasrevealed in 36 patients (34,3%). Significant differences in potential sources ofcerebral embolism were revealed in group I in comparison with group II (p<0,05)and group III (p<0,001). Multiple lacunas or combination of lacunas and corticalstrokes is reliable cerebral embolism marker. Neurological deficiency was moresevere in LS patients with embolism (p<0,01).Conclusions: Causes of LS are heterogeneous. Choice of secondary preventionregime demands of determination true cause of LS.

Stroke and lipids

1 Stroke and lipids

DEPLETION OF THE LIPID RAFT COMPONENTS CHOLESTEROL ANDSPHINGOMYELIN PREVENTS NMDA-INDUCED NEURONAL DEATHJ. Ponce, N. Perez de la Ossa, O. Hurtado, M. Dolade, M. Millan, J. Arenillas,A. Davalos, T. GasullFundacio Institut d’Investigacio en Ciencies de la Salut Germans Trias i Pujol,Badalona, Spain

Background and purpose: Excess brain extracellular glutamate in cerebral is-chemia leads to neuronal death through overactivation of N-methyl-D-aspartate(NMDA) receptors. The cholesterol lowering-drugs statins have been reported toprotect from NMDA-induced neuronal death but, so far, the mechanism underlyingthis protection is unknown. Since NMDA receptors have been reported to beassociated with the cholesterol- and sphingomyelin-rich membrane domains knownas lipid rafts, we have investigated the effect of treatments that deplete cholesterolor sphingomyelin levels on NMDA-induced neurotoxicity. In addition, we haveinvestigated the effect of simvastatin on the percentage of NMDA receptorsassociated to lipid rafts.Methods: Primary neuronal cultures were pre-treated with simvastatin, the inhibitorof cholesterol synthesis AY9944, or the inhibitor of sphingomyelin synthesis fu-monisin B1. Cell death was determined 24 h after the addition of NMDA. Lipidrafts from control and simvastatin-treated neurons were isolated, and Western blotswere performed using an antibody specific for the subunit 1 of NMDA receptors(NMDAR1).Results: Sustained treatment with either simvastatin, AY9944, or fumonisin B1,protected neurons from NMDA-induced neuronal death by 70%, 56% and 30%,respectively. Simvastatin (250 nM) reduced by 40% the association of NMDAR1to lipid rafts and did not change total expression of NMDAR1.Discussion: The inhibition of the synthesis of main components of lipid raftsprotects from NMDA-induced neuronal damage. This protection might be mediatedby a reduced association of NMDA receptors to lipid rafts. Taken together these


findings suggest that lipid raft integrity is necessary for signalling to death throughNMDA receptors.

2 Stroke and lipids

STATIN USE, LIPID PROFILE AND SYMPTOMATIC INTRACEREBRALHAEMORRHAGE FOLLOWING IV THROMBOLYSISM.W. Koch, M. Uyttenboogaart, P.C. Vroomen, J. De Keyser, G.J. LuijckxUniversity Medical Centre Groningen, Groningen, The Netherlands

Background: Intravenous thrombolysis with tissue plasminogen activator (tPA)improves outcome of acute ischaemic stroke, but increases the risk of symptomaticintracerebral haemorrhage (sICH). It has been suggested that lower cholesterollevels may be associated with higher incidence of primary ICH as well as withsICH following tPA treatment.Methods: From a prospective hospital based tPA registry comprising 309 patients,we selected all patients with known serum concentrations of total cholesterol (TC),HDL, LDL and triglycerides (TG), modified Rankin Scale (mRS) scores at threemonths and statin use (n=214). Lipid profiles were compared between patients withand without sICH following tPA treatment and between patients with and withoutfavourable outcome (mRS 0-2).Results: Patients with sICH (n=10) had significantly higher TG (2.6 SD 1.9 vs 1.8SD 1.0 mmol/l, p=0.049) and lower HDL levels (1.0 SD 0.3 vs 1.2 SD 0.4 mmol/l,p=0.049) Neither LDL (3.0 SD 0.8 vs 3.2 SD 1.0 mmol/l, p=0.59), TC (5.0 SD1.0 vs 5.1 mmol/l SD 1.2, p=0.78) levels nor statin use (2 vs 30 patients, p=0.65)were significantly different between patients with and without sICH. There were nosignificant differences for any of these variables between patients with favourableand unfavourable outcome: TC (5.2 SD 1.3 vs 5.0 SD 1.1), HDL (1.2 SD 0.4 vs1.2 SD 0.4), LDL (3.2 SD 1.1 vs 3.1 SD 0.9), TG (1.9 SD 1.2 vs 1.7 SD 1.0), statinuse (15 vs 17 patients).Conclusion: Although patients with sICH had marginally higher TG and lowerHDL levels, statin treatment or overall lipid profile seem unrelated to functionaloutcome at 3 months.

3 Stroke and lipids

HIGH TRIGLYCERIDE LEVELS IN SINGAPOREAN ACUTE ISCHEMICSTROKE PATIENTS REDUCES THE RISK OF POOR OUTCOME AT 1 YEARAFTER STROKEJ.L. Pascual, H.M. Chang, M.C. Wong, C.P. ChenNational Neuroscience Institute, Singapore General Hospital Campus,Singapore, Singapore

Background: Cholesterol reduction lowers risk for recurrent vascular events instroke patients. However, low total cholesterol (TC) and low triglyceride (TG) levelshave been linked to increased risk for hemorrhagic stroke, and poorer outcomesafter stroke, respectively. We investigated the effect of lipid-lowering therapy onthe outcomes of Singaporean acute ischemic stroke patients.Method: Consecutive acute ischemic stroke patients had fasting TC, high-densitylipoprotein(HDL), low-density lipoprotein (LDL) and TG determinations. Elevatedlipids were defined as follows: TC ≥ 5.2 mmol/L, LDL ≥ 2.6 mmol/L, and TG≥ 1.7 mmol/L. Prior lipid-lowering therapy was documented. At 1 year follow-up,functional outcome was assessed using the modified Rankin score (MRS).Results: 805 patients gave informed consent for fasting lipid profiles and 1 yearfollow-up. The mean TC was 5.66 mmol/L, mean LDL was 1.31 mmol/L, andmean TG was 1.73 mmol/L. Patients with prior stroke, ischemic heart disease orlipid therapy had lower TC and LDL levels (p < 0.05). At 1 year after stroke, 28%of patients had mRS 3 or worse. Elevated TG independently predicted for goodfunctional outcome (HR = 0.41, 95% CI: 0.23 – 0.75).Discussions: In agreement with previous studies, high TG is associated with betteroutcomes after stroke. Whether elevated TG is itself protective or is associated witha higher probability of receiving lipid-lowering drugs remains to be elucidated.

4 Stroke and lipids

ROLE OF STATINS IN FUNCTIONAL OUTCOME FOLLOWING AN ACUTEISCHEMIC STROKEL. Vaidyanathan, G. Kumar, D. Nash, W.W. Decker, L.G. SteadMayo Clinic, Rochester, USA

Hypothesis: Similar to the beneficial effect in acute coronary syndrome, dailystatin improves functional outcome following acute ischemic stroke by mechanismsother than lowering LDL levels.

Methods: The cohort included all patients presenting to the Emergency Departmentwith an acute ischemic stroke over a 22-month period (March 04 to December05). The lipid profile (cholesterol, triglyceride, LDL and HDL levels) measuredprior to or following admission was abstracted. Measurement of the panel rangedbetween 15 days prior to the stroke to 17 days after the event (Mean 0.79 days ±SD 2.53). Daily statin intake prior to the event and prescription on discharge wasalso recorded. The primary outcome variable, functional disability, was determinedusing the modified Rankin scale (mRs, 0-2=good outcome, 3-6=bad outcome),which was calculated for each patient at the time of discharge. A Pearson’schi-square test was performed analyzing the relationship between the functionaloutcomes at discharge and statin intake at the time of the event.Results: Of 508 patients, 207 (40% female) presented with an LDL of ≤100mg/dL. This group was divided into those who were on a statin on admission(n=100) and those who were not (n=107). There was no significant difference inthe admission stroke severity measured by the NIHSS (p=0.18), age (p=0.31) andgender (p=0.06) between the 2 groups. The patients on a statin were significantlymore likely to have a good functional outcome, (OR 0.5; 95% CI 0.29-0.95;p=0.033). Following adjustment for age and NIHSS, statin intake still predicted abetter functional outcome (p<0.0001).Conclusion: Daily statin intake appears to be associated with a better functionaloutcome following an acute ischemic stroke despite ideal LDL levels (≤100)and similar stroke severity on admission. Other mechanisms of action of statinslike plaque stabilization, improved endothelial cell function, anti-inflammatory,antiplatelet, anti-oxidant and antithrombotic effects may play a role in a betterfunctional outcome.

5 Stroke and lipids

SERUM CHOLESTEROL LEVELS AND SHORT TIME OUTCOME INSTROKE PATIENTSP. Milia, M. Paciaroni, V. Caso, S. Biagini, M. Venti, A. Billeci, F. Palmerini,A. Alberti, A. Baldi, G. AgnelliUniversity of Perugia, Perugia, Italy

Background: Although cholesterol and stroke disorders has been extensivelystudied, the relationship between serum cholesterol levels and short time outcomeafter stroke has not been widely investigated.Objective: To identify if serum cholesterol levels measured at admission afterstroke have any prognostic value on outcome at discharge.Methods: Patients consecutively admitted to our stroke unit suffering of any typeof stroke were analysed. Fasting serum cholesterol was measured at 24 hours afteradmission. Outcome was evaluated using mRS: we identified adverse outcome asmRS ≥3 at discharge. Data were analysed by univariate and logistic regressionanalysis.Results: We collected 935 patients suffering of stroke (mean age 74.65±11; 789ischemic, 146 hemorrhagic; mean NIHSS 8.67±6.5). Mean values of cholesterolat admission was 195.2±51 in the overall group with no differences between alltypes of stroke (I 195.65±50; H 192.43±38). Functional outcome at discharge(mRS&≥3) was not influenced by levels of cholesterol in all the population afterlogistic regression analysis (OR 0.99 CI 0.99-1.0, p 0.4) and also in either ischemic(OR 0.99 CI 0.99-1.0 p 0.7) and hemorrhagic strokes (OR 0.99 CI 0.98-1.0, p 0.2).Conclusions: Serum levels of cholesterol are not associated with outcome in theearly phase after ischemic and/or hemorrhagic stroke, suggesting that it doesn’tneed to be treated in the acute phase as negative prognostic risk factor. Still remainsuncertainty about its role at early and medium time on stroke patients.

Acute stroke: clinical patterns and practise

1 Acute stroke: clinical patterns and practise

EVALUATING THE USE OF HAND MITTENS IN POST STROKE PATIENTSWHO DO NOT TOLERATE NASO-GASTRIC FEEDINGY.-Y.K. Kee, W. Brooks, R. Dhami, A. BhallaEpsom and St. Helier University Hospitals NHS Trust, United Kingdom

Background: Early naso-gastric (NG) and consistent feeding in acute strokepatients has been shown to improve patient morbidity and mortality. However, afteran acute stroke, patients can be agitated and may frequently pull out NG tubes.Recurrent NG tubes placements are associated with complications such as traumaand chest infections. The use of restraints such as hand mittens may improvenutrition and reduce complications of NG placements. This practice althoughuncommon in the UK, is used commonly in other countries. Few studies have been

Poster SessionAcute stroke: clinical patterns and practise


done to evaluate the use of such restraints in acute stroke patients. This study aimsto evaluate the use of hand mittens in such patients.Methods: We carried out a retrospective case control study with 18 patients overa period of one year period (8 with, 10 without mittens). The following data wascollected: number of NG tubes inserted during the patient’s admission, number ofaspiration pneumonias treated with antibiotics, number of chest x-rays the patientreceived, amount of feed received, weight loss/gain during admission, length ofstay. Results were analysed using GraphPad Prism 4 softwareResults: The median age in both groups was 81.5 years. 89% of patients had totalanterior circulation stroke. Patients in the mitten group needed fewer ng tubes; 7vs 10 p<0.05. There were less episodes of aspiration pneumonia, p<0.05. Patientswith mittens experienced less weight loss; 0.7kg vs 5.2kg, p<0.05. There werefewer deaths in the mitten group, 2 vs 7, p<0.05. There was a small reduction inthe length of stay 40 vs 48.5 days, p=0.15.Discussion: The use of physical restraint is not universally accepted in the UK. Thedecision to use mittens in our patients was taken after discussions with the patientor next of kin. The use of mittens resulted in better nutrition in our patients, as wellas a reduction in mortality and should be considered in patients who do not toleratetheir NG tubes.


HEMOGLOBIN: A PREDICTOR OF FUNCTIONAL OUTCOME FOLLOWINGAN ACUTE ISCHEMIC STROKEL. Vaidyanathan, D. Nash, M.F. Bellolio, S. Enduri, S. Mishra, R. Kashyap,R.D. Brown, W.W. Decker, L.S. SteadMayo Clinic College of Medicine, Rochester, USA

Aim: To assess if the hemoglobin levels measured at the time of presentation tothe Emergency Department in a patient with Acute Ischemic Stroke (AIS) wouldpredict the severity and functional outcome.Methods: The cohort included all patients presenting to the Emergency Departmentwith an AIS over a 3.5 year period (from December 2001 through June 2005).Hemoglobin levels measured at the time of admission was recorded. Stroke severityon presentation was assessed retrospectively using the 42-point NIHSS scoringsystem and the functional disability was scored at discharge from the stroke serviceusing the modified Rankin scale (mRs). A statistical analysis of the data wasconducted using the JMP statistical software. Analysis of variance (ANOVA) wasused to analyze the variables.Results: Of the total cohort, (n=1018), 47.2% were female. The mean age was 72.3years ±SD 14.7. Hemoglobin levels were documented in 96.2% (n=979) of thepatients. Lower levels of hemoglobin predicted a more severe stroke (p=0.0067) andpoorer functional outcome (p<0.0001). This significance was retained followingadjustment for age in men (RANKIN p<0.0001 and NIHSS p=0.0004) and women(RANKIN p<0.0001 and NIHSS p=0.0004).Conclusion: Lower hemoglobin levels measured at the time of admission to theEmergency Department seem to predict more severe strokes with poorer functionaloutcome at discharge regardless of the gender probably due to greater ischemiaresulting from the decreased oxygen carrying capacity of the blood. With earlyrecognition, active methods could be taken to raise the patient’s hemoglobin and,thereby, possibly improve functional outcome.


PURE MIDBRAIN INFARCTION: CLINICAL RADIO ANATOMYCORRELATIONSR. Allibert, F. Vuillier, L. Tatu, T. MoulinUniversity Hospital Besançon, Besançon, France

Background: Clinical anatomical correlation in pure midbrain infarct has not beenwidely established in a standardised manner. The aim of the study was to analysepatients with pure midbrain infarcts using standardised protocols (clinical andimaging analyses).Method: Using a previously published arterial territories mapping, a correlationbetween the clinical patterns, arterial territories and anatomical structures involvedwas established. Patients with pure midbrain infarct were selected from all patientsadmitted with first-ever stroke in our stroke unit. Clinical and imaging data werestandardised according to Besançon stroke registry criteria. All MRI images wereretrospectively reviewed by 2 neurologists to confirm infarct location, vasculararterial territories and brain stem structure.Results: 14 pure midbrain infarcts were included. Clinical manifestations in-cluded gait ataxia in 2/14 patients, dysarthria in 4/14, limb ataxia in 10/14, sensorysymptoms in 1/14, third nerve palsy in 9/14, definitive limb weakness in 6/14 and in-ternuclear ophthalmoplegia. In most cases (9/14), the infarct was located in the high

level of the midbrain. Only 1 arterial territory was involved in 7/14 cases and whichcorresponded to the antero-medial territory in 6/7 cases. Antero-medial infarctswere always present in patients with diplopia and controlateral cerebellar ataxia.Among patients with III nerve palsy, infarct affected the nuclear (2/3) or fascicularfibers (1/3). Motor deficit was associated with anterolateral infarct (5/6 patients).Discussion: The link between diploplia and controlateral cerebellar ataxia seemedto relate specifically to a midbrain infarct located in the antero-medial territory andwith no predictive value for an antero-posterier extension of the infarct. A brachio-facial motor deficit clearly implied the involvement of the anterolateral territory, asexpected by the somatotopy of the corticospinal tract. The use of a practical toolto determine location of pure midbrain infarct enables the establishment of a goodcorrelation.


PURE DYSARTHRIA AND DYSARTHRIA-FACIAL PARESIS SYNDROMEK. Yonemura, Y. Hashimoto, M. UchinoKumamoto City Hospital, Kumamoto, Japan

Background: Pure dysarthria (PD) and dysarthria-facial paresis (DFP) are rarelyencountered clinical stroke syndromes, and seem to be regarded as an atypicallacunar syndrome. We sought to clarify the clinical characteristics of PD and DFPin acute ischemic stroke patients.Methods: We selected patients with PD or DFP from 1,043 consecutive patientswith first-ever acute ischemic stroke admitted over a 7-year period. Vascular riskfactors, emboligenic heart disease, occlusive cerebral artery disease, infarct sizeand location, stroke subtype according to the TOAST classification, and outcomedata were evaluated. Acute infarcts were all identified by diffusion-weighted MRI(DWI).Results: A total of 34 patients were reviewed. They consisted of 16 patients withPD and 18 patients with DFP. Hypertension was the most frequent in the vascularrisk factors (13 patients with PD and 14 patients with DFP). Emboligenic heartdisease and/or cerebral artery disease were detected in 13 (38%) patients (5 with PDand 8 with DFP). DWI identified infarcts on the cortical motor area, centrum ovale,or corona radiata in 25 (74%) patients (12 with PD and 13 with DFP), whereasbrainstem infarct was demonstrated only in 1 patients with PD. 18 (53%) patients(11 with PD and 7 with DFP) had infarcts <15mm in maximal diameter, and 15(44%) patients (10 with PD and 5 with DFP) met the diagnostic criteria of lacunarstroke. Neurological deterioration occurred after hospitalization in 2 patients withDFP, but patients were discharged usually with mild or no disability.Discussions: Infarcts in the higher levels of cerebral hemisphere are responsiblefor the majority of the PD and DFP syndromes. The predictive value of thesesyndromes for lacunar stroke is rather low, particularly in DFP.


ACUTE VERTIGO OF UNDETERMINED ORIGIN: DIAGNOSTIC VALUE OFMAGNETIC RESONANCE IMAGINGL. Huang, A. Villringer, A. HartmannCharité Campus Benjamin Franklin, Berlin, Germany

Background: The origin of acute vertigo often remains undetermined after neuro-logical, otological, and CT examination in the emergency room. We investigatedthe diagnostic yield of magnetic resonance imaging (MRI) in these patients.Methods: Patients were included in the study if they had presented with suddenonset of vertigo to our Department between 01/2002 and 06/2005, and completeneurological, otological and cranial CT investigation allowed no definite allocationto peripheral or central origin of the vertigo. Results of cranial MRI includingdiffusion-weighted imaging (DWI), clincal, and epidemiological information weretaken to compare the patients with (group 1) and those without (group 2) acutelesions on MRI using univariate statistics.Results: In the 108 patients with acute vertigo (mean age 61 years, 62% women),acute ischemic lesions were detected in 12% on DWI. Affected regions weremedulla oblongata, cerebellum, pons, thalamus, corpus callosum, temporo-occipitallobe, parietal lobe, both periventricular areas, and frontal lobe. Higher prevalence ofvertigo-unrelated ischemic CT changes (p=0.01) and non-significant trends towardselevated serum cholesterol (p=0.06), older age (0.07) and higher blood glucoselevels (0.09) were observed in group 1 compared with group 2. All other clinicaland epidemiolgical variables were similar in both groups.Conclusion: In patients with acute vertigo and inconclusive clinical and CTexaminations, the diagnostic yield of cranial MRI is low. Patients with old ischemiclesions on CT are more likely to have a central vertigo origin. Clinical andepidemiological characteristics are not associated with higher detection rates ofacute brain lesions.



COURSE OF CEREBROVASCULAR INCIDENTS IN PATIENTS WITHPERMANENT AND PAROXYSMAL ATRIAL FIBRILLATIONJ. Staszewski, J. Kotowicz, A. StepienMilitary Medical Institute, Warsaw, Poland

Permanent atrial fibrillation (pAF) is an important risk factor for cerebrovascularincidents (CVI). The frequency and prognosis of CVI in patients with paroxysmalAF (plAF) is not well known. The aim of this single center study was to assess thecourse and risk factors of unfavorable outcome of CVI in ptAF and plAF.Patients with established AF and acute, ischemic CVI were included to prospec-tive study with 6-months follow-up. The patients were classified on admission,discharge and after 6-month as non-dependent (mRS ≤ 2) or dependent (mRS≥ 3 ≤ 5). The course of CVI was categorized as favorable (non-dependence) orunfavorable (dependence or death). The risk of embolic complication in AF wasassessed in CHADS2 scale.Of 178 included patients 70 (39%) had plAF and 108 (61%) had ptAF. Groupswere similar in vascular risk profile. Cardioembolic strokes were more common inptAF than in plAF group (56% vs 44%, p=0.03). The proportions of dependentpatients at admission, discharge and at 6-month visit in ptAF were higher thanin plAF (50% vs 19%, 81% vs 45%, 71% vs 40%; p<0.001). In-hospital and6-month mortality rates were higher in ptAF than in ptAF group (13% vs 3% and32% vs 13%, p<0.001). Unfavorable risk factors for short- and long-term outcomewere: ptAF (OR 2.05, p=0.01), lack of chronic anticoagulation treatment (OR3.48, p=0.02), chronic heart failure (OR 2.54, p=0.05), diminished left ventricularejection fraction (LVEF) (OR 2.15, p=0.01), non-lacunar cerebral infarction (OR1.74, p=0.001) and increased (moderate to very high) risk of embolic complications(OR 1.92, p=0.004).Results suggest significantly different course of CVI in patients with ptAF andplAF. Those with plAF had more favorable short- and long-term prognosis of CVI.Risk factors for unfavorable stroke outcome were: ptAF, chronic heart failure, lackof chronic anticoagulation, diminished LVEF, non-lacunar cerebral infarction andpresence of increased risk of embolic complications.


COMPARISON BETWEEN THE OCSP AND TOAST CLASSIFICATIONSYSTEMS IN THE DIAGNOSIS OF CEREBRAL SMALL VESSEL DISEASE INTHE SOUTH LONDON BLACK AND WHITE POPULATIONSU. Khan, P. Jerrard-Dunne, L. Kalra, A. Rudd, C. Wolfe, H. MarkusSt. George’s University of London, London, United Kingdom

Background: Stroke classification systems based on clinical criteria, such as theOxfordshire Community Stroke Classification (OSCP), have been widely used inepidemiological studies to diagnose lacunar stroke but may be inaccurate comparedwith systems based on investigation results. We compared OSCP with a pathophys-iological classification (modified TOAST) in diagnosis of lacunar stroke in blacksand whites in the South London Ethnicity and Stroke Study.Methods: African and African-Caribbean strokes presenting to three South Londonhospitals were prospectively recruited (N=600). 600 consecutive Caucasian strokespresenting to the same three centres were also recruited. All cases underwentstandardised clinical assessment, demographic and risk factor data collection andinvestigations (brain imaging (CT 65.4%, MRI 8.3%, both CT and MRI 26.3%),imaging of the extracranial cerebral vessels (97%), echocardiography (56.7%)).All cases were subtyped using modified TOAST criteria (excluding the use ofhypertension as a criterion for diagnosis) by one observer with review of originalimaging. Cases were also subtyped using the OCSP classification.Results: Using TOAST, lacunar stroke was more prevalent in blacks (OR 2.94(1.97-4.39)p<0.001) compared to whites. Similarly, OCSP-defined lacunar infarction wasincreased in blacks but the association was weaker (OR 1.94(1.39-2.73)p<0.001).Taking TOAST classification as the gold standard the sensitivity of OCSP fordetection of lacunar stroke was 84% (76% specificity). Positive predictive value(PPV) for lacunar stroke detection was 56.5% (negative predictive value (NPV):92.8%). Accuracy of OSCP for lacunar stroke diagnosis was worse in blackpatients: PPV 66.7% (NPV: 88.6%) compared to a PPV of 41.4% (NPV: 96.0%) inwhites.Conclusions: Lacunar stroke is increased in blacks compared to whites but useof OCSP underestimates the difference. OCSP is less accurate at lacunar strokediagnosis in blacks, and this may introduce bias in studies comparing subtypedifferences between ethnic groups if OSCP is used.


ISCHEMIC STROKE REVEALING SMALL INTRACRANIAL ANEURYSM.NATURAL HISTORY, MANAGEMENT AND RECOMMENDATIONH. Desal, B. Daumas-Duport, F. Herisson, E. Auffray-Calvier, B. GuillonHopital Laennec, Centre Hospitalier Universitaire de Nantes, Nantes, France

Background: Ischemic stroke may be the first manifestation of small intracranialaneurysm, secondary to clot embolization from the aneurysmal sac. Pathophys-iology, clinical characteristics and outcome are not clearly identified leading toundetermined management.Methods: Patients admitted over a 6-year period in our stroke unit with ischemicstroke distal to small (< 25 mm) sacciform intracranial aneurysm, in the ab-sence of other causes for the infarctions, were selected. Patients demographics,characteristics of aneurysms, outcome and management were analysed.Results: Eight patients fulfilled our selection criteria (5 women; mean age 50, range38-58). Ischemic stroke and intracranial aneurysm involved the anterior circulationin 3 patients and the vertebrobasilar system in 5. The mean size of the aneurysmswas 11 mm (range 3-18). Digital angiography showed partial or complete aneurysmthrombosis in 4 patients with occlusion of the parent artery in 2. An unexpectedsubarachnoid haemorrhage was diagnosed in 3 patients. Two patients died duringthe 72-hours period following their admission because of severe subarachnoidhaemorrhage. In the other cases, prognosis was excellent after early endovascularembolisation (3 patients) or spontaneous thrombosis (3 patients) of the aneurysm.However, in these last 3 patients, 2 had a late partial recanalisation of the aneurysmthat justified endovascular treatment.Discussion: Our results suggest that ischemic strokes revealing intracranialaneurysm might be associated with an asymptomatic subarachnoid haemorrhage,that should be ruled out by CSF analysis (with spectrophotometry). Antithromboticsshould also be used cautiously in these high risk patients for haemorrhage. Earlyendovascular or surgical treatment could prevent subsequent subarachnoid haem-orrhage or stroke recurrence. Radiological follow up is required to detect furtherrecanalisation, particularly in cases with early spontaneous aneurysm thrombosis.


CEREBRAL MICROEMBOUS DETECTION IN PATIENTS WITH ACUTEISCHEMIC STROKEJ. Lee, S.J. LeeYeungnam University College of Medicine, South Korea

Background: Microembolic signals(MES) detected by transcranial Doppler(TCD)have been considered as an independent predictor of recurrent ischemic stroke.However, the association between the presence of MES and the risk of stroke haspredominantly been studied on small and selected patients. To evaluate the clinicalsignificance of MES in patients with acute ischemic stroke, we investigated theprevalence of MES and analyzed the relationship between MES and stroke subtype.Methods: We intended to perform TCD monitoring for 30 minutes to detect MESfrom the bilateral middle cerebral arteries in patients within 15 days of stroke onset.The strokes were subtyped using the TOAST classification criteria.Results: Of a total of 884 consecutive ischemic stroke patients admitted to ourstroke unit within 7 days of stroke onset, TCD study performed 590 patients, and65 were excluded because there was a long interval between onset of symptomsand examination, an artificial heart valve, and inadequate temporal bone window.MES were detected in 23(4.2%) despite the fact that all patients were receivingan antiplatelet or an anticoagulant treatment. Among patients with positive MESdetection, 7(31.8%) had recurrent MES during follow-up TCD monitoring within3 days after the first examination. MES were detected in 3.1% of patients withlarge-artery atherosclerosis stroke, 4.1% of cardioembolic stroke, 2.1% of lacunarstroke, 9.1% of cryptogenic stroke, and 4.1% of undetermined stroke(p=0.241). In6 of the 8 patients(75%) with the anterior circulation infarct, MES were observedipsilateral to the affected territories.Discussion: During antithrombotic treatment in patients with acute ischemic stroke,the prevalence of MES is low and MES detection dose not improve ischemic strokesubtype classification. The MES are frequent in the territories of symptomaticarteries in the anterior circulation stroke and recurrent MES are common during ashort follow-up.




FIRST-DAY BODY TEMPERATURE DYNAMICS – A CLINICIAN’S TOOL FORMONITORING PENUMBRAL TISSUE TRANSFORMATION AND UPDATINGPROGNOSIS AFTER ISCHEMIC STROKE?J. Mau, S.D. JayavelHeinrich Heine University Hospital, Duesseldorf, Germany

Background: Early normal but later rising body temperature (BT) was recentlyshown to robustly and independently reduce the likelihood of complete dependencyor death after hemispheric ischemic stroke. To substantiate this unexpected finding,first-day BT dynamics were to be studied in data from the first European Cooper-ative Acute Stroke Study (ECASS) and explained within a model of progressivestroke.Methods: The ECASS had randomized and treated 615 patients to either 1.1mg/kgrt-PA or placebo IV within 6 hours from symptom onset; 461 patients hadcomplete 0-2-24h profiles of 3 BT measurements. BT profiles were groupedinto 37(10%) "ever constant", 34(7.4%)"early only rising", 123(26.7%) "late onlyrising", 107(23.2%) "ever rising", 49(10.6%) "decreasing", 35(7.6%) "convex",and 76(16.5%) "concave", their association with baseline covariates tested withstratified chi-squares and dichotomized 90-day modified Rankin scores (mRS)logistically regressed on known predictors and trial medication.Results: Profile groups were equally distributed between rt-PA and placebo and allbaseline covariates but for initial BT (P<0.0001), initial stroke severity (P=0.0298),hypo-density extent in infarct territory (P=0.0164), and presence of infarct signs(P=0.0033) on baseline computed tomography. Specific interactions were signif-icant for the latter (P=0.041) and in BT > 37°C (P<0.0001) and BT<36°C(P=0.0025). "Ever rising" profiles increased 90-day fatality and "ever constant"profiles reduced risk of long-term dependency (mRS 4-5) among survivors after24h, independently. Hence, associations of BT profiles were found for only thelesion parameters and for specific long-term outcomes.Discussion: A dynamic model of progressive stimulation of metabolic and apop-totic regulation after focal ischemia explains that BT dynamics appear closelycorrelated with evolving transformations in the ischemic penumbra.Conclusion: Beyond clinically updating baseline prognosis after 24 hours, moni-toring BT might also be used for timing neuro-protective interventions.


DOES ACUTE STROKE UNIT CARE CHANGE DURING A REHABILITATIONCLINCIAL TRIAL (AVERT PHASE II)?J. Collier, J. BernhardtNational Stroke Research Institute, University of Melbourne, Heidelberg West,VIC, Australia

Objectives: During A Very Early Rehabiliation Trial (AVERT) Phase II, werandomised recruited stroke unit patients from two hospitals to receive either avery early mobilisation (VEM) protocol or standard (usual) care. Both patientgroups were treated on the same ward. Given the potential for contamination effectsusing this design, we aimed to investigate whether levels of physical activity fornon-recruited patients changed as a response to the trial embedded within eachstroke unit.Methods: Prior to the trial, behavioural mapping procedures were used to evaluatethe proportion of the day patients were in bed, sat out of bed, and were standing orwalking. Behavioural mapping requires structed observation and recording of pa-tient activity over a single day. During AVERT, mapping procedures were repeatedto determine activity levels of patients receiving standard care. All patients <14days post stroke were eligible, with the exception of those requiring palliative care.We excluded clinical trial participants. Ten-minute observations were conductedbetween 0800 to 1700 with patient activity documented, who was present andwhere patients were located. Stepwise binomial logistic regression was used toassess difference in activity between time periods (2001/2; 2004/5), controlling fordifferences in patient characteristics.Results: 51 patients (mean age 74.0 years, 51% male) were recruited at AustinHealth and St. Vincents Hospital, Melbourne. Patients spent 65% of the day restingin bed and 9% of the day in standing/walking activities. This was similar tothe activity patterns of patients in the pre-trial sample (60% of day in bed, 13%standing/walking). No statistical differences between time periods was found (bed:CI 95%-1.02.0, P=0.926; stand/walk: CI 95% -6.22.1, P=0.315).Conclusion: The level of physical activity of stroke patients receiving standardstroke unit care did not change during conduct of an acute rehabilitation trial.This finding supports the feasibility of conducting an individually randomisedrehabilitation clinical trial.


ALBUMIN TO CREATININE RATIO (ACR) IS ASSOCIATED WITH THESEVERITY OF ACUTE STROKE AND PREDICTS THE OUTCOME OF ACUTESTROKE AND TRANSIENT ISCHEMIC ATTACK (TIA)K. Koniari, E. Gialouri, K. Makris, I. Drakopoulos, O. GlezakouGeneral Hospital KAT, Athens, Greece

Background: Although microalbuminuria is associated with clinical risk factorsfor stroke, there is surprising little information regarding it as an independent riskfactor for stroke or as a predictor of stroke outcome.Methods: In our study we investigated the prevalence of microalbuminuria in acutestroke patients and its association with the patients’ clinical status at admissionand outcome. We studied 60 patients (mean age 75 years) who were admittedin our hospital’s ER within 6 hours of their first neurological symptom. A urinesample was collected at the time of admission, along with morning collectionsat 24, 48, 72 hours and at the 7th day of hospitalization. ACR was measured ona POCT instrument (Bayer DCA-2000). CT-scan was performed on all patients.Neurological deficit and clinical status was assessed by the Glasgow Coma Scale(GCS) on admission and on days 1, 2, 3 and by Glasgow Outcome Score (GOS) ondischarge from the hospital.Results: The mean value of ACR, on the admission sample, in patients with severeclinical status (GCS<8) was 354 mg/g, significantly higher than in patients withbetter clinical status (GCS 9-13 and GCS 14-15) 114 and 122 mg/g respectively(p<0,01). The mean value of ACR on the last day of hospitalisation was highlycorrelated with the outcome (reflected by GOS). It was 348 mg/g in patientswith GOS=1, 209 mg/g in patients with GOS 2 -4 and 53 mg/g in patients withGOS=5 (p<0,01). In addition concerning the question if ACR levels can predictthe outcome, only in non-diabetic patients with poor outcome (GOS=1), the meanvalue of ACR of the 24 hours measurement was significantly higher than in thosewith better outcome (GOS 2-4 and GOS=5) 262, 199 and 58 mg/g respectively(p<0,01).Discussion: Our preliminary results from this ongoing study suggest that thismarker, when determined on admission and within 24 hours might be useful indetermining the severity of the stroke independently of the type of stroke, and thatthese early measurements can be of value in predicting outcome in non-diabeticpatients.


RATE OF INTRACEREBRAL HAEMORRHAGE IN PATIENTS WITH MINORSTROKE: A CLINICAL RULE TO REDUCE CT MISDIAGNOSISC.E. Lovelock, J.N. Redgrave, D. Briley, P.M. RothwellUniversity of Oxford, Oxford, United Kingdom

Background: Most studies of acute stroke management focus on the initial hoursfollowing symptom-onset. However many patients with non-disabling strokespresent late. In a recent clinic-based Scottish study of patients presenting late (>4days after the event) with minor stroke, around 4% had intracerebral haemorrhage(ICH) on MRI, 75% of which appeared as infarcts on CT brain imaging. MRI wasrecommended when patients could not be CT scanned within one week of a minorstroke. We aimed to determine the frequency of ICH in two cohorts of patients withminor stroke, and to identify clinical predictors for ICH, which might be used toprioritise patients for MRI where resources were limited.Method: We studied 343 consecutive patients with minor stroke (NIHSS≤3) inpatients ascertained in the Oxford Vascular (OXVASC) Study (scanned using CTafter a median delay of 4 days) and 245 consecutive patients presenting to ahospital-based stroke clinic, all of whom had MRI.Results: The rates of ICH were 4.1% (95%CI: 2.5-6.8%) on CT in OXVASC and4.5% (95% CI 2.5-7.9%) on MRI in the clinic cohort. In a pooled analysis (25 ICHin 588 patients), severe hypertension (BP ≥ 180/110 mmHg) on assessment (OR5.4, 95%CI 2.3-12.3, p<0.001), vomiting (OR 9.9, 3.7-26, p=0.001), confusionat onset (OR 8.2, 3.1-21.4, p=0.001), and premorbid anticoagulation (OR 6.1,2.1-17.7, p=0.01) were predictive of ICH. The 178 (30%) patients who had at leastone of these risk factors included 92% of patients with ICH.Conclusion: The rate of ICH in minor strokes in our cohorts is consistent with thatin the only previous study. Several clinical variables were predictive of ICH, and ifindependently validated could form the basis of a simple rule to select patients whorequire MRI.



MRI PERFUSION-DIFFUSION MISMATCH IN PATIENTS WITH ACUTEISCHEMIC STROKES. Di Legge, M. Diomedi, F. Sallustio, S. Napolitano, G. Koch, B. Rizzato,R. Floris, P. StanzioneUniversity of Tor Vergata, Rome, Italy

Background: The MRI perfusion-diffusion mismatch is viewed as a marker ofstill-salvageable tissue amenable to reperfusion therapies. Whether its presence isan independent predictor of clinical outcome in non-thrombolytic series has notbeen extensively investigated.Objectives: To evaluate the occurrence, evolution, and clinical correlates of MRIperfusion-diffusion abnormalities in patients with acute ischemic strokes not treatedwith thrombolytics.Methods: Patients with acute ischemic stroke (AIS) admitted to our emergencyroom (ER) within 12 hours of symptom onset were screened for MRI eligibility.The MRI protocol included DW, PW MRI and intracranial MR angiogram (MRA).All patients were admitted to a stroke unit and managed according to the currentguidelines. Stroke severity was assessed by the NIH Stroke Scale (NIHSS) score atscheduled times. Patients with and without PW-DW mismatch were compared fordemographic, clinical and imaging variables.Results: Over a 12-month period 189 patients with symptoms suggestive of AISwere seen by the stroke team. Of them, 116 (61%) had an MRI study within 12hours. Thirty (26%) patients did not complete the MRI study for lack of complianceor clinical instability. A perfusion-diffusion mismatch (M+) was detected in 29/86(34%) patients. The presence of PW-DW mismatch was associated with higherbaseline NIHSS scores (p=0.02), intracranial stenosis (p=.001), lesion enlargementon follow-up MRI (p=.001) and higher three-month mRS (0.04). At logisticregression analysis the only independent predictor of poor outcome (mRS 3-6) at 3months was onset NIHSS (OR 1.7, 95% CI 1.1-2.8; p=0.02).Conclusions: Up to one third of our AIS patients who completed the MRI protocolwithin 12 hours of stroke onset had a PW-DW mismatch. Its detection wasassociated with more severe strokes, intracranial artery occlusion, lesion growing,and worse outcome. This information may help in establishing the efficacy ofthrombolitic therapy beyond the 3-hour window based on MRI parameters.


SYSTEMIC THROMBOLYSIS WITH RT-PA IN POSTERIOR CIRCULATIONSTROKEB. Dimitrijeski, A. Villringer, H.C. Koennecke, A. HartmannCharité Campus Benjamin Franklin, Berlin, Germany

Objectives: Ischemic stroke in the posterior circulation (PCS) accounts for 10-15%of all strokes, representing a major cause for disability and death in stroke patients.Treatment with rt-PA for acute ischemic stroke within a 3-hour time window hasbeen proven to be effective and reduces significantly disability. However, most dataon systemic thrombolysis refer to stroke in the anterior circulation (ACS).We compared the clinical outcome at 3 months in patients with PCS and ACStreated with rt-PA.Methods: 242 patients were treated between 1998 and 2006 within a 3-hour timewindow according to the NINDS-trial protocol, 216 (89%) with ACS, 26 (11%)with PCS, 3 (1%) with basilar occlusion.Infarct localisation was n=11 brain stem, n=6 occipital lobe, n=2 thalamic, n=2cerebellar and n=5 combined.Neurological status was measured at admission and at 3-month follow-up using theNIH-Stroke-Scale (NIHSS) and the modified Rankin Scale.Results: A total of 26 patients suffered from PCS (42% female, Mean NIHSS atadmission 13, mean age 68y).Good functional outcome defined as Rankin ≤ 2 occurred in 16 patients (62%)with PCS compared to 107 patients (50%) with ACS (p=0.25).The mortality rate was n=4 (15%) in PCS and n=30 (13%) in ACS (p=0.51) andsymptomatic intracranial hemorrhage occurred in n=1 (4%) in PCS and in n=7(3%) in ACS (p=0.60).Conclusions: In our study population clinical outcome at 3-month follow up,mortality and intracranial hemorrhage rates are similar in patients with ACS andPCS after treatment with systemic thrombolysis.It seems to be a safe and effective treatment in patients with posterior circulationstroke.


TRANSIENT ISCHEMIC ATTACK IN ADULT ONSET MOYAMOYA DISEASEJ.-M. Kim, S.-H. LeeSeoul National University Hospital, Seoul, South Korea

Background and objective: There have been few reports about initial manifes-tations in adult onset moyamoya disease. In this study, we described the initialmanifestations of adult onset MMD including TIA symptom characteristics, andinvestigated the relationship between the initial manifestation and intracranialstenosis.Method: Between 1999 and 2006, total of 65 patients who are older than 14 yearswere admitted and diagnosed as moyamoya disease at Seoul National UniversityHospital. The patients were categorized into TIA, ischemic stroke, hemorrhagicstroke, and nonspecific group due to the initial symptoms. The outcome aftersurgical revascularization was evaluated in terms of two domains, which are thenumber of TIAs and stroke recurrence after surgery.Result: Out of 65 subjects, the numbers of patients who had initial manifestationas TIA, ischemic stroke, hemorrhagic stroke and nonspecific symptom were 29(44.6%), 11 (16.9%), 15 (23.1%), and 10 (15.4%), respectively. TIA manifestationswere variable among subjects. Twenty subjects out of 29 experienced motordominant symptoms, whereas only two had sensory dominant symptoms. Isolatedcognitive dysfunctions such as language dysfunction occurred in four subjects. Themeans of Suzuki grade in TIA group and ischemic stroke group were 2.90 ± 0.9and 3.64 ± 0.8 (p=0.022), showing significantly severe stenosis in ischemic strokegroup compared to TIA group. There was no statistically significant difference inthe surgical outcome in terms of TIA numbers and stroke recurrence between theTIA and stroke groups.Discussion: This study demonstrated that TIA is frequent as initial symptomamong adult onset MMD, and TIA manifestations are variable among subjects.Patients whose initial manifestation is TIA have lower intracranial arterial stenosisthan stroke group, implying that TIA is earlier manifestation in the disease processthan ischemic or hemorrhagic stroke.


INSULAR INVOLVEMENT IN ACUTE MIDDLE CEREBRAL ARTERYTERRITORY INFARCTIONB.G. Yoo, J.K. Kim, J.H. Ko, E.G. KimKosin University College of Medicine, Busan, South Korea

Background: The frequency and extent of insular involvement in middle cerebralartery (MCA) territory infarction and its relationship with stroke severity andclinical relevant disorders are not well established. The purpose of this study is todetermine insular involvement in MCA territory infarction and its relationship withclinical and laboratory parameters.Methods: We analyzed a total of 73 consecutive patients with acute non-lacunarMCA territory stroke proved by an MRI scans.Results: Insular involvement were present in 52 patients (73%); 34 (65%) hadmajor insular lesions and 18 (35%) had minor lesions. The major insular in-volvement was associated with elevated serum CK-MB (p=0.044) and fibrinogen(p=0.024), and size of MCA infarction (p=0.018) than minor insular lesion. Theanterior insular alone was involved in 14 (27%) patients, and the posterior insularalone was involved in 4 (7.7%) patients. Among patients with insular involvement,twenty-three patients (44%) had lenticulostriate territory infarction. Insular involve-ment was associated with larger MCA territory infarctions, more severe clinicaldeficits, and mortality. Isolated anterior insular infarcts were often accompanied byother infarcts in the superior territory, whereas posterior insular infarcts were oftenaccompanied by inferior division infarction.Conclusions: The insular involvement is a common in patients with acute non-lacunar MCA territory infarction. Major insular involvement is associated withlarge MCA territory infarction, proximal MCA occlusion, elevated serum CK-MB,and greater neurologic severity than minor and no insular infarction.


COMPARATIVE ELIGIBILITY FOR ACUTE HEMORRHAGIC ANDISCHEMIC STROKE TREATMENTS IN A DISTRICT GENERAL HOSPITALM.O. McCarron, M. Armstrong, P. McCarronAltnagelvin Hospital, Derry, United Kingdom

Background: Thrombolysis treatment fot acute ischemic stroke (AIS) benefitspatients. Recombinant activated factor VII (rFVII) is emerging as a similartime-dependent treatment for acute intracerebral hemorrhage (ICH). We sought to



determine the relative proportions and absolute numbers of patients eligible foracute stroke treatments in an Irish district general hospital.Methods: In a prospective observational study delays in admissions, demographicdetails, and stroke severity were recorded in consecutive stroke patients admitted toa district general hospital over a 12 month period. The eligibility criteria for acutetreatments were adapted from the NINDS study group for ischemic stroke and aphase two study of rFVII treatment, which excluded patients with any history ofthrombotic or occluusive disease.Results: 171 patients (96 men, 75 women, mean age 69.9±12.7 years) wereassessed. Multiple logistic regression showed that less severe strokes, living aloneand attending a general practitioner all independently delayed hospital admission.Patients with ICH arrived in hospital faster than AIS patients, p=0.03. ICH patientshad more severe strokes than AIS patients (median NIHSS 8 versus 4, p=0.006).Nineteen of 152 or 12.5% of AIS patients were in hospital within 2.5 hours ofstroke onset with a NIHSS>4 and fulfilled the NINDS thrombolysis criteria. Sevenof 19 or 37% of ICH patients were admitted within 3.5 hours and were eligible forrFVII treatment.Discussion: Although proportionately more ICH patients may potentially avail ofacute treatment than AIS patients, almost three times as many AIS patients wereeligible for acute treatment in this district general hospital. There may be lesspotential to decrease admission delays for ICH patients.


A LINK BETWEEN LUNAR PHASE AND MEDICALLY UNEXPLAINEDSTROKE SYMPTOMSF. Ahmad, T.J. Quinn, M. Walters, J. DawsonGardiner Institute of Medical and Cardiovascular Sciences, Glasgow, UnitedKingdom

Background: Possible lunar effects on health have been postulated for centuries.Association between phase of the moon and vascular; neurological and psychiatricdisease have been reported. There are no published studies on the effect of lunarphase on cerebrovascular disease. A consistent proportion of stroke unit admis-sions remain medically unexplained despite extensive investigation. This cohort ofMedically Unexplained Stroke Symptoms (MUSS) patients have been previouslydescribed and show a high rate of psychiatric comorbidity. We hypothesised thatadmission to an acute stroke unit, with true stroke or MUSS, may be influenced bylunar cycle.Methods: All admissions to our Acute Stroke Unit are recorded in a comprehensivedatabase. Those admitted between 1st January 1993 and 30th September 2006(MUS) were included in the study. The association between admission rate andphase of the moon was calculated using a X2 test across the groups. We observedadmission rate for confirmed stroke and MUSS.Results: There were 7219 admissions during the study period, which comprised167 complete lunar cycles. Stroke admissions were evenly spread throughout lunarphases (p=0.72). Admission with medically unexplained stroke-like symptoms wassignificantly increased during full moon phases (p=0.023).Discussion: There was a statistically significant association between full moonlunar phase and diagnosis of medically unexplained stroke-like symptoms. Therewas no association with other stroke diagnoses. This study adds to the growingliterature regarding lunar effects on health.


IMPLEMENTATION OF FAST-TRACK ASSESSMENT OF PATIENTS WITHTRANSIENT ISCHAEMIC ATTACKS IS MORE EFFECTIVE THAN WEEKLYONE-STOP CLINICSS. Goode, N. Altaf, J. Riley, J. Gladman, S. MacSweeneyQueens Medical Centre, United Kingdom

Introduction: The risk of stroke is highest immediately after an initial transient is-chaemic attack (TIA). Current guidelines highlight the need for the rapid assessmentof patients with TIA. There is, however, a significant delay in the assessment ofpatients with TIA in a weekly one-stop clinic. The aim of this study was to ascertainthe impact of an urgent daily TIA clinic on waits for assessment and treatment.Methods: Retrospective analysis was performed on the delay between initial TIA,referral and clinic dates. This data was collected for a 3 month period during whichthe weekly one stop clinic was used to assess patients with suspected TIA. Inaddition, data was also collected during the 3 month period during which the pilotfast-track TIA assessment as well as the weekly one-stop clinic was functional.Results: 288 patients with suspected TIA were assessed over the study period. Themean age of the patients was 68 years (SD 10) and 51% were male. The meaninterval between TIA (n= 120) and review in the weekly TIA clinic in the initial

3 month period was 66.9 [72.0] days. The mean wait for fast-track assessment ofpatients (n=62) from initial TIA to review was 4.5 [6.9] days. This decreased theoverall wait (n=168) in the 2nd 3 months to 29.6 [33.2] days, p<0.005.Conclusion: The urgent daily assessment of TIAs provides significantly fasterassessment, investigation and treatment compared with weekly one-stop clinics.This is essential to decrease the risk of recurrent stroke in patients with TIAs, inparticular those with high grade carotid stenosis that may need subsequent CarotidEndarterectomy.


ATRIAL FIBRILLATION IN ISCHEMIC STROKE PATIENTS: EVALUATIONOF THE USAGE OF ORAL ANTITHROMBOTICS IN EUROPEB.M. Hamad, P. Nasuti, E. YeungIMS Health, London, United Kingdom

Background: It is clinically known that atrial fibrillation (AF) is associated withhigher in-hospital mortality in ischemic stroke patients. The purpose of this studyis to assess the use of oral antithrombotics among patients with AF and to illustratethe diffusion of recommended guidelines into clinical practice.Methods: We conducted a patient-diary study in 296 hospitals in France, Germany,Italy, Spain and UK. They treated total of 2,659 ischemic stroke patients, duringthe period of July 2005 and December 2005, with a mean of 9 patients per hospital.Hospitals were selected to be representative by geographic regions, size and strokeunit facilities in each country.Results: Among the 2,659 patients, 478 (18%) pts were found to have chronic AFof whom 123 (26%) also developed AF acute event during the stroke hospitalisa-tion. Also, 70 (3%) patients had experienced an acute event of AF for the first time.We report data from 548 ischemic stroke patients with AF, 75% were classifiedas cardioembolic, 20% non-cardioembolic and 5% of uncertain cause. On-therapyanalysis showed that 54% of pts received aspirin, 46% oral anticoagulants (OAC),13% clopidogrel, 3% dipyridamole and 1% Aggrenox. We also analysed monoand dual therapies; 31% pts had aspirin only, 26% OAC only, 16% aspirin+OAC,6% clopidogrel only, 6% aspirin and clopidogrel only, and interestingly 10%didn’t receive any oral antithrombotic therapy. In-hospital mortality accounted for12% of the patients, 27% discharged home independently, 15% discharged homedependently and 43% transferred to rehab and long term care facilities.Conclusion: European guidelines recommend oral antithrmobotic therapy forischemic stroke associated with AF for secondary prevention and this has beenadopted widely in European hospitals treating ischemic stroke patients, howeverstill 10% of patients are not receiving any form of oral antithrombotic therapy toprevent recurrent stroke, and only 45% are receiving OAC.


TRANSIENT ISCHEMIC ATTACK IN ADULT MOYAMOYA DISEASEJ.-M. Kim, S.-H. Lee, J.-K. RohSeoul National University Hospital, Seoul, South Korea

Background: Transient ischemic attack (TIA) is a frequent initial manifestation inadult onset moyamoya disease (MMD). However, clinical characteristics of TIAin adult MMD have not been fully understood, as compared with the numerousreports on child MMD. In this study, we investigated the initial manifestations ofadult MMD including TIA symptom characteristics, and the relationship betweenthe initial manifestations and the extent of the disease progression.Method: We recruited 65 MMD patients of adult onset between 1999 and 2006,who visited and diagnosed at the Seoul National University Hospital. We obtainedtheir basic demographic data, imaging files, and clinical information. The studygroups were categorized into TIA, ischemic stroke (IS), hemorrhagic stroke (HS),and nonspecific (NS) according to the initial manifestations. Symptoms of TIAwere described via detailed interview. The outcomes after surgical revascularizationwere evaluated in terms of stroke or TIA recurrence during the follow up.Result: Out of 65 subjects, there were 29 (45%) TIA, 11 (17%) IS, 15 (23%) HS,and 10 (15.4%) NS patients. TIA manifestations were variable among subjects.Twenty subjects out of 29 experienced motor dominant symptoms, whereas onlytwo had sensory dominant symptoms. Isolated cognitive dysfunctions such aslanguage dysfunction occurred in four subjects. Stages of MMD evaluated bySuzuki’s method were significantly higher in IS group (3.64 ± 0.8) than in TIAgroup (2.90 ± 0.9; p=0.022). There was no significant difference in the surgicaloutcome in terms of TIA and stroke recurrence among the groups.Conclusion: We showed that TIA is very frequent initial manifestation amongadult MMD and that involvement of intracranial arteries are less extensive in TIAgroup. Our results suggest that TIA as initial manifestation mainly occurs in theearlier stage of the MMD.



THE ULTRASOUND SIGNS OF ENDOTHELIAL DYSFUNCTION IN ACUTEISCHEMIC STROKEM.A. Domashenko, A.O. Chechetkin, Z.A. SuslinaInstitute of Neurology, RAMS, Moscow, Russian Federation

Background: The aim of the study was to evaluate the ultrasound signs ofendothelial dysfunction in patients with acute ischemic stroke [IS].Methods: The ultrasound evaluation of the endothelium-derived vasodilatation ofthe brachial artery [BA] was performed in 27 patients (age 67 [55; 79] years; 14males, 13 females) with IS in the first 48 hours of stroke onset. The maximalincrease of BA diameter after “cuff test” [CT] with the transient occlusion of BA bythe cuff of manometer was evaluated. 20 patients (age 64 [51; 72] years; 10 males,10 females) with the chronic ischemic cerebrovascular diseases were investigatedin the control group.Results: The neurological deficit in IS patients was 79 [58,5; 90] on ESS and 4,5[2,5; 9] on NIHSS. The initial linear speed of the blood flow [LSBF] on BA was 62cm/s [54; 65] and 58 cm/s [50; 65] and the initial diameter of BA was 4,6 mm [4,0;5,1] and 5,1 mm [4,6; 5,4] in patients with IS and control group accordingly. Afterthe CT the LSBF on BA increased on 121% [103; 219] and 184% [126; 223] inpatients with IS and control group (p=0,07). The increase of BA diameter after theCT was 5,4% [4,3; 9;5] and 8,5% [6,8; 11,5] in patients with IS and control group(p=0,035). The increase of BA diameter was directly correlated with ESS score (R0,35; p=0,047) and indirectly correlated with NIHSS score (R -0,33; p=0,049).Conclusion: The BA dilatation after CT in patients with IS was decreased com-pared to patients with the chronic cerebrovascular diseases. The ultrasound signsof endothelial dysfunction in patients with IS and their correlation with the strokeseverity were demonstrated.


S100B AS A BIOMARKER: ITS OPTIMAL ROLE IN STROKEP. Dassan, G. Keir, R. Jager, M.M. BrownUCL, Institute of Neurology, London, United Kingdom

Background: Blood biomarkers may be important in three areas of acute stroke:diagnosis; as a surrogate marker for severity of brain damage; and predictingprognosis. The S100B protein has been studied in each area individually in selectedpatients but there are no studies directly comparing its utility in these areas. Wecorrelated all three measures with serial measurements of S100B in an unselectedseries to determine its optimal role.Method: Blood samples and National Institute of Health Stroke Scale (NIHSS)scores were taken on arrival to hospital and daily, where possible, for up to 6 daysafter onset of symptoms in 40 consecutive patients with suspected ischaemic stroke(26 acute infarcts and 14 stroke mimics). Serum S100B was measured by enzyme-linked immunosorbent assay. Infarct volumes were measured on diffusion-weightedimages.Results: In patients venesected within 24 hours of symptom onset there was nosignificant difference in S100B levels between acute infarction and stroke mimics(mean 0.19 ng/ml vs 0.12 ng/ml). Peak S100B levels after 24 hours however,correlated well with both infarct volume and maximum NIHSS scores (r = 0.89,P<0.001 and r = 0.81, P<0.001 respectively). The highest level was seen in apatient with malignant middle cerebral artery infarction. Peak S100B level wasa good predictor of dichotomised outcome after discharge (independent mean0.14ng/ml vs dependent mean 0.36ng/ml, P<0.05).Conclusion: Serum S100B measurements are not helpful in distinguishing infarc-tion from stroke mimics in the emergency room. Measurement of blood S100Blevels after cerebral infarction is a useful measure of the severity of brain damageand predicts prognosis. It may also be a useful predictor of malignant infarction.


PREHOSPITAL TRANSPORT OF ACUTE STROKE PATIENTS AND TIME TOINITIAL MEDICAL ASSESSMENT IN AN IRISH GENERAL HOSPITALR. Purcell, G. Bergin, C. Cooney, E. Farrelly, R. Morton, H. Logan, R. Lynch,S. MurphyMidland Regional Hospital, Mullingar, Co. Westmeath, Dublin, Ireland

The aims of acute stroke management are to minimise neurological impairmentand maximise functional recovery. Rapid patient transfer to acute stroke units isessential to achieve these aims. In Ireland, there is no nationally agreed policy onacute stroke care.A retrospective study was performed to examine prehospital transport of acute

stroke patients to a general hospital. HIPE identified patients with acute strokepresenting to our hospital from 01-01-05 to 30-06-05. Data was collected fromthe emergency service’s and hospital’s records. Time intervals from initial contactwith the emergency services to medical assessment in hospital were recorded. Dataon time of symptom onset was available for patients who arrived via personaltransport.Forty-six patients presented with an acute stroke (84.78% infarcts). Thirty-onepatients arrived by ambulance [12 by 999 call= (1), 19 non-999 call= (2)],15 viapersonal transport (3). Mean interval from the time of the ambulance call to arrivalat scene was 24 mins (1) vs 28 mins (2) [NS]. Mean interval from time of call toarrival in the emergency department was 66.3 mins (1) vs 74.4 mins (2) [NS]. Meaninterval from time of emergency department arrival to medical assessment was 95mins(1), 105 mins(2) and 43 mins(3) [(1)/(2) vs (3) p=0.01/p=0.002].The averagetime from symptom onset to arrival in the emergency department was 861.6 minsfor patients arriving by personal transport.The rapid ambulance response contrasts with the delayed medical response to acutestroke at our institution. Better public and medical awareness of the urgency ofacute stroke management is necessary. This study has provided useful baseline dataon our current performance regarding the transport and medical assessment of acutestroke patients. We plan to put in place an integrated response for acute stroke inorder to maximise patient outcomes.

Acute stroke: complications and early outcome

1 Acute stroke: complications and early outcome

THE RELATIONSHIP BETWEEN THE LOCATIONS OF DEEP-VEINTHROMBOSIS AND MOTOR IMPAIRMENT IN ACUTE ISCHEMIC STROKEPATIENTSD.G. Sherman, G.W. Albers, C. Bladin, C. Fieschi, A.A. Gabbai, C.S. Kase,W. O’Riordan, G.F. Pineo, for the PREVAIL InvestigatorsUniversity of Texas Health Science Center at San Antonio, San Antonio, TX,USA

Introduction: Studies of major orthopaedic surgery have shown that deep-veinthromboses (DVT) do not always occur on the same side of the body as the surgicalintervention suggesting that reduced mobility is not the only factor triggeringthrombus formation. We assessed the relationship between the side of the bodyaffected by motor impairment and the side with DVT in PREVAIL, a study of VTEprophylaxis in acute ischemic stroke patients.Methods: Patients with acute ischemic stroke, confirmed by CT scan or MRI, andunable to walk unassisted due to motor impairment of the leg were randomizedwithin 48 h of stroke symptoms to enoxaparin 40 mg SC qd or UFH 5000 IUSC q12h for 10±4 days. DVT was confirmed by venography, or ultrasonographywhen venography was not practical. PE was confirmed by VQ or CT scan, orangiography.Results: The PREVAIL study showed a 43% relative reduction in the risk ofsymptomatic or asymptomatic deep-vein thrombosis (DVT), symptomatic pul-monary embolism (PE), or fatal PE with enoxaparin compared with UFH in acuteischemic stroke patients (10.2% vs 18.1%; p=0.0001), with no increase in clinicallyimportant bleeding. A post-hoc analysis showed that 7.0% of patients developed aDVT on the same side as the motor impairment, and 3.5% developed a DVT on thecontralateral side.Conclusion: Although more DVT events occur on the same side as the motorimpairment, about one third occur in the contralateral leg. This suggests that whileflow-dependent thrombogenic factors (i.e. stasis) are triggers for thrombus forma-tion, some other factors may also be important. Rehabilitation and nursing careshould focus on mobilization of the patient as well as providing VTE prophylaxis.


PREVALENCE AND RISK FACTORS OF FAECAL INCONTINENCE INSTROKE PATIENTS ADMITTED TO THE ACUTE STROKE UNIT AND TOREHABILITATION WARDS (PILOT STUDY)U. KhanOxford Redcliff, Abingdon, United Kingdom

Background and purpose: Faecal incontinence [FI] commonly affects patientsafter stroke. This is an observational study to assess the prevalence of and riskfactors for FI in stroke patients admitted to the acute stroke unit & to rehabilitationwards.Methods: 50 patients admitted in Oxford Radcliffe Hospital stroke unit from May

Poster SessionAcute stroke: complications and earlyoutcome


2006 for a period of 10 wks were enrolled and followed up in rehabilitation wardsand in the community. The prevalence of FI was assessed over a time period of3 months and risk factors were assessed in the1st wk after admission. The maincomparative statistical tool used was the Chi squared test or Fisher’s exact tests.Results: Pre-stroke FI was 2%.Post-stroke FI at wk1 was 34%, at wk 4 was 26%& at wk 12 was 21%.The characteristics of 17 FI patients were compared with 33 without FI.Totalanterior circulation stroke syndrome was strongly associated with FI (P<0.0001)as was stroke severity, measured by NIHSS >14; GCS<15 (P<0.0001).Therewas a significant association between poor cognition [MMSE score <25] and FI(P=0.009). Disability [Barthel Index<15] was also significantly associated withFI (P=0.002). Concurrent urinary incontinence [UI] (P<0.0001) and diarrhoea(P=0.027) were but constipation (P=0.70) and faecal impaction (P=0.327) were notassociated with FI. Enteral feeding was strongly associated with FI (P<0.0001).Advanced age (>80) was not a significant factor when adjusted for gender(P=0.061) but showed a positive trend.Conclusion: FI is common in older stroke patients & resolves in <50% within3 months. Stroke severity, conscious state, level of disability, co-existing UI,diarrhoea and enteral tube feeding are significant risk factors. This pilot providesdata to plan future larger studies of FI in stroke patients.


COMPARISON OF 2 TYPES OF PROGRESSION AFTER ACUTE ISCHEMICSTROKE: CONTINUOUS DETERIORATION VS. FLUCTUATION OF STROKESEVERITYY.-J. Cho, K.-S. Hong, J.-.S. Koo, K.-H. Yu, H.-J. Bae, M.-K. Han,M.-K. Jeong, D.-W. Kang, J.-M. Park, B.-C. LeeInje University Ilsan Paik Hospital, Goyang-si Gyeunggi-do, South Korea

Background: To investigate the frequency, possible attributable factors and theprognosis of 2 types of progression after acute ischemic stroke.Methods: All consecutive patients with first-ever ischemic stroke within 24 hoursfrom onset were recruited prospectively, who admitted 4 university hospitals inSeoul metropolitan region. Baseline demographics, stroke subtypes, past medicalhistory, medical complications after stroke, and modified Rankin Scale at 3 monthswere assessed by predetermined protocol. Stroke severity was assessed by NIHStroke Scale (NIHSS) at admission, on hospitalization days 1, 2, 3, and week 1 and2. Clinical deterioration was defined as decrease of 2 points in total NIHSS scoreor 1 point in motor scale score.Results: Among 566 patients recruited, 142 (25.1%) worsened. One hundred thirtypatients (91.6%) were deteriorated within 3 days after stroke onset. Continuousdeterioration (CD) was found in 94 (66.2%), and returning to initial status afterfluctuation (F) was found in 48 (33.8%). CD patients are older (68.6±11.2 vs.62.0±12.4), and more likely to have higher initial NIHSS (median=7.5 vs. 6.0),preceding systemic infection (24.5% vs. 12.5%), and worse functional outcome(mRS 3-6=78.7% vs. 56.2%) at 3 months than F patients. After adjusting sex, hy-pertension, diabetes, and stroke subtypes, age was the only significant independentpredictor of CD (OR=1.04, 95% CI=1.01-1.08) by logistic regression analysis.Discussion: Continuous deterioration of stroke severity after acute ischemic strokeresults in poor functional outcome. Age, initial NIHSS, and preceding systemicinfection might predict further decline.


COMPARISON OF NUTRITIONAL INTAKE IN POST STROKE PATIENTS ONNORMAL, MODIFIED AND NG DIETT. Nagarajan, A. Addison, M. Winder, A.G. DykerFreeman Hospital, Newcastle upon Tyne, United Kingdom

Patients with swallowing problems after stroke are at risk of developing nutritionaldeficiencies.Methods: Three groups of patients were studied: Normal diet (n=4), modifiedthickened diet (n=5), and naso-gastric feeding(n=4). Patients were studied for sevendays. Full records were kept of patients’ nutritional intake. Patients were weighedon day one and day 7. Results were tabulated using Excel (2003) and statisticalanalysis was carried out using MINITAB (14). Test for normality was carried outon all analysed data and students T test was used to compare normally distributeddata while Mann–Whitney Tests and Wilcoxon Signed Rank test were used tocompare non-normally distributed data.Results: Mean daily calorie intake was significantly lower in those being feda modified diet compared with NG (difference 546, p<0.0005) and normal diet(difference 868, p<0.0005).There was no difference in daily calorie intake betweenNG and normal diet. Average daily calorie intake was significantly less than

recommended (2000) in all groups, particularly those on modified diet (normaldiet mean 1311; SD 520, modified diet 765 SD 464, NG diet 1633 SD 780: all pvalues<0.0001). Median daily protein intake was significantly lower in those on amodified diet (39.2g) compared with NG feeding (56.9g, p=0.012) a normal diet(50g, p=0.02) and recommended levels (50grams). Mean seven day total calorieintakes were significantly lower than recommended levels (14000) in patients ona normal diet (8854 p=0.023), and modified diet (5052 p=0.001) but not in thosebeing fed with NG (11436p=0.32). This interim analysis was not powered todemonstrate reductions in weight, but patients on a modified and normal diet lost3.5 kg in 7 days while those on NG lost 0.3kg. There was a strong negativecorrelation between weight loss and both calorie and protein intake (Pearson -0.79,and 0.-0.89 p= 0.001).Conclusions: Patients on modified diet are at risk of malnutrition. Consideration offeeding supplementation should be given to patients with swallowing difficulties


THROMBOLYSIS WITH RT-PA DOES NOT PROMOTE EDEMA FORMATIONIN ACUTE ISCHEMIC STROKEV. Sachsenmaier, I. Dzialowski, C. Disque, G. GahnTechnical University of Dresden, Dresden, Germany

Background: There is ongoing controversy whether recombinant tissue plas-minogen activator (rt-PA) treatment is associated with excessive edema formationfollowing acute ischemic stroke. Possible mechanism might be the toxic disruptionof the blood brain barrier and subsequent reperfusion injury. We studied thehypothesis that patients with large middle cerebral artery (MCA) infarctions treatedwith rt-PA develop larger cerebral edemae than controls.Methods: We retrospectively studied patients with large MCA infarctions from07/2001-10/2006 defined by an Alberta Stroke Program Early CT Score (AS-PECTS) < 5 on day 1 – 7 follow-up imaging. Exclusion criteria were space-occupying secondary hemorrhage, hemicraniectomy, and thrombolysis with non-t-PA agents. We recorded baseline characteristics and rt-PA treatment status from thepatient chart. We prospectively assessed the extent of cerebral edema applying a 5-scale grading system (0, no edema, 1, compression of external subarachnoid spaces,2, compression of lateral ventricle, 3, midline deviation, 4, obstructive dilatation ofopposite lateral ventricle, Borovich scale). We compared Borovich scores amongrt-PA treated patients and controls using non-parametric statistical testing.Results: We identified 90 patients for our study with a mean age of 66 ±13 years,42% female, mean onset-to-imaging-time of 59 ±40 hours, median ASPECTS of 2(iqr2-4). Of the 90 patients, 30 were thrombolysed and 60 were controls. Baselinecharacteristics did not differ across the two groups. Distribution of Borovich scalescores for the treatment (and control) group was 0 (0) % for grade 0, 0 (8) % forgrade 1, 43 (40) % for grade 2, 37 (33) % for grade 3, and 20 (18) % for grade 4.Median Borovich scores among both groups did not differ significantly (median 3,p=0.48).Conclusion: In our retrospective study, t-PA treatment did not exacerbate cerebraledema in patients with large MCA infarctions. Further prospective evidence isneeded to confirm this result.1723/1800 characters


ACUTE MORTALITY PREDICTION IN STROKE PATIENTSM. Delobel, A. Viguier, M.C. Turnin, V. LarrueCHU de Toulouse France, Toulouse, France

We examined the use of the Simplified Acute Physiology Score II (SAPS II) forthe prediction of in-hospital mortality in a large number of stroke patients managedin a neurological intensive care unit.Data on SAPS II were prospectively collected in patients with ischemic stroke,cerebral hemorrhage, transient ischemic attack, or subarachnoid hemorrhage, con-secutively admitted to a tertiary neurological intensive care unit. We constructedreceiver operating characteristic curves (ROC) to determine the ability of SAPS IIto predict in-hospital mortality.2214 patients were included in this analysis. 321 (14.5%) patients died in hospital.The area under ROC curve [95% confidence interval] was 0.83 [0.81-0.86]. Witha cut-off point of 30 the positive predictive value of SAPS II was 38.3%, and thenegative predictive value 94.8%. Findings were similar in an analysis restricted topatients older than 40 years.The findings indicate that SAPS II is a reliable tool to predict acute mortality inpatients managed for stroke in a neuroligical intensive care unit.



A NOVEL POLYMORPHISM IN THE PROMOTER REGION OF THESURVIVIN GENE IS RELATED TO HEMORRHAGIC TRANSFORMATION INPATIENTS WITH ACUTE ISCHEMIC STROKEM. Castellanos, C. Gubern, J. Serena, J. Castillo, M.A. Moro, M. Millán,R. Rodríguez, F. Nombela, O. Hurtado, J. MallolasHospital Universitari de Girona Doctor Josep Trueta, Girona, Spain

Background: The expression of survivin, a member of the inhibitor antiapoptoticproteins family, has been shown to increase after cerebral ischemia in responseto the release of angiogenic growth factors. This protein has been mainly locatedat the microvaculature within the infarcted and peri-infarterd area, and so weaimed to investigate whether survivin gene polymorphisms were associated withhemorrhagic transformation (HT) of cerebral infarction.Methods: DNA was isolated from peripheral blood samples of 97 patients witha hemispheric ischemic stroke and 38 controls. Polymorphism screening of thesurvivin gene was performed by polymerase chain reaction, single-strand con-formation polymorphism and sequencing analysis. DWI sequences were obtainedwithin 12 hours from symptoms onset and at 72±12 hours by using a 1.5-T MRI.The presence of HT was determined on the second DWI sequence and classifiedaccording to ECASS II criteria.Results: Forty-seven patients (48%) had HT: 27 patients had hemorrhagic infarc-tion and 20 had parenchymal hemorrhage (PH). Thirty-two patients (32%) receivedrt-PA. A novel polymorphism was identified in the promoter region of the survivingene which corresponded to an C-to-T transition at -241 bp from the transcriptionstart site. The prevalence of the mutant allele (T) was similar in patients andcontrols (14% vs. 16%, respectively; p=0.985). Seven patients (26.9%) with alleleT had HT compared to 40 (56.3%) of wild-type (p=0.009). Logistic regressionanalysis showed that the presence of the polymorphism was associated with alower risk of HT (OR 0.28; 95% CI, 0.08 to 0.97; p=0.045) independently of age,baseline stroke severity, temperature, platelet count, glucose levels, systolic bloodpressure, DWI lesion volume and rt-PA administration.Discussion: The -241 C/T polymorphism in the promoter region of the survivingene is associated with a lower risk of HT in patients with acute ischemic stroke.This polymorphism might be related with a decrease of survivin expression andsecondary down-regulation of the angiogenic process.


STROKE PATIENT PROGRESSION IN DIJON FROM 1985 TO 2004P. Decavel, Y. Bejot, G.V. Osseby, B. Parratte, T. Moulin, M. GiroudBesançon University Hospital, Besançon, France

Background: Development of stroke management over a number of years haschanged the vital and functional prognosis of patients.Method: The main aim was to test the outcome of patients with first-ever strokeover a long period among a random population. A study was carried out intothe progression of the number of deaths over 20 years, handicap developmentaccording to the modified Rankin scale and the clinical state of patients cominginto the department with a first-ever stroke identified in an ongoing registry of thepopulation between 1985 and 2004.Results: The distribution of stroke type was as follows: 3142 infarctions, 341hematomas and 74 subarachnoid hemorrhages. Over 20 years, mortality has de-clined by 0.94% per year (p<0.01), the number of patients able to walk 28 daysafter stroke has increased by 0.78% (p=0.02) per year, with no increase in thenumber of patients severely handicapped (p=0.43). If the age at which the first-everstroke took place has risen, the number of patients initially comatosed has notchanged (p=0.06).Discussion: The decline in mortality observed in Dijon confirms the tendencyobserved in the majority of other registries. The Dijon registry is, however, the onlyone to be ongoing. The improvement in stroke patient progression is significantdespite the increase in the age at which the first event took place. Over a period of20 years, a 50% decrease in stroke patient mortality with no rise in severe handicaphas been observed in Dijon.


EARLY HEMORRHAGIC EVENTS AFTER INTRAVENOUS THROMBOLYSISOF HEMISPHERIC AND BASILAR ISCHEMIA: RESULTS OF THE HELSINKISTROKE THROMBOLYSIS STUDYT. Bogoslovsky, O. Häppölä, L. Soinne, O. Salonen, P.J. Lindsberg, M. KasteHelsinki University Central Hospital, Biomedicum, Neuroscience Program,Helsinki, Finland

Early cerebral hemorrhages after ischemic stroke are the most feared adverse eventfollowing the theapeutic use of recombinant tissue plasminogen activator. Thereason for bleeding is not known, but their incidence has been associated withvarious premorbid factors, such as diabetes, the use of antithrombotics, the durationand extent of cerebral ischemia as well as the perithrombolytic levels of bloodglucose and blood pressure (BP). Onset-to-treatment times tend to be longer inbasilar occlusions, but it is not known, if their bleeding tendency is different.We aimed to asses the rate of hemorrhagic events within 24 hours after thrombolysisand the associated factors in consecutive anterior and posterior circulation strokepatients treated in Helsinki University Central Hospital during years 2003 to 2005.Of 335 strokes 304 were hemispheric (HS) and 31 basilar occlusions (BAO). BAOpatients had more severe strokes (median NIH Stroke Scale 20 vs. 10, p<.001) andlonger onset-to-treatment times (873±1453 min vs. 127±43 min, p<.0001) andless antiaggregatory treatment (22.6% vs 43.1%, p=.03), but similar age (64±17vs. 69±12 years, p=.14) and prevalence of diabetes [4(15%) vs. 33(12%),p=.72].BAO patients had comparable prethrombolytic glucose (7.0 vs 7.1 mmol/l) and BPlevels (systolic 152 vs 156 mmHg, diastolic 81 vs 82 mmHg). Overall incidence ofpostthrombolytic hemorrhages was 64 (21%) in HS and 5 (16%) in BAO (p=.52).There was no difference in distribution into subgroups of hemorrhagic eventsaccording to ECASS II classification or extraischemic or subarachnoidal bleeding.Despite 7-fold longer treatment delay and 2-fold stroke severity, thrombolysis ofBAO is not associated with more major hemorrhages than that of HS.


ENDOTHELIAL DYSFUNCTION IN ACUTE ISCHEMIC STROKE ISCORRELATED WITH STROKE SEVERITY AND THE SIZE OF THE BRAININFARCTIONM.A. Domashenko, S.V. Orlov, M.M. Tanashyan, V.G. Ionova, M.V. Kostyreva,R.M. Umarova, A.S. Suslin, M.V. Krotenkova, Z.A. SuslinaInstitute of Neurology, RAMS, Moscow, Russian Federation

Background: The aim of the study was to evaluate the concentration of vonWillebrand factor [vWf] in acute ischemic stroke [IS] compared with the strokeseverity and the size of the brain infarction.Methods: The concentration of serum vWf was investigated in 40 patients (age 65[57; 74] years; 22 males, 18 females) with IS in the first 48 hours and on 21 dayof IS onset. The size of the brain infarction was measured on diffusion-weighted(DWI) MRI images (in the first 48 hours of IS) and on T2 MRI images (on 21day of IS). Correlation analysis of vWf concentration, NIHSS score and the size ofbrain infarction was performed.Results: The vWf concentration was 158% [130; 181] and 170% [147; 200] inthe first 48 hours and 21 day of IS accordingly (p=0,03). The neurological deficiton NIHSS was 4,5 [2,5; 9] and 1,5 [0; 4] in the first 48 hours and 21 day ofIS accordingly. The vWf concentration in the first 48 hours of IS was directlycorrelated with NIHSS score (R 0,33; p=0,049). The size of the brain damage onDWI was 16,1 sm3 [4,7; 40,4] in the first 48 hours of IS and was directly correlatedwith vWf concentration (R 0,59; p=0,046). The size of brain infarction on T2 MRIwas 11,2 sm3 [6,8; 32,2] on the 21 day of IS and also was directly correlated withvWf concentration (R 0,71; p=0.009).Conclusion: The serum vWf level is correlated with the stroke severity and thesize of the brain infarction in patients with IS.


IS THE ASSOCIATION OF COAGULATION ACTIVATION MARKERSWITHPROGRESSING STROKE DUE TO THE ACUTE-PHASE RESPONSE?J.M. Barber, P. Welsh, P. Langhorne, A. Rumley, G.D. Lowe, D.J. StottRoyal Infirmary, University of Glasgow, Glasgow, Airdrie, Scotland, UnitedKingdom

Introduction: Early progression of ischaemic stroke is common, occurring inaround 25% of patients. This complication is associated with poor outcome. Wehave demonstrated that progression is associated with elevation of markers of co-agulation activation. We aimed to determine whether the association of progressing



stroke with haemostatic activation is due to an underlying enhanced acute phaseinflammatory response.Methods: Consecutive ischaemic stroke patients were recruited. Progressing strokewas defined by deterioration in components of the Scandinavian Stroke Scaleover the first 72 hours. Measures of Interleukin (IL)-6, IL-18 and tumour necrosisfactor-alpha (TNF-A) were made in addition to highly sensitive C-reactive protein(hsCRP).Results: We studied 150 patients of whom 35 (23%) had progressing stroke byEuropean Progressing Stroke Study criteria. IL-6 levels [11.3 (6.0-16.6) v. 7.0(3.3-13.8), p=0.02] and hsCRP [8.66 (3.69-30.4) v. 5.26 (1.64-18.4), p=0.05] werehigher in the progressing group. IL-18 [median 295 (216-452) v. 280 (212-375),p=0.42] and TNF-A levels [2.53 (1.86-3.67) v. 2.24 (1.58-3.10), p=0.18] weresimilar in progressing and non-progressing cases. In binary logistic regression noneof the acute phase markers were independent predictors of progressing ischaemicstroke; thrombin-antithrombin complexes (odds ratio (OR) 7.74), admission meanarterial blood pressure (OR 1.4 for each 10mmHg rise) and age (OR 1.05) wereindependent predictors of progressing stroke.Conclusions: Circulating levels of IL-6 and hsCRP are elevated in subjects withprogressing ischaemic stroke. However, these inflammatory markers are not inde-pendent predictors of stroke progression. The acute phase inflammatory responseis unlikely to be a major contributor to haemostatic activation in progressingischaemic stroke.


THE EFFECT OF DYSPHAGIA ON COMPLIANCE AND OUTCOME IN THE‘EFFICACY OF NITRIC OXIDE IN STROKE’ (ENOS) TRIALG.M. Sare, L. Gray, T.J. England, P.M.W. Bath, for the ENOS InvestigatorsInstitute of Neuroscience, University of Nottingham, Nottingham, UnitedKingdom

Introduction: About 50% of stroke patients have dysphagia at presentation andthis may result in the discontinuation of pre-stroke medical therapy and delay theinitiation of acute oral therapy. Enteral access is unreliable since naso-gastric tubescan be difficult to insert and are often pulled out. Here we examine the relationbetween the administration of oral and transdermal medication in patients withacute stroke.Methods: ENOS is an international, randomised controlled trial in 5,000 patientswith acute ischaemic or haemorrhagic stroke which is investigating the safetyand efficacy of lowering BP with transdermal glyceryl trinitrate (GTN, given for7 days). Patients taking pre-stroke antihypertensive therapy are also randomisedin a partial-factorial design to continue or temporarily stop this. We comparedtreatment compliance and outcome (Modified Rankin Scale (mRS) at 90 days postrandomisation) between those with and without dysphagia.Results: As of 10 January 2007, 559 patients (28 centres, 9 countries) had beenrecruited into ENOS. 283 (51%) of these patients had dysphagia. In analysesblinded to treatment assignment, 87% of dysphagic patients received all of theirrandomised treatment between days 1-4 in the GTN patch versus No GTN arm ofthe trial compared to 76% in the stop versus continue arm (p=0.01). Those withdysphagia had more severe strokes at baseline and significantly worse outcome at90 days compared to those without: median mRS 3 (inter quartile range, IQR 2-5,n=283) compared to median 2 (IQR 1-3, n=276) (p<0.0001).Discussion: Stroke patients with dysphagia were more likely to receive transdermaltreatment than oral antihypertensive medication. These data highlight the potentialbenefits of transdermal medications in this high risk population.


SIDE DIFFERENCES ON THE EXTENT OF SALVAGEABLE ISCHEMICTISSUE IN HYPERACUTE STROKER. Delgado-Mederos, M. Ribo, A. Rovira, J. Alvarez-Sabin, M. Rubiera,J. Munuera, E. Santamarina, P. Delgado, O. Maisterra, C.A. MolinaHospital Vall d’Hebron, Barcelona, Spain

Background: Previous research has suggested that right hemisphere stroke (RHS)may achieve worse outcome after thrombolysis. We aimed to evaluate the influenceof the side of affected hemisphere on the extent of baseline MRI abnormalitiesin acute stroke and to assess the value of MRI-based selection approach for tPAtreatment in RHS.Methods: We prospectively studied 145 acute stroke patients due to proximal MCAocclusion imaged with MRI within the first 6 h from symptoms onset. Those withunclear onset time were excluded. DWI and time-to peak (PWI) lesion volumeswere measured. MRI inclusion criteria for tPA were PWI/DWI mismatch>20%and DWI volume<50% of MCA territory. Recanalization was assessed by TCD at

6 h and hemorrhagic transformation (HT) on MRI at 24-48 h. mRS score was usedto assess 3-month outcome.Results: 66 (45.5%) patients had RHS. Baseline NIHSS was lower in RHS (median17 vs 19, p=0.0001). Time to MRI was similar between both hemispheres. RHSpatients presented with larger DWI volume (73.9 vs 38.5 cc; p=0.004) and smallerPWI/DWI mismatch (63% vs 79%, p=0.011). Only 13 (16.5%) patients with LHSdid not meet MRI criteria for thrombolysis, compared to 21 (31.8%) of thosewith RHS (p=0.03). 85 (58.6%) patients were treated with tPA (34 RHS, 51 LHS;p=0.112). Among those treated with tPA, baseline NIHSS was lower in RHS (16 vs19; p=0.0004), whereas DWI and PWI/DWI mismatch volumes were comparable.Recanalization rates (RHS 53.1% vs 52.2%, p=0.934) did not differ between bothhemispheres. Only 1 patient had symptomatic HT. At 3 months, good clinicaloutcome (mRS 0-2, RHS 33.3% vs 39.1%, p=0.609) and mortality (RHS 15.2% vs23.9%, p=0.339) were similar between RHS and LHS.Conclusion: Patients with a RHS present with larger infarct volumes and lessersalvageable penumbral tissue, suggesting a possible less efficient pattern of collat-eral circulation in the right hemisphere. A MRI-based screening for thrombolysis,irrespective of the time window, may improve the selection of patients with RHSlikely to benefit from tPA.


IS RENAL FAILURE A PREDICTOR FACTOR OF POOR EVOLUTION INACUTE STROKE?L. Gabaldón, B. Fuentes, J. Fernández, L. Idovro, P. Martínez, E. Díez-TejedorUniversity Hospital La Paz, UAM, Madrid, Spain

Background: Previous studies have pointed out that renal failure is an independentfactor of poor outcome in patients with cardiac failure, myocardial infarction andcoronary surgery. However, no studies analysing its possible influence on strokeoutcome are available.Methods: Observational study including consecutive first-ever acute stroke in-patients with a two-years recruitment period. Renal failure was defined ascreatinine level 3 1.2 mg/dl on admission or previous diagnosis of it. In-hospitalmortality and outcome at discharge (modified Rankin Scale) were the main outcomemeasures.Results: 445 patients were included, mean age 69.7±13.2. In the univariateanalysis creatinine >1.2 mg/dl was associated to more in-hospital mortality (31.7vs 20.7%; p=0.04) but not to poor outcome at discharge. In multivariate logisticregression analysis the predictive factors independently associated to in-hospitalmortality were: stroke severity on admission (OR 0.49; 95% IC 0.39-0.63) and thedevelopment of systemic (OR 17.97; 95% IC 5.47-59) or neurologic complications(OR 23.49; 95%IC 7.25-76.01) without any influence of renal failure.Conclusions: Renal failure measured by creatinine serum level does not signif-icantly influence in-hospital mortality or outcome at discharge in acute strokepatients. However, new studies analysing other parametres of renal function suchas creatinine clearance are need to get definite conclusions.


SIGNIFICANCE OF COMMON COMPLICATIONS DURING THE FIRSTWEEK POST STROKE: IMPACT ON FUNCTIONAL OUTCOME AT 90 DAYSG. Rohweder, E. Naalsund, P. Oksnes, B. IndredavikSt. Olavs University Hospital, Department of Neuroscience, NTNU,Trondheim, Norway

Background: Little is known about the prognostic impact of acute post-strokecomplications. The aim of this study was to define the longterm functional outcomein patients who experienced a common (>2.5%) complication during the first weekafter a stroke, while admitted to an acute comprehensive stroke unit and followedup by an early supported discharge service.Methods: 244 patients consecutively admitted to our stroke unit and with a mod-ified Rankin Scale (mRS) of < 2 were included on admission and followed withassessment for 8 complications: fever, diffuse pain, progressing stroke, urinary tractinfection (UTI), isolated Troponin T elevation (Trop T), chest infection, non-seriousfalls, and acute myocardial infarction (AMI). After discharge, the patients werefollowed up for 90 days with weekly telephone assessments and a home-visit andfunctional assessment after 3 months.Results: The mean age of patients was 77 yrs., 56% were women. The frequencyof complications were as follows: fever 26%, diffuse pain 25%, progressing stroke18%, UTI 17%, Trop T 13%, chest infection 12%, non-serious falls 7%, AMI 6%.48% of all patients had an improvement >1 on the mRS. Frequencies of mRS >

1 were as follows: In pts with a chest infection: 18%, without 52% (p=0.001). Inpts with Trop T: 34%, without 50% (p=0.089). In pts with AMI: 30%, without


49% (p=0.229). In pts with UTI 46%, without 48% (p=0.777). In pts with fever49%, without 48% (p=0.876). In pts with progressing stroke 60%, without 46%(p=0.080). In pts with diffuse pain 62%, without 44% (p=0.012). In pts withnon-serious falls 81%, without 46% (p=0.006).Discussion: This study suggests that the occurrence of chest infection during thefirst week after a stroke leads to a worsened outcome, while the occurrence ofTroponin T elevation and acute myocardial infarction may do so. Frequent fallsand diffuse pain are indicators of improved functional outcome at 90 days after astroke. Progressing stroke does not confer a worsened outcome under the currentmanagement.


EARLY PULMONARY EMBOLISMS IN PATIENTS TREATED WITH FACTORVIIA FOR INTRACEREBRAL HEMORRHAGE. CHANCE OR SAFETYCONCERN?W. Vadot, A. Jaillard, P. Bedouch, C. Chevallier, K. Garambois, O. Detante,B. Hommel, M. HommelUniversity Hospital of Grenoble, Grenoble, France

Introduction: Intracerebral hemorrhage (IH) was the least treatable type of stroke.Activated Factor VII (rFVIIa) IH Trial (FIHT) suggested that treatment of IH withrFVIIa within 4 hours improves outcomes despite thromboembolic adverse events(Mayer et al. 2005). Our aim was to evaluate feasibility and safety of rFVIIa in astroke Unit.Methods: We consecutively included patients with IH admitted during years 2005and 2006, who met criteria for inclusion in FIHT. Patients admitted in 2006 wereadministered 80 μg of rFVIIa per kilogram of body weight and were compared withthe historical controls admitted in 2005. Clinical outcome were NIHSS, Rankinscore and mortality. Safety was assessed using serum troponin and D-dimer atbaseline, 6 hours and on day 1, 2, 3 and later when necessary. Lower limb Dopplerand thoracic CT scan were performed in patients with high D-dimer levels in orderto assess deep venous thrombosis (DVT) and pulmonary embolism (PE).Results: Among the 30 patients included (NIHSS=14), 15 received RFVIIa. Inthis group, none had myocardial infarction, one died of cerebral edema, and twosuffered PE. The first PE occurred at day one and the other at day 2 with re-bleedingat day 3. Both patients with EP recovered. Because each received 7.2 mg of RFVIIain relation to high weight and had no other risk factor, we limited later on thehighest dose at 4,8 mg. None of the next patients suffered PE but one had PVT. Inthe control group, 2 patients died, 3 had DVT but none had PE. Clinical outcomewas not different in the two groups.Discussion: At our knowledge, early PE was not reported as thromboemboliccomplications in patients treated by RFVIIa for IH. A dose effect relation wassuggested by the high total dose received the two patients with PE.


ASPIRIN RESISTANCE IS ASSOCIATED WITH INFLAMMATION, ISCHEMICSTROKE SEVERITY, AND POORER FUNCTIONAL OUTCOME AT 6MONTHSN.A. Englyst, G. Horsfield, C.D. ByrneUniversity of Southampton, Southampton, United Kingdom

Background: Stroke is the largest cause of disability in the UK but little isknown about which factors influence recovery. Aspirin is used in primary andsecondary prevention of stroke. Aspirin resistance is associated with a higher riskof developing stroke but its relationship with severity of stroke and functionaloutcome after stroke is unclear. The aim of this study was to investigate therelationship between aspirin resistance and inflammatory cytokines, stroke severityand functional outcome at 6 months.Methods: Aspirin resistance was assessed by thrombelastography in 100 peoplewith ischemic stroke and 100 community-based controls. Stroke outcome (degreeof disability) was assessed using the Rankin’s Stroke Scale within 72 hours ofstroke and at 6 months. Plasma interleukin IL-6 was measured by Enzyme LinkedImmunosorbent Assay (ELISA).Results: Aspirin resistance was associated with a higher Rankin’s scale at baseline(p=0.013), suggesting that aspirin resistance is associated with more severe strokes.Aspirin resistance at baseline was also associated with higher Rankin’s scaleat 6 months (p=0.048). Aspirin resistance was associated with increased IL-6(p=0.034) and higher levels of IL-6 were associated with poorer outcome fromstroke (p=0.017). IL-6 was independently associated with aspirin resistance inmultivariate analysis.Discussion: Aspirin resistance in conjunction with increased plasma IL-6 mayindicate a high risk of poor functional outcome from stroke. These data suggest that

screening for aspirin resistance after a stroke might identify a sub-group of peoplewho may benefit from higher doses of aspirin and/or combination therapy.


WHAT FACTORS INFLUENCE EARLY RECANALISATION DURINGTHROMBOLYSIS IN ACUTE ISCHEMIC STROKE?L. Sekoranja, H. Yilmaz, K. Lovblad, R. Grandjean, P. Temperli, R. SztajzelUniversity Hospital of Geneva, Geneve, Switzerland

Purpose: We evaluated the factors determining récanalisation after thrombolysis.Patients and methods: Patients with acute ischemic stroke (AIS) of < 3 hunderwent either IV or combined IV-IA lysis, if no recanalisation after 30’.CT-angiography was done in all p and monitoring with TCCD during 60’ in case ofIV and during 30’ in case of IV-IA lysis. TIBI was used to assess the residual MCAflow before the lysis and to evaluate presence or absence of early recanalisation(TIBI >1 at 30’).Results: 54 patients, 30 M, mean age 68.1 y. 33 p had IV and 21 IV-IA lysis. 43(80%) p had an MCA and 11 (20%) a T occlusion. Hypertension had 30 (55%),DM 8 (15%), and AF 23 p (43%); 18 (33%) were on ASA and 11 (20%) on statins.Thrombolytic was given within 60 to 230’. NIHSS ranged from 5 to 21. Fifteen(27%) p had a TIBI 0 before lysis, 9 a grade of 1(16%), 12 (22%) grade of 2 and18 (33%) grade of 3; 17 (31%) p received contrast because of insufficient temporalwindow. After lysis 25 (46%) p improved > 4 points on NIHSS. Factors increasingthe early recanalisation in univariate analysis: TIBI 1 to 3 respectively OR 8.1,95% CI 1.1 to 59, p= 0.039; OR 8.7, CI 95% 1.3 to 59, p= 0.021 and OR 15.6;95% CI 2.7 to 103, p= 0.002. The factors decreasing the early recanalisation: Tocclusion OR 0.02, 95% CI 0.003 to 0.20, p 0.001; M1 occlusion OR 0.11 95% CI0.02 to 0.6, p 0.011. In a multivariate analysis factors increasing the likelihood ofearly recanalisation: TIBI grades 1 to 3 respectively OR 11.9 95% CI 0.4 to 4.1, p=0.025; OR 13.8, CI 95% 1.3 to 105, p= 0.014 and OR 24.7; 95% CI 0.07 to 1.1, p=0.075 and atrial fibrillation OR 0.28, 95% CI 0.07 to 1.1, p 0.075.Conclusion: Presence of a residual flow of the MCA on TCCD (TIBI 1 to 3) wasthe best predictor of early recanalisation; presence of T or M1 segment occlusion onAngio-CT were associated with a lower one, however only on univariate analysis.


THE IMPACT OF HYPERTHERMIA AND INFECTION ON THE ACUTEISCHEMIC STROKE PATIENTS ADMITTED TO INTENSIVE CARE UNITW. Seo, S. Yu, J. Kim, S. KohKorea University College of Medicine, Seoul, South Korea

Background: Despite well recognized deleterious effect of the hyperthermia oncritically ill neurological patients, few investigations were performed for the issueabout the fever after ischemic stroke in intensive care unit (ICU) setting. We triedto prove the effect of hyperthermia on in-hospital outcome and the role of infectionon hyperthermia in the acute ischemic stroke patients admitted to ICU.Methods: We reviewed medical records retrospectively for the acute ischemicstroke patients admitted to ICU within 24 hours from the onset between March2004 and December 2006. The patients were assigned into normothermia, mildhyperthermia (MH, ≥37.6°C and <38.0 ≥1;, at least one time during ICU stay)and severe hyperthermia (SH, ≥38.0 ≥1;). Causes of hyperthermia were dividedinto infection and non-infectious cause. Outcomes were measured by in-hospitalmortality or long ICU stay (≥4 days). The logistic regression tests with factorspresumed to be related with hyperthermia were performed to predict the outcomes.Result: Among the 150 patients included (63.38 ± 12.13 years old, male 57.3%),MH and SH were observed in 15 and 40 patients, respectively. SH was indepen-dently related to in-hospital mortality (OR 10.3, p < 0.01) and long ICU stay (OR7.8, p < 0.01). MH was related with long ICU stay (OR 4.2, p = 0.03). Among theother factors, Glasgow coma scale was associated with in-hospital mortality (OR0.77, p = 0.02) and long ICU stay (OR 0.82, p < 0.01). The patients with infection(39 patients) was more prevalent in SH than in MH (p < 0.01) and had longer ICUstay than non-infectious group (p = 0.01).Discussion: Careful concern for the infection as well as effort for lowering bodytemperature per se is needed to the acute ischemic stroke patients admitted to ICU.




MAJOR NEUROLOGICAL IMPROVEMENT AFTER STROKETHROMBOLYSIS IN A FRENCH COMMUNITY HOSPITALB. Barroso, C. Morisset, E. Bertandeau, H. Mangon, A. Dakar, J.M. LarrieuF. Mitterrand Hospital, France

Objective: A significant advancement in the treatment of acute ischemic strokemanagement has been the use of recombinant tissue plasminogen activator (rtPA).In practice, French community hospitals experience difficulty in treating theirpatients because stroke units (SU) are missing in Neurology wards. To improvequality of care and to provide modern stroke therapy, we decided to assess whetherthrombolysis was feasible in the Emergency department of our community hospitallocated in Pau (south-west of France).Methods: Select patients were treated with rtPA and observed for 24 hours in theEmergency department. They were then transferred to the neurological departmentfor follow-up. By reference to French legislation this use of rtPA therapy wasdefined as an "off-label use". NIHSS was obtained on admission, immediatelyafter treatment and 24 hours post treatment. At that time we collected all majorneurological improvements defined as an NIHSS score equal to 0 or 1 or animprovement of more than 8 points compared to baseline.Results: Results from the first 40 patients are reported. They were treated betweenSeptember 2004 and June 2006. A total of 1169 patients were admitted for strokeduring this period. We collected 349 transitory ischemic attacks, 192 hemorrhagicstrokes and 628 ischemic strokes. A total of 3.4% of strokes was treated by rtPA. Amajor neurological improvement was present in 14 patients (35%).Conclusions: According to the infrastructural local criteria and preparatory pre-requisites described in this study, thrombolysis is a viable and feasible treatmentoption for stroke patients in Emergency departments of French community hospitalseven in lack of SU. It has been shown elsewhere that good outcome in the first 24hours is an independent predictor of good functional outcome at 3 months. So, suchan emergency organisation could be used as a compromise as long as SU are notavailable enough.


OUTCOME OF PATIENTS NOT TREATED WITH RT-PA DUE TOSPONTANEOUS IMPROVEMENTJ. Martí-Fàbregas, E. Martínez, S. Martínez-Ramírez, D. Alcolea, D. Cocho,M. Martínez-Corral, M. Marquié, M. Suárez, L.A. Querol, J.-L. Martí-VilaltaHospital de la Santa Creu i Sant Pau, Spain

Background: Rapid and significant spontaneous clinical improvement is an exclu-sion criteria for intravenous thrombolysis. However, there is controversy about theshort- and long-term outcome of these patients. We report a prospective study ofconsecutive patients.Methods: We studied patients with a focal neurological deficit admitted within3 hours of onset of symptoms with the following characteristics: 1) Duration ofsymptoms >30 minutes, 2) NIHSS score of 5 or more points, obtained either byanamnesis (when it occurred outside the hospital) or by neurologic examination.3) A spontaneous decrease in the NIHSS score to values <5 points that occurredbefore the 3 hours-limit. Improvement was assessed either by anamnesis (when itoccurred outside the hospital) or by direct examination. We used the SITS-MOSTcriteria for intravenous thrombolysis. Favourable outcome at 24 hours was definedas a NIHSS below 5 points. Favourable outcome 1 and 3 months after stroke wasdefined as a score < 2 on the Rankin scale. The diagnosis of an acute infarctiondemonstrated by neuroimaging within the study period was recorded.Results: We evaluated 15 patients, with a mean age of 69 ± 13.2 years, and80% were men. Median NIHSS score at onset was 8, and median NIHSS score atinclusion was 3. Favourable outcome was observed in 13 (87%) patients at 24h, 11(73%) at 1 month and 8/12 (67%) at 3 months. During the study period 11 (73%)patients developed a cerebral infarction. Mortality at 3 months was 13%.Discussion: The outcome of patients not given rt-PA due to spontaneous im-provement is not uniformly favourable. One third of patients had an unfavourableoutcome and 73% developed an acute infarct.


PREDICTORS OF INTRACEREBRAL HEMORRHAGE AFTERINTRAVENOUS RTPA THERAPY FOR ACUTE ISCHEMIC STROKE INCLINICAL PRACTICED. Gasecki, G. Kozera, M. Swierkocka-Miastkowska, K. Chwojnicki,B. Karaszewski, S. Szczyrba, M. Wisniewska, W.M. NykaMedical University of Gdansk, Gdansk, Poland

Background: Intravenous recombinant tissue plasminogen activator (rtPA) is aneffective therapy for acute ischemic stroke, but it is associated with risk ofintracerebral hemorrhage (ICH). Our aim was to identify baseline factors that areassociated with thrombolysis-related ICH and to assess the clinical course of thosepatients compared to patients without ICH.Methods: we analyzed 52 patients (18 women) with acute stroke treated with IVrtPA within 3 hours of stroke symptom in Medical University of Gdansk, Poland,between 2000 and 2006.Results: 2 (3,8%) patients developed symptomatic ICH, and 4 additional patients(7,7%) had asymptomatic ICH identified on a routine follow-up CT (within 22-36hours of stroke symptoms). In analyses based on clinical and radiological variables,the attributes associated with ICH were advancing age (p<0,05), early ischemic CTchanges (p<0,05), a history of atrial fibrillation (p<0,05), and elevated pre-bolusdiastolic blood pressure (p=0,06). No association between diabetes mellitus, serumcholesterol, the initial stroke severity, neurological outcome at 7 or 90 day and ICHwas found. Clinical relevant was only parenchymal type of ICH, in 2of 3 casesassociated with early neurologic deterioration.Conclusions: Advanced age, embolic stroke, elevated diastolic blood pressure andischemic changes on CT could be predictive of ICH. Only PH-ICH seems to be ofclinical significance.


THE "NO-REFLOW" PHENOMENON AFTER THROMBOLYSIS IN ACUTESTROKE - A MARKER OF POOR STROKE OUTCOME?A.Y. Jin, O. Islam, F.W. Saunders, A.M. Demchuk, D.G. BrunetUniversity of Calgary, Calgary, Canada

Background: The “no-reflow” phenomenon after thrombolysis-induced recanal-ization in acute stroke has not been well-described. We assessed the hypothesisthat reperfusion failure despite thrombolysis-induced recanalization, i.e. no-reflow,is common and is associated with a poor early stroke outcome.Methods: Patients treated with thrombolysis over a three year period (2001-2004)at Kingston General Hospital were considered in this retrospective study. Inclusioncriteria were: admission noncontrast CT (NCCT) scan with either transcranialDoppler sonography (TCD), CT angiography (CTA), or CT perfusion (CTP) scandone before thrombolysis; follow-up NCCT and CTP scans done between Days 1 to5 with either serial TCD examinations or CTA; and a modified Rankin Scale (mRS)score evaluated before hospital admission and at hospital discharge. Cerebral bloodvolume (CBV) and cerebral blood flow (CBF) maps were used to compare theaffected and contralateral hemispheres before and after thrombolysis. Reperfusionfailure was defined as any area with decreased CBV and CBF after thrombolytictherapy. Recanalization of the primary arterial occlusive lesion was evaluated byeither serial TCD examinations or follow-up CTA. Reperfusion failure despiterecanalization of the primary arterial occlusive lesion was classified as “no-reflow”.Results: 20 patients were included in this study. 19 patients showed recanalizationof the primary arterial occlusive lesion at follow-up. Of these patients, 7 (37%)showed reperfusion failure on follow-up CTP scan. Among patients with recanal-ization, those with no-reflow had a median hospital discharge mRS of 4 (range 2 to6), compared to a median mRS of 1 (range 0 to 4) in patients without no-reflow(Mann Whitney U test: p (two-tailed) = 0.002).Discussion: The no-reflow phenomenon is common despite thrombolysis-inducedrecanalization and is associated with a poor early stroke outcome. Possible causesof no-reflow may include distal embolization, arterial reocclusion, and persistentbranch vessel occlusion.


Acute stroke: treatment concepts

1 Acute stroke: treatment concepts

OXIDATIVE STRESS, DOCUMENTED BY DETERMINATION OF MDA, INACUTE ISCHEMIC STROKE SUBTYPES, ACCORDING TO TOASTCRITERIAA. SimionFaculty of Medicine and Pharmacy/Clinical Hospital of Neurology andPsychiatry, Oradea, Romania

Background: In ischemic stroke, oxidative stress has been shown to worsen theextent of cerebral injury. Our study tries to correlate oxidative stress with thesubtype and outcome of ischemic stroke.Material and method: We studied a consecutive series of 104 ischemic strokes,evaluated with CT and/or MRI or MRA, Duplex sonography, transthoracic echocar-diography, and electrocardiography. Strokes were divided into large-artery disease(LAD), small-artery disease (SAD) and cardioembolism (CE), according to theTOAST criteria. Cases with uncertain or unknown etiologies were excluded. Eachpatient was scored on the NIHS scale at admittance and had the Barthel index(BI) evaluated at discharge. Beside a complete laboratory evaluation, oxidativestress was assessed by measuring the seric malondialdehyde (MDA) levels with thetiobarbituric acid method at admittance (day 1), and on days 3 and 7.Results: We found different patterns for the course of oxidative stress in the strokesubtypes. By day 3 we obtained significantly (p<0,001) higher level of MDA in allischemic subtypes (view Table 1), but by day 7 MDA levels were increased only inCE (p<0,001) whilw in LAD they decreased (p<0,02) and showed non-significantvariations (p>0,05) in SAD (view Figure 1). MDA values do not correlate withNIHSS, or with BI. Complications, especially infectious ones, raise the MDA valueson the second and third determination, but only bronchopneumonia significantlyinfluenced the outcome.

Table 1. MDA values in stroke subtypes and overall

LAD SAD CE Ischemic stroke

MDA2 2,43±0,31 1,83±0,34 2,17±0,32 2,2±0,4MDA3 4,18±0,43 3,86±0,46 3,07±0,43 3,7±0,6MDA3 3,13±0,43 4,06±0,64 4,61±0,52 3,9±0,5

Fig. 1

Conclusions: We presume that the high levels of MDA on day 7 in CE could betied to reperfusion. If so, antioxidant therapy would be most beneficial if given topatients with embolic strokes or after thrombolysis.


SYSTEMATIC REVIEW OF THE EFFECT OF CONTRAST AGENTS ONRECANALISATION RATE AFTER INTRAVENOUS THROMBOLYSIS FORACUTE STROKEK.A. Dani, K.W. MuirUniversity of Glasgow, Glasgow, United Kingdom

Background: Recent assertions that contrast agents impair fibrinolysis are basedon limited evidence from animal models. We undertook a systematic review

of recanalisation rates after intravenous (IV) thrombolysis for stroke comparingstudies where contrast was administered to those where it was not.Methods: Search results for MEDLINE and Embase from inception to October2006 were screened and studies reporting recanalisation rates for IV thrombolysisfor anterior circulation stroke were selected. Additional data were sought fromtwo authors. We compared recanalisation rates for contrast (CS) and non-contrast(NCS) studies.Results: We identified 31 studies (7 CS, 24 NCS). Non-ionic contrast was admin-istered for CT perfusion images (CTp) in 2 studies (40mls, n=1; 50mls, n=1), CTp+/- angiography (CTa) in 1 study (50mls +/- 50mls), and CTp and CTa in 1 study(weight adjusted; max 140mls). 3 studies involving catheter angiograms did notspecify type or volume of contrast. The mean time limit for thrombolysis was 6.1h(SD 2.1) in CS and 4.9h (SD 1.6) in NCS. Recanalisation rates were assessed byCTa, magnetic resonance angiogram and transcranial doppler ultrasound. Recanali-sation was assessed late (>24h) in 6 CS, and 15 NCS. Precise occlusion site wasspecifically indicated in 4/7 CS and 19/24 NCS.Recanalisation was non-significantly more frequent in CS (56/89, 63%) comparedto NCS (841/1496, 56%) OR 1.32 (95% CI 0.85, 2.06). This remained true for laterecanalisation (OR 1.34, 95% CI 0.79, 2.29). and for MCA M1 or M2 occlusions(OR 1.17, 95% CI 0.47, 2.91).Discussion: This was an indirect comparison and not a randomised study, and there-fore limited. However, we found no evidence that contrast impaired recanalisationrates with IV fibrinolytic therapy.


PROGNOSTIC FACTORS OF THROMBOLYTIC THERAPY IN HYPERACUTEISCHEMIC STROKEJ.H. Rha, B.N. Yoon, K.H. Ji, J. LeeInha University Hospital, Seoul, South Korea

Background: To investigate the prognostic factors of intravenous thrombolysis, weevaluated 121 consecutive patients treated with IV tPA. Methods: Demographic andclinical profiles, laboratory results, transcranial Doppler, and brain imaging wereevaluated. Clinical assessment was done by National Institutes of Health StrokeScale (NIHSS) for one week, and by modified Rankin Scale (mRS) at baseline andthree months. Early improvement was defined as the complete resolution of theneurological deficit or an improvement of 4 or more points by NIHSS within 24hours of the stroke onset, and good outcome as mRS score of 2 or less at threemonths. We assessed the possible relationship of the factors with early improvementand good outcome, and also analyzed the correlation of TCD grade with NIHSSscore. Comparisons of variables were performed using Fisher’s exact test, t-test andMann-Whitney test. The predictors of early improvement and good outcome wereanalyzed by logistic regression, and the correlation of TCD grade and NIHSS scorewere analyzed by Spearman correlation.Results: On univariate analysis, younger age, absence of abnormal CT findings(hyperdense middle cerebral artery sign [HMCAS], focal hypodensity in the totalMCA territory > 33%) were significantly associated with early improvement.Good outcome was associated with younger age, lower levels of baseline NIHSSscore, mean blood pressure, fasting glucose, lipoprotein (a) [Lp(a)], and absenceof abnormal CT findings. Multivariate analysis revealed age < 63 years and noHMCAS as independent predictors of early improvement. Thrombolysis in brainischemia grade by TCD monitoring significantly correlated with NIHSS score for24 hours.Conclusions: These results suggest that younger age, normal CT findings areimportant prognostic factors of acute thrombolytic therapy, whereas age, CTfindings, baseline NIHSS, blood pressure, blood sugar and Lp (a) level might beassociated factors of long term outcome. TCD can be a useful indicator of clinicalimprovement.


STATES OF LEPTOMENINGEAL COLLATERALS AND RESPONSE TOTHROMBOLYSIS IN PATIENTS WITH ACUTE MCA INFARCTSK.H. Kang, H.C. Park, E.H. Kim, Y.S. Kim, C.K. Suh, Y.H. HwangKyungpook National University Hospital, South Korea

Background: MR-based thrombolysis using DWI/PWI mismatch is feasiblemethod in selecting thrombolysis candidate. However, state of leptomeningeal col-laterals in conventional angiography (CA) may also affect response to thrombolysisand clinical outcome. The purpose of this study is to correlate state of collateralswith response to thrombolysis.Methods: We retrospectively analyzed 16 patients from May to December 2006who performed CA for Intra-arterial thrombolysis (IAT) in presumed MCA occlu-

Poster SessionAcute stroke: treatment concepts


sion with DWI/PWI mismatch. Angiographic states of leptomeningeal collateralswere defined as minimal (retrograde flow to M4 segment), moderate (retrogradeflow to M3 or M2 segment), and maximal (retrograde flow to distal to site ofocclusion). IAT was done using urokinase or tPA infusion and/or mechanicalthrombus crushing.Results: Mean values of onset to door and door to IAT were 231.9±53.3 min and143.4±11.5 min. Among them, 11 patients regained angiographic recanalization(69%, TIMI Grade 2 or 3) and 6 patients had good clinical outcome (38%, mRS 0-1at 1-3 months). There was no statistically significant results in recanalization rateand clinical outcome among each collateral groups [response to thrombolysis - 3/3in maximal, 4/5 in moderate, and 4/8 in minimal, (P=0.23); good clinical outcome- 2/3 in maximal, 3/5 in moderate, and 1/8 in minimal, (P=0.12)]. However, therates of recanalization and good clinical outcome were much higher in maximalcollateral group.Discussion: Unfortunately, we did not show statistically significant results abouteffect of collateral circulation on response to thrombolysis and clinical outcome.However, we found trend that good collateral circulation is related to recanalizationsuccess and better clinical outcome. In future, standardized method of collateralgrading system is needed for multi-center controlled study.


PERSONAL EMAILS: A SIMPLE MEASURE TO IMPROVE THE QUALITY OFCARE IN THE STROKE PATIENTSM.F. Bellolio, R. Kashyap, L. Vaidyanathan, S. Enduri, A.M. Hoff, A.S. Yassa,S. Mishra, R.D. Brown, W.W. Decker, L.S. SteadMayo Clinic College of Medicine, Rochester, USA

Background: Embolic or thrombotic arterial occlusion is a frequent cause ofcerebral infarction, making antithrombotic therapy an important part of the care instroke patients. Aspirin (ASA) reduces the risk of early recurrent ischemic strokewhen given within 48 hours of initial symptom onset.We hypothesized that sending a friendly remainder by email to the consultant whosaw a patient with Stroke or TIA in the Emergency Department (ED) and for anyreason did not give ASA during the ED stay, could improve the rate of patientsreceiving aspirin.Methods: A consecutive cohort of patients presenting into the ED with a suspecteddiagnosis of TIA or Stroke were prospectively enrolled. We excluded patients withsuspected diagnosis of intracerebral hemorrhage and those receiving thrombolytics.The intervention was to track ASA administration and send out timely follow up tothose who did not provide the drug. This was done via individual (rather than group)email, specifically stating the name and clinic number of the patient, and notinglack of aspirin administration as well as lack of documentation for not doing so.Results: Our study group was 64 patients in the pre-intervention period (Aug-Sept)and 59 post-intervention (Oct-Nov).During the pre-emails period 43.8% of the patients received ASA, and 66.1% aftersending the emails (p=0.013)After the intervention, the patients with stroke or TIA were 2.51 times more likelyto receive aspirin in the ED (95%CI 1.21-5.21)Discussion: This simple method of personal emails has been improving not onlythe rate of patients who receive aspirin, but also the quality of the documentation(reasons why the patient is not a candidate for ASA, such as suspected hemorrhageor ASA given by the pre-hospital services personnel) and has been maintained overthe months.


CAN WE USE THE ABCD SCORE TO SCREEN PATIENTS FOR A WEEKLYTIA CLINIC? RETROSPECTIVE APPLICATION OF THE SCORE TO ASSESSIT’S SUITABILITYB. Brady, M. Sekiguchi, B. Silke, J. HarbisonStroke Service, St James’s Hospital, Dublin, Ireland

Background: Prior to the establishment of a weekly TIA in St James’s Hospitalin 2006 suspected TIA patients were routinely admitted for investigation. Indeveloping protocol for the clinic we considered the ABCD score (1) as a potentialmeans of screening patients for the clinic rather than admission. The original paperreports a score of 5 or 6 being associated with a recurrence rate of >25% implyingneed for immediate admission. To assess the scores utility we applied the scoreretrospectively to a sample of patients admitted with suspected TIA.Method: 100 sets of notes on patients admitted with suspected TIA were reviewedby 2 doctors. Diagnosis was reviewed, ABCD score applied from available admis-sion data and prior and subsequent admissions with TIA/stroke identified. Changesin management resulting from admission were noted.

Results: Of the 100 subjects, 83 were felt to have suffered a cerebrovascularevent. Twenty eight of the 83 subjects (34%) had symptoms lasting >24 hours. Adecision had been made that such patients should be admitted. Of the 55 remainingsubjects, 15 (27%) had suffered other events in the preceding month (1-5 events,median 1) necessitating admission. Of the 40 left, 19 (47%) had an ABCD scoreof 5 or 6. Two patients had subsequent events in the following month. One hadan ABCD score of 3 and the other a score of 5. The patient with the score of 3was young, normotensive but suffered a prolonged event (>1 hour). None of thepatients underwent intervention other than the commencement of anti-platelet andantihypertensive therapy. None had suffered an intracerebral haemorrhage.Conclusion: Use of the ABCD score 1-4 as a means of selecting suitable subjectsfor outpatient management would have prevented the admission of 38% of thereferrals (non-strokes and TIAs). It would have missed one subject who suffered arecurrent event, however admission of that subject did not alter his outcome. Wedecided to adopt the ABCD score but not use it in isolation and to admit subjectswith events >30 minutes duration.


ANTICOAGULATION AND THE RISK OF ICH AFTER CARDIOEMBOLICSTROKEH. Hallevi, K.C. Albright, A.D. Barreto, S. Martin-Schilde, A. Khaja,E.A. Noser, N.R. Gonzales, K. Illoh, J.C. GrottaUniversity of Texas-Houston, Fannin, TX, USA

Background: Heparin has not been shown to be effective in reducing mortalityand morbidity after acute cardioembolic stroke, however anticoagulation (AC)eventually needs to be instituted for secondary stroke prevention. We aimed tostudy the timing and mode of starting AC in cardioembolic stroke patients.Methods: We conducted a retrospective analysis of all cardioembolic strokes caredfor by our Stroke Team over 3 years. Patients were monitored in our Stroke Unit.Neuroimaging was done on admission, at 24 hours and with any neurologicaldeterioration.Results: We included 204 patients with cardioembolic stroke in the analysis.Full dose AC with IV Heparin or Low Molecular Weight Heparin (LMWH) wasgiven to 73 patients. Warfarin was started subsequently in 87% (63/73). Low doseAC (DVT prophylaxis dose) with or without aspirin was given to 131 patients.Warfarin was given subsequently to 26.7% (35/131). There were 22 (10.8%) casesof asymptomatic hemorrhagic transformation (HT). All but one occurred during thefirst three days. Three patients experienced symptomatic HT 6 to 12 days from theirstroke. In all three, warfarin was started while patients were “bridged” with fulldose LMWH and aspirin (3/21, 14.3%, p=0.001). Two cases of serious systemichemorrhage occurred among patients treated with IV heparin (2/48, 4.2%, p=0.054).Overall acute anticoagulation was associated with 6.8% (5/73) serious bleeding(p=0.05). Recurrent strokes occurred in two cases despite effective anticoagulationin one (0.1%, p=0.53). There were no cases of hemorrhage in patients treated withlow dose AC and warfarin.Discussion: Symptomatic HT after cardioembolic stroke occurred late and wasassociated with aggressive anticoagulation and simultaneous ASA use in our series.Early asymptomatic hemorrhage seems unrelated to anticoagulation. Low doseAC appears safe, even in the setting of asymptomatic HT. Our data suggest that“bridging” patients with full dose LMWH until anticoagulated with warfarin maynot prevent early stroke recurrence and is associated with an unacceptable risk ofsymptomatic HT.


EFFECTIVENESS OF THROMBOLYSIS IN PATIENTS OVER 80 YEARS OFAGEC. Rueckert, T. Staudacher, D. BengelSt. Elisabeth Hospital, Ravensburg, Ravensburg, Germany

Background: Despite the fact that the clinical trials leading to the approval ofthrombolysis excluded patients over 80 years, in many centers this therapy isoffered to patients above this age.Methods: In our series of 324 patients treated with thrombolysis between 12/2000and 01/2007 we analyzed the outcome of all 77 patients over the age of 80 (max:92 years).Results: Overall median age was 71 years; 45,2% were women, 54,8% were men;in the subgroup of over 80 years 59,7% were women, 40,3% were men. 29,4% of allpatients but 40,8% of the patients 80+ were admitted to the hospital within 1 hourof onset of symptoms. There was no difference in mean door-to-needle-time (55minutes), mean time from onset of symptoms to therapy (131 minutes), and strokeseverity at onset (mean National Institute of Health Stroke Scale=NIHSS 16).


57,8% of all patients versus 77,5% of patients over 80 had a cardio-embolic origin,17,7% (8,4%) an arteriosclerotic origin.At discharge mean NIHSS was 10 (overall) versus 11 (80+). While medianModified Rankin Stroke Scale (mRS) at discharge was 4 in both groups, there wasa difference at three months: 2,7 (overall) versus 3,4 (80+). Accordingly, 25,6% ofall (14,7% of 80+) have reached Rankin 0/1, 36,6% of all (35,3% of 80+) havereached Rankin 2/3.Overall in-hospital mortality was 18,5%, in the group of 80+ 29%, but the inci-dence of symptomatic intracerebral hemorrhage was only slightly different: 3,7%vs. 5,3%.Discussion: In good correlation with previous reports our data show a slightlybut not significantly higher incidence of symptomatic intracerebral hemorrhage inelderly patients. In-hospital mortality is higher and the proportion of favourableoutcome lower than in younger patients. Nevertheless more than 20% of treated pa-tients over 80 years reached Rankin 0-2 and about 50% Rankin 0-3. In our opinionthese data show the practicability of thrombolysis and support the effectiveness ofthis therapy in selected patients over 80.


PREVIOUS TREATMENT WITH ANGIOTENSIN II RECEPTOR BLOCKERSCOULD PLAY A POSSIBLE PROTECTOR EFFECT IN ACUTE STROKEM.A. Ortega-Casarrubios, B. Fuentes, B. San José, M.J. Aguilar-Amat, I. Ybot,P. Martínez, E. Díez-TejedorStroke Unit, Department of Neurology, La Paz University Hospital, UAM,Madrid, Spain

Background: Previous studies with angiotensin II receptor blockers (ARB) havedemonstrated a protective effect in spontaneously hypertensive rats from cerebralischemia. It is not known if all the hypotensor drugs share this property. Our goal isto analyse the impact of pre-stroke use of hypotensor drugs in stroke severity andoutcome.Methods: Observational study from the Stroke Unit data bank of the Departmentof Neurology, with inclusion of consecutive stroke patients (January 2000-October2005). Parameters analysed: Risk factors, previous hypotensor drug treatment,severity on admission(Canadian Stroke Scale, CSS),in-hospital complications,mortality and functional state at discharge (Modified Rankin Scale, mRS)Results: 1738 patients were included in the study, 55,9% were men. Averageage: 69,61 ±12 years. 63,3% had high blood pressure and 27,9% DM. 39,4%received hypotensor drugs (90,2% of them for hypertension). Patients with ARBpresented lower stroke severity on admission (EC ≤ 16 vs 29,4%, p=0,011) andbetter functional state at discharge (mRS ≤ 78 vs 63,6%, p=0,008) The multivariatelogistic analysis showed that ARB pre-treatment was a predictive factor of lowerstroke severity, independent of age, sex and stroke subtipe (OR 0,4; 95% IC 0,2-0,7)Patients treated with diuretic drugs had higher severity on admission (EC ≤ 34,5vs 27,3, p=0,023) and worse outcome (mRS ≤ 58,9 vs 65,8%, p= 0,044). Othergroups of hypotensor drugs did not show significant benefit on stroke severity.Conclusions: Previous treatment with ARB was associated with a lower acutestroke severity on admission and better evolution, being an independent predicitivefactor of lower stroke severity. More studies are needed to confirm this posibleprotector role.


LACK OF KRINGLE 2 DOMAIN AND HIGH FIBRIN SPECIFICITYDIFFERENTIATE THE NOVEL PLASMINOGEN ACTIVATORDESMOTEPLASE FROM RT-PAD.B. Bharucha, M.K. Pugsley, K.-U. Petersen, M. SoehngenForest Laboratories/PAION Deutschland GmbH, Jersey City, NJ, USA

Background: Thrombolytic therapy with recombinant tissue plasminogen activator(rt-PA) is effective in treating acute ischemic stroke (AIS) within the first 3 hoursafter symptom onset. Desmoteplase, or rDSPAα1 (recombinant Desmodus SalivaryPlasminogen Activator α1), is a novel plasminogen activator (PA) shown to treatAIS up to 9 hours post-stroke onset with a positive risk:benefit ratio (in a PhaseIIa study in patients with penumbra) not shown in previous trials with rt-PA whentreating patients beyond 3 hours. The observed clinical benefit of rDSPAα1 mayrelate to its unique structural features.Methods: Pharmacological and structural characteristics of desmoteplase are re-ported using in vitro studies. Fibrin specificity was determined for desmoteplaseand rt-PA using an enzymatic assay for the kinetics of plasminogen activation.Neurotoxicity was determined using ischemic murine models.Results: rDSPAα1 is a glycosylated serine protease structurally similar to rt-PA(with finger, epidermal growth factor-like, kringle 1, and serine protease domains),

but it lacks the plasmin-sensitive cleavage site and the lysine-binding kringle 2domain found in rt-PA. The lack of the kringle 2 domain may explain the fact thatrDSPAα1 has the highest fibrin selectivity among PAs, with a 100,000-fold increasein catalytic activity compared to a 550-fold increase for rt-PA and an absence ofneurotoxicity in animal models of ischemic and hemorrhagic stroke.Discussion: The unique structural features of rDSPAα1 suggest an advantageousrisk:benefit ratio and may explain the observed favorable clinical results includingthe potential for use in later time windows.


BRIDGING WITH GPIIB/IIIA-RECEPTOR-ANTAGONISTS COMBINDEDWITH INTRA-ARTERIAL PHARMACOMECHANICAL THROMBOLYSIS INISCHEMIC STROKER. Dabitz, U. Leppmeier, L. Fuhry, V. Collado-Seidel, R. Michailow,K. Schoeneboom, S. Triebe, H. Gunselmann, G. Ochs, D. VorwerkKlinikum Ingolstadt, Ingolstadt, Germany

Background: The prognosis of ischemic stroke due to occlusion of the internalcarotid artery, the middle cerebral artery in M1 and the basilar artery is even afterintravenous rt-PA usually very poor.Patients: We report in a retrospective analysis about 76 consecutive patients (39m,37f; age 64 ± 12.2 years), who were treated with GPIIb/IIIa-Rezeptor-Antagonistscombined with pharmacomechanical intra-arterial thrombolysis. 25 pat. with tan-dem occlusion of the ICA and MCA, 26 pat. with occlusion of the MCA and 25patients with basilar artery occlusion.The mean NIH-Stroke-Scale was 16.4, analogical to a mean modified Rankin Scale(mRS) of 4.7.Materials: After exclusion of ICH 39 patients were bridged with Abciximab (dueto the Abestt trial) and in 37 cases with Tirofiban followed by an intra-arterialthrombolysis with rt-PA.Results: At the time of discharge the mean mRS was 2.8.23 pat. mRS: 0-2 (=31%); 2 pat. mRS: 3 (=3%); 21 pat. mRS: 4 (=28%); 6 pat.mRS: 5 (=8%).At the time of re-evaluation after rehabilitation the mean mRS was 1,8.32 pat. mRS: 0-2 (=42%); 13 pat. mRS: 3 (=17%); 5 pat. mRS: 4 (=7%), 2 patmRS: 5 (=3%).24 of 76 (=32%) patients died nevertheless23 pat. (=30%) showed ICH in the CT-scan, 10 (=13%) of them were symptomatic.8 of these 10 patients showed already malignant infarction at the time of bleedingdue to unsuccessful rekanalisation. One patient died due to perforation of thebasilar artery.Conclusion: Bridging with GPIIb/IIIa-Rezeptor-Antagonists combined with intra-arterial pharmakomechanical thrombolysis is feasible and may help to reduce thelethality and morbidity of ischemic stroke due to occlusions of the ICA, MCA andBA. The rate of complications and intracranial haemorrhage is in the range of thosereported in other cohorts


WHY ARE EARLY ADMITTED STROKE PATIENTS EXCLUDED FROM TPATHERAPY?S. Debiais, I. Bonnaud, B. Giraudeau, D. Saudeau, D. Perrotin, B. de Toffol,A. AutretCHRU Tours, Tours, France

Introduction: Our University hospital (300 000 inhabitants) receives patients whomay benefit from IV thrombolysis from the whole region. Since June 2003, anacute stroke network comprising 2 beds of admission and thrombolysis was createdin the intensive care unit.Objective: To assess conditions of treatment with IV thrombolysis and to determinewhy stroke patients admitted within 3 hours of symptoms onset (SO) are not treatedwith IV TPA.Methods: During 18 months were prospectively recorded the following data foreach patient: demographic data, delays of arrival and imaging, treatment andoutcome. For the patient arrived in the first 3 hours, the reason why IV TPA wasnot administered was noticed.Results: During 18 months, 364 patients were admitted, with a median delay ofadmission after SO of 2 h 50. Two hundred patients were admitted within threehours and among them, 17 (8,5%) patients received IV TPA. Among the patientsadmitted in the first 150 minutes, the main reasons for exclusion were: mild strokeor clinical improvement (NIHSS < 6) for 48 patients, a non vascular diagnosis in41 patients, aged older than 80 yo for 22, intracerebral hematoma for 17, NIHSS >

23 for 13. For 13 patients the only cause was network internal dysfunction.



Discussion: In our institution, 18% of ischemic stroke patients are treated with IVTPA. For the early admitted patients, main reasons of therapy exclusion are nonvascular diagnosis, mild stroke with NIHSS< 6, and age > 80 yo, as found inprevious studies (Barber 2001; Huang 2006). Improvement of the pre-admissionselection, decrease of the delays of arrival and correction of the network dys-functions could increase dramatically the proportion of treated patients. Moreover,some patients with mild or improving deficits could also benefit from thrombolytictreatment.


NOVEL APPROACH IN ACUTE STROKE MANAGEMENT THROUGHL-LYSINE MONOHYDROCHLORIDE INDUCED ANGIOGENSIS ANDREVASCULARIZATIONS.C. Mukhopadhyay, G. Guha, M. Alam, A. Mukherjee, M. HashiniGreen Cross Therapeutics Pvt. Ltd., India

Background: The efficacy and safety of L-Lysine Monohydrochloride (LMH) asangiogenic agent was studied in acute ischaemic stroke management.Material and methods: 120 patients (mean age 61.3 yrs) with CT/MRI evidenceof ischaemic stroke was studied. 80 received LMH,1gm I.V. 6 hrly x 7 days, 40did not receive LMH. Both LMH and non- LMH grs. did not receive thrombolytictherapy but had routine stroke management. Modified NIHSS,MRS and BI wereused to assess clinical outcome, at baseline, 1 wk, 6wk, 3 mths and 6 mths. Basichaematological biochemical and urine analyses were done at baseline, 1 wk, 1 mth.MRI was done on a 1.5T whole body imager for T1 and T2 weighted, FLAIR,DWI, Angiography imaging sequences at baseline, 7th day and 6 wks. Adverseevents noted.Result: Patient Inclusion time for therapy varied between 6-98 hrs. There was nomortality or significant change in biochemical, hematological and urine analyses atany stage of therapy. Mean baseline Glasgow Coma Scale Score was 13.9. Imagingrevealed acute infarct in anterior, middle and posterior cerebral artery territory, butmost in MCA area (68.75%). Clinical outcome by NIHSS in LMH gr. revealed15.05 ±3.52 at baseline, 7.95 ± 2.31 at 1 wk, 4.14 ±1.16 at 6 wks., 1.81 ± 0.88 at6 months, vs. to 15.17 ± 3.50 (baseline)12.17 ± 2.92 (1wk), 8.61 ±1.85 (6wks),4.33 ± 0.84 (6months) in non LMH gr. with P Value 0.08(1wk), 0.02 (6 wk), 0.01(6 months). Decrease in NIHSS become more pronounced as time progresses inLMH gr. Parallel improvements were seen in B.I at 90 days (≥95) and MRS (≤1).Infarct area in DWI/T2 imaging at 7th day showed remarkable improvement inLMH gr. compared to non- LMH gr.Discussion and conclusion: LMH significantly increases both short term and longterm clinical, functional and imaging recovery and was due to revascularizationthrough angiogensis.Both anterior and posterior circulation improved equally. LMHcan be given with much wider therapeutic window (upto 96 hrs) and has no adverseeffect.


COMPARISON OF INTRA-ARTERIAL AND INTRAVENOUSTHROMBOLYSIS FOR ISCHEMIC STROKE WITH HYPERDENSE MIDDLECEREBRAL ARTERY SIGNH.P. Mattle, M. Arnold, D. Georgiadis, C. Baumann, K. Nedeltchev,D. Benninger, L. Remonda, C. von Büdingen, G. Schroth, R.W. BaumgartnerUniversity Hospitals Bern and Zurich, Bern, Switzerland

Background: It is unclear whether intra-arterial (IAT) or intravenous (IVT) throm-bolysis is more effective for ischemic stroke with hyperdense middle cerebralartery sign (HMCAS). The aim of this study was to compare IAT and IVT in suchpatients.Methods: Comparison of data from two stroke units with similar management ofstroke associated with HMCAS, except that one unit performed IAT with urokinaseand the other IVT with plasminogen activator. Time to treatment was up to 6 hoursfor IAT and up to 3 hours for IVT. Outcome was measured by mortality and themodified Rankin Scale (mRS), dichotomized at three months into favorable (mRS0-2) and unfavorable (mRS 3-6).Results: 112 patients exhibited a HMCAS, 55 of 268 patients treated with IAT and57 of 249 patients who underwent IVT. Stroke severity at baseline and patient agewere similar in both groups. Mean time to treatment was longer in the IAT group(244 ± 63 minutes) than in the IVT group (156 ± 21 minutes; p=0.0001). However,favorable outcome was more frequent after IAT (n=29, 53%) than after IVT (n=13,23%; p=0.001) and mortality was lower after IAT (n=4, 7%) than after IVT (n=13,23%; p=0.022). After multiple regression analysis IAT was associated with a morefavorable outcome than IVT (p=0.003) but similar mortality (p=0.192).Conclusion: Intra-arterial thrombolysis was more beneficial than IVT in stroke

associated with HMCAS, even though IAT was started later. Our results indicatethat a randomized trial comparing both thrombolytic treatments in patients withmiddle cerebral artery occlusion is warranted.


IS INTRAVENOUS TPA TREATMENT BENEFICIAL IN ACUTE ISCHEMICSTROKE RELATED TO INTERNAL CAROTID DISSECTION?B. Fuentes, M. Alonso de Leciñana, J. Masjuán, J. Egido, P. Simal,F. Díaz-Otero, A. Gil-Nuñez, E. Díez TejedorUniversity Hospitals La Paz, Ramón y Cajal, Clínico San Carlos and GregorioMarañón, Madrid, Spain

Background: Small series reported the safety of intravenous tPA treatment inacute ischemic stroke (IS) related to extracranial internal carotid artery dissec-tion (eICAD). However, no studies analysing specifically the posible benefits onoutcome are available.Methods: Multicentre, prospective study conducted in 4 university hospitals. Con-secutive IS patients were included. Stroke severity (NIHSS) and 3-months outcome(mRS) were compared: (1) tPA-treated patients with IS related to eICAD vstPA-treated patients with other causes of stroke; (2) tPA-treated vs non tPA-treatedeICAD patients.Results: 265 IS patients received intravenous tPA (7 of them with eICAD). Therewere no diferences in baseline NIHSS between patients with or without eICAD(14.3 vs 14.3; ns). However, NIHSS scores at 24 h and day 7 were significantlyworse in eICAD patients (17 vs 9.6 at 24h; 15.6 vs 7.3; p<0,05). No eICADpatients developed a significant improvement at 24h (decrease in NIHSS ≥ 8points) as compared to 67 (32%) of patients with other IS causes. When comparingtPA-treated eICAD (n=7) with non-treated eICAD patients (n=7), a trend to higherimprovement in 24h and day 7 was found in the non tPA-treated eICAD group(NIHSS 7.6 vs 17 at 24h; 6.4 vs 15.6;p=0.205) with no differences in baselineNIHSS. At 3 months, 80% of tPA-treated eICAD and 20% of non-treated eICADpatients were dependent (mRS>2).Conclusions: Although intravenous tPA treatment in IS related to eICAD seems tobe safe, the benefit on outcome is significantly minor than in other causes of IS,and possibly worse than in non tPA-treated eICAD.


AN EMERGENCY CLINICAL PATHWAY FOR THE MANAGEMENT OFCRITICAL STROKE PATIENTS: RESULTS OF A RANDOMISED CLINICALTRIAL IN THE LAZIO REGION (ITALY)A. De Luca, D. Toni, L. Lauria, M.L. Sacchetti, M. Barbolini, E. Puca, M. Ferri,M. Prencipe, G. GuasticchiPublic Health Agency, Lazio Region, Rome, Italy

Background: Emergency Clinical Pathways (ECP) may play a crucial rule in themanagement of critical stroke patients. Objects To evaluate the effectiveness ofintroducing an ECP for the management of critical stroke patients in the emergencysystem of Lazio region (Italy).Methods: A cluster-randomized controlled trial (ISRCTN41456865) was designedto compare the practice of a test group of health professionals (HP) pertaining toEmergency Medical Services (EMS) and to Emergency Rooms (ERs), trained touse the ECP, with that of non trained EMS and ERs control groups. Groups werecompared by chi2 or Fisher’s exact tests.Results: the two groups were similar at baseline as type and number of EMSambulances and ERs. Over six months in 2005, 3298 suspected stroke patientswere enrolled (1353 in the test groups: 573 by EMS and 780 by ERs; 1945 inthe control groups: 485 by EMS and 1460 by ERs). Both the test groups referredto our hospital more suspected stroke patients than the control groups: EMS:219(38.2%) vs 8 (1.6%) (p<0.05); ERs: 147 (18.8%) vs 116 (7.9%) (p<0.05).Confirmed ischemic stroke were (test groups: EMS =70, Ers=26; control groups:EMS=4, ERs=13). Among ischemic stroke patients eligible for i.v. thrombolysis(test groups: EMS=19, ERs=17; control groups: EMS=2, Ers=10), those referredby the test groups were treated more frequently than those of the control groups(EMS: 8 (42%) vs 0 (p>0.05); ERs:7 (41%) vs 2 (20%) (p>0.05).Discussion: Adherence to the ECP improved the appropriateness of stroke patientreferral and treatment in the SU, particularly by the EMS. Hence, the educationalprogram on early detection and timely transportation of stroke patients to theappropriate ward will be extended to all emergency health personnel.



REMOTE CEREBRAL HEMORRHAGE IN PATIENTS TREATED WITHINTRAVENOUS RT-PAJ. Martí-Fàbregas, E. Martínez, M. Marquié, D. Alcolea, D. Cocho,M. Martínez-Corral, S. Martínez-Ramírez, M. De Juan, R. Marín,J.-L. Martí-VilaltaHospital de la Santa Creu i Sant Pau, Spain

Background:The frequency, clinical presentation and prognosis of remote cerebralhemorrhage (rPH) after thrombolysis is poorly known. We report our experience inpatients with ischemic stroke treated with intravenous rt-PA.Methods: A retrospective review of consecutive patients treated at our Hospitalfrom 1999 to 2006, according to the ECASS II and the SITS-MOST (since 2004)criteria. A control Computed Tomography scan was routinely obtained in allpatients within the first 36 hours of treatment. Cerebral hemorrhagic complicationswere classified according to the ECASS study in hemorrhagic infarction (HI-1and HI-2) and parenchymal hematoma (PH-1 and PH-2). rPH was defined asany extraischemic hemorrhagic lesion observed in the control CT. Neurologicworsening was defined as an increase in more than 3 points on the NIHSS score. Afavourable outcome was defined as a score <2 on the Rankin scale.Results: We studied 163 patients (mean age 67.6 ± 11.8 years, 57% of them weremen). The frequency of hemorrhagic complications was: HI-1 (2.4%), HI-2 (4.9%),PH-1 (3.6%), PH-2 (2.4%), rPH (2.4%), rPH+PH-2 (0.6%). Patients with rPH(n=5) had a mean age of 70.4 ± 8.1 and 40% were men. The median NIHSS scorewas 15 and mean time to treatment was 134 ± 47.6 minutes. rPH were multifocalin 2, single in 2, and associated with a PH-2 in one patient. The location of rPHwas lobar in 4 patients and in brainstem in one patient. rPH were symptomatic in4 patients and asymptomatic in one. The neurologic worsening occurred 8, 14, 17and 30 hours after rt-PA. The outcome was unfavourable in all patients, with 3deaths.Discussion: Remote parenchymal hemorrhage is an uncommon complication afterrt-PA (3%). It is usually lobar and symptomatic and has an unfavourable outcome.


A RETROSPECTIVE CLINICAL CASE-NOTE STUDY OF THE NUTRITIONALMANAGEMENT PRACTICES OF STROKE PATIENTS IN A SCOTTISHPOPULATIONS. Ray, B. Al Falasi, P. Rana, M.A. Haleem, M. Rajput, S. AtkinsonUniversity of Dundee, Dundee, United Kingdom

A Retrospective Clinical Case-note Study of the Nutritional Management Practicesof Stroke Patients in a Scottish PopulationBackground: Stroke is a leading cause of death and adult disability. Feeding instroke remains important as a number of stroke patients are undernourished onadmission and nutritional status declines during hospital stay, with increased mor-bidity and mortality. The “Scottish Intercollegiate Guidelines Network”, “QualityImprovement Scotland” and “Council of Europe” provide guidance for nutritionalmanagement of patients with dysphagia. This study aimed to assess the impact ofguidelines on the nutritional management of stroke patients.Methods: Retrospective case note analysis was done for ischemic strokes admittedto Ninewells Hospital, Dundee. This audit-type study looked at a 3-month periodprior to the establishment of guidelines and compared with 3 months post guideline.Results: There were 126 and 204 suspected cases of stroke in Periods 1 and 2respectively. The corresponding number of CT diagnosed strokes was 78 and 107respectively. Out of these, 33 records were retrieved for patients in Period 1 and 43in Period 2. The weight recording rates were 27% in Period 1 vs. 35% in Period 2.One- fifth of those weighed were monitored in Period 1 compared to one-third inPeriod 2.Discussion: There appear to be specific barriers to the implementation of guidelinesin daily practice. A continuing clinical nutrition education programme as an adjunctto guidelines for stroke carers may help to overcome these limitations.


IS REMOTE TELE RT-PA TREATMENT SAFE? THE EXPERIENCE OF THEEMERGENCY NEUROLOGY NETWORK IN FRANCHE-COMTE (RUN)E. Vidry, E. Medeiros, E. Revenco, F. Vuillier, P. Decavel, T. MoulinUniversity Hospital Besançon, Besançon, France

Background: The only validated treatment for acute ischemic stroke is thromboly-sis with intravenous alteplase (rt-PA) performed in a stroke unit (SU). Its efficiencyis strongly time-dependent. Our aim was to evaluate the efficiency and safety of

rt-PA administered in distant hospitals using telemedicine tools imlemented in theRUN, compared to that of patients treated directly in the Besançon SU.Method: All patients admitted to the SU who were treated with rt-PA for ischemicstroke since 2005 were included. Patients were either treated in the SU afteradmission or were transferred to the SU after receiving rt-PA in a distant hospital.The decision to administer rt-PA in distant hospitals was made by the SU neurologyteam using “tele rt-PA” with telemedicine tools (video/imaging transfers). Thethrombolysis decision respected contra-indication and followed the usual criteriafor patients admitted within 3 hours or was guided by MRI findings (after 3hours). NIHSS scores were measured on admission, modified Rankin scores (mRS)were determined after treatment and at discharge as well as stroke causes andhaemorrhagic transformations (HT).Results: Of 34 patients treated by rt-PA, 16 were in distant hospitals and 18 inthe SU. Median treatment times were 3hrs in the SU (4hrs25 for those transferredthere) and 2hrs30 in distant hospitals. Following rt-PA, 7/34 (20%) of patientsdeveloped HT, which was symptomatic in only 3 (8.8%), 1/3 of whom were treatedin distant hospitals. mRS distribution was not statistically different between the2 patient groups: mRS 0-1 in 6/34 (18%), of which 4/6 (66%) were in a distanthospital, mRS (2-3) in 9/34 (26%), of which 4/9 (44%) were in a distant hospital,mRS (4-5) in 16/34 (47%) of which 8/16 (50%) were in a distant hospital, and 3/34(8%) of patients died, all of whom were treated in the SU.Discussion: No significant difference was found concerning outcome or haemor-rhagic complications between patients treated in the SU and distant hospitals. Thestudy shows that “tele rt-PA” is safe and can improve patient outcome.


DIURNAL FLUCTUATIONS IN THE EFFICACY OF RT-PA IN ACUTEISCHEMIC STROKEJ. Gracia, S. Martinez-Ramirez, A. Ayo, D. Cocho, J. Martí-Fabregas, T. SeguraDepartment of Neurology, University Hospital of Albacete, Albacete, Spain

Background: several studies have demonstrated that a circadian pattern of efficacyexists in myocardial acute infarction, with a greater efficacy in the evening. Thisfact has been explained for the presence of different levels of plasminogen-activatorinhibitors and probably other haemostatic factors. There are not published datacommunicating this fact in intravenous thrombolytic therapy in stroke patientsObjetives: to investigate possible diurnal fluctuations in the efficacy of intravenousthrombolysis in ischemic stroke patientsMethods: one hundred forty-eight patients with acute stroke treated with intra-venous rt-PA were prospective enrolled in this study, conducted in two differenthospitals between January 05 and May 06. Efficacy of thrombolysis was deter-mined according to accepted clinical criteria: neurological response was assessedby NIHSS scale performed at 2 and 24 h after therapy, and long term functionaloutcome was evaluated by modified Rankin scale at three months. We comparedthe results dividing patients into 2 different groups, day (08-20 h.) and evening(20-08am) and again dividing into 4 intervals of 6 hours: 0-06am, 06-noon, noon-06pm and 06-00.Results: The study population consists of 85 males and 63 females, age range 24-82years (mean 68.47, SD 11.2). There were no demographic or clinical differencesbetween the two cohorts from the different hospitals (74 patients every one) oramong the pre-specified groups of time of thrombolysis. In the whole group,median NIHSS before treatment was 16 and median NIHSS at 2 h and 24 h were11 and 9 respectively. There was no significant relationship between the clinical orfunctional result of thrombolysis and the time of the treatment.Conclusions: Our results do not support the hypothesis that exist circadianvariations in the response to thrombolytic therapy in ischemic stroke.


NO PREVIOUS EXPERIENCE WITH INTRAVENOUS THROMBOLYSIS FORACUTE ISCHAEMIC STROKE DOES NOT INFLUENCE THE PROPORTIONOF PATIENTS TREATEDA. Kobayashi, M. Skowronska, T. Litwin, A. CzlonkowskaInstitute of Psychiatry and Neurology, Medical University of Warsaw, Warsaw,Poland

To determine the eligibility of ischaemic stroke patients admitted to the 2nd De-partment of Neurology for intravenous thrombolysis, identify the major exclusionsand assess if changes of the in-hospital pathway and informative campaign in thelocal community and medical services can increase the number treated. To establishif lack of previous experience with thrombolytic treatment or trials is predictive ofa low proportion of patients treated.A survey of the database of stroke patients admitted during the first 30 months



following introduction of intravenous thrombolysis for acute ischaemic stroke inorder to identify eligible patients. This included patients admitted within 2 hoursof symptom onset (assuming a 1-hour door-to-needle time), age < 80 years,National Institute of Health Stroke Scale (NIHSS) score from 5 to 22, seizuresat onset, platelet count >100,000 per ml, glycaemia from 50 to 400 mg per dland international normalized ratio (INR) <1.6. We have compared the number ofpatients eligible with the number treated.745 patients with acute ischaemic stroke were admitted. 18.4% were admittedwithin 2 hours of onset, 71.0% were aged under 80, 55.4% had an NIHSS scorebetween 5 and 22. 96.1% had INR < 1.6, 98.9% had a platelet count higher than100,000 per ml, 99.4 had blood glucose between 50 to 400 mg per dl and 97.4%had no seizures at onset. After adjusting for all inclusion criteria 7.1% of patientswere found potentially eligible and 8.7% were actually treated (p=0.250). Out ofthe 65 treated patients 63.1% were independent after 3 months, 16.9% had diedand none had a symptomatic intracranial haemorrhage.The proportion of ischaemic stroke patients treated with intravenous thrombolysisin a previously inexperienced centre is not lower than in other centres and countrieswere this treatment is provided for a longer period of time. The number of patientstreated is higher than estimated mainly due to organizational changes introduced inour centre.


NEUROPROTECTIVE EFFECT OF XG-102 ALONE OR IN COMBINATIONWITH TISSUE PLASMINOGEN ACTIVATORK. Wiegler, C. Bonny, D. Coquoz, L. HirtCentre Hospitalier Universitaire Vaudois, Lausanne, Switzerland

Background: XG-102 (formerly D-JNKI1), a TAT-coupled dextrogyre peptidewhich inhibits the c-jun N-terminal kinase (JNK), is a powerful neuroprotectivedrug in rodent models of cerebral ischemia (Borsello et al., Nat Med, 2003; Hirt etal., Stroke, 2004) when administered by intra-cerebro-ventricular injection (i.c.v.).We studied the feasibility of systemic administration. We also evaluated in vitrothe effect of combined administration of XG-102 and tissue plasminogen activator(TPA), known to exacerbate excitotoxicity.Methods: Young adult male ICR-CD1 mice were subjected to 30 min transientsuture MCAo. XG-102 was administered intravenously (i.v.) 6 or 9 hours afterischemia. Neurological outcome was evaluated by neurological scores and rotarodtests. Rat organotypic hippocampal slice cultures subjected to oxygen (5%) andglucose deprivation (OGD) for 30 minutes.Results: XG-102 administered i.v. 6 hours after ischemia onset significantlyreduced the infarct volume at 48 hours. The lowest dose with maximal neuroprotec-tion, was 0.3 μg/kg, which reduced the infarct volume from 62±19mm3 (n=18) forthe vehicle-treated group to 18±9mm3 (P<0.001, n=5). Administration of XG-102(1mg/kg) 9 hours after MCAo did not significantly reduce the infarct volume. Thebehavioural outcome after transient MCAo was also considerably improved. Inorganotypic slices, TPA alone (0.9 μg/ml) administered immediately after OGD,increased cell death. XG-102 (12 nM), 6 hours after OGD onset, induced a strongreduction (P<0.001) of cell death compared in the presence of TPA (49±19%,n=20 vs 12±14%, n=24).Conclusion: XG-102 is a powerful neuroprotectant in our mouse stroke model,and can be administered i.v. up to 6 hours after MCAo. In vitro XG-102 can induceneuroprotection also in presence of TPA.Funding: CTI #7057.2


EFFECTS OF ASPIRIN PLUS CILOSTAZOL TO REDUCE THE PLATELETACTIVATION IN ACUTE ISCHEMIC STROKEJ.-K. Cha, S.S. Kim, H.W. Jeon, Jk ChaCollege of Medicine, Dong-A University Hospital, Busan, South Korea

Background: Aspirin has been still considered to be the most evidenced therapeuticregimen to prevent the recurrence of ischemic events in acute ischemic stroke.Recently, Cilostazol, a Phosphodiesterase III inhibitor, has been known as usefulantiplatelet agents to curb the progression of atherosclerotic ischemic stroke. In thisstudy, we investigated the usefulness of cilostazol on the top of aspirin to regulatethe expression of P-selectin and PAC-1 on activated platelet in acute ischemic stroke.Methods: We analyzed seventy-seven patients with acute ischemic stroke (<72hrs). Among them, 50 patients were prescribed aspirin 100 mg per day and another27 aspirin 100 mg plus cilostazol 200 mg per day. All patients were seriallyvaluated the expression P-selectin and PAC-1 on activated platelets at admissionday and 5 days later. We also serially measured clinical progression by using NIHstroke scale at same time period.

Results: After 5 days of using aspirin plus cilostazol, the extent of PAC-1 (58.5± 19.2%, P<0.005) on activated platelet was significantly reduced compared withthe baseline (71.0 ± 12.1%). But there was no any difference (34.1 ± 13.3 vs 30.9± 10.5%) in aspirin only. In case of P-selectin, both aspirin and cilostazol groupshowed any differences of their expressions after 5 days of initiating treatment.Compared clinical progression between two groups, there were no any significantchanges of NIHSS in the observation period.Conclusions: In this study, we showed that the combined regimen of aspirin andcilostazol had beneficial effect to reduced PAC-1 activity on activated plateletsin acute ischemic stroke. However, this regimen did not showed better clinicaloutcome than aspirin only. Therefore, we need the more detailed future study aboutthe clinical benefit of cilostazol in acute ischemic stroke.


CARDIO-PROTECTION IN ACUTE STROKE: HYPOTHESIS FORINTERVENTION FOR MORTALITY REDUCTIONJ.C. Sharma, I.N. Ross, M. VassalloKings Mill Hospital, United Kingdom

Background: Measures of damage limitation for acute stroke have not producedsubstantial benefit to reduce stroke mortality. Search continues for measures toreduce stroke mortality.Methods: Literature review for influence of cardiovascular factors, specifically thevalue of NT proBNP (a sensitive index of cardiac impairment) for stroke mortality,Results: Cardiovascular factors, in particular cardiac failure, adversely influenceacute stroke mortality. Recent studies reveal that Troponin and NT-proBNP areelevated in acute stroke patients, in response to the activated Renin-Angiotensin-Aldosterone-System and other neurohumoral changes, as a protective mechanismfor sympatho-inhibitory activity. Elevated NT-proBNP has been reported to beassociated with higher short and long term mortality. In one study all patientswho died at 4 months had NT-proBNP levels above the median, no patient withNT-proBNP below the median value died. Two studies revealed that NT-proBNP ismore significant than clinical stroke severity for stroke mortality. Raised Troponinindicates myocardial injury, raised NT-proBNP indicates occult cardiac impairmentin acute stroke patients. Protection of myocardium in stroke patients may bepossible by the use of drugs such as beta-blockers and the drugs acting on RAAS.Reduction of mortality in studies of candesartan (ACCESS) and prior betablockersis one such example.Conclusion: Some stroke patients die due to occult cardiac impairment in acutephase due to common risk factors. This relationship between brain and heart needsevaluation. Protection of heart with currently available or new drugs in acute strokesis worth investigating since this intervention could be applied to a large proportionof acute stroke patients over a wide time window.


A NEW DEVICE FOR THE TREATMENT OF THROMBO-EMBOLICSTROKESG. Pearce, N.D. Perkinson, J.H. PatrickORLAU, RJAH Hospital Oswestry, United Kingdom

Each year in the UK 130,000 strokes occur (Arnold 2006), and it remains a heavyburden on hospital and social service provision. We have invented a clot extractiondevice for use in the acute situation-to reduce the continuing neuronal cell death inthe penumbral area, (the “GP” Clot Removal Device). This device was invented byPearce and Perkinson (2005) has recently been published as a international patent(W0206120464). The unique interior surface of this device facilitates controlledremoval of blood clots in arteries that block during thrombo-embolic strokes. Ourdevice has an inner surface that generates a helical vortex which actually removesthe clot. It has advantages are (i) it has no moving parts and (ii) it does not makeintimate contact with the blood clot or arterial wall. Complications should be lower.In-vitro experimental results (Pearce et al 2006a, 2006b, 2006c, 2006d, 2007) showdifferences between fluid flow patterns when fluid is sucked through the device(mounted within the catheter) compared to fluid sucking through a simple catheter.We have also established that the device removes clots quicker and with lessvolume of blood being removed, than with simple catheter being use. Methyleneblue dye has been used to effectively demonstrate the mechanism of the device.Pearce and Perkinson, 2007 have also undertaken radial flow rates within the “GP”clot removal device.We have used both artificial clots and actual blood clots (abattoir sourced) toshow clot removal in plastic tubes and porcine abattoir arteries. At vacuum suctionpressures of 66 KPa, when using the “GP” clot removal device (of internal diameter2.5 mm) occlusive clots of about 1cm long are removed.



RAPID ALTERNATE DAY DOSE TITRATION OF DIPYRIDAMOLE IS WELLTOLERATED IN PATIENTS WITH ISCHEMIC STROKEJ.L. Pascual, H.M. Chang, M.C. Wong, C.P. ChenNational Neuroscience Institute, Singapore General Hospital Campus,Singapore, Singapore

Background: The combination of aspirin (ASA) and dipyridamole (DIP) has beenshown to be efficacious in the secondary prevention of acute ischemic strokeof arterial origin. However, DIP-induced headache is common and may affectcompliance. We aimed to determine the feasibility and tolerability of rapid dosetitration of DIP in Singaporean acute ischemic stroke patients.Method: Observational study of non-aphasic ischemic stroke patients requiringanti-thrombotic therapy and without hypersensitivity or contraindication to DIP.Patients were allocated to either open label DIP starting at 25mg tds followedby75mg tds and finally 150mg tds in either every other day(EOD) or week (EOW)titration regimes at the choice of the attending physician. Follow-up was at 1 monthand occurrence of headache or other adverse effects, as well as any reason forstopping the drug were documented.Results: 188 acute ischemic stroke patients were started on ASA + DIP overa 6-month period. There were 113 patients on EOD and 75 on EOW titrationschemes. No significant differences in baseline characteristics between treatmentgroups were seen. The EOW regimen was associated with more adverse events(22% vs. 9%, p < 0.05), predominantly headaches (72%), leading to more drugdiscontinuations(12% vs 5%, p < 0.05). The majority of headaches in the EODgroup (70%) occurred at 25 mg tds, while in the EOW group, patients hadheadache at 75mg tds or higher (88%). Discontinuation rates were similar for bothgroups(13% vs 9%, p = 0.19).Discussions: Rapid titration of DIP is well tolerated. These findings need furtherconfirmation with randomized trials but may form the basis of a treatment regimeto be considered in acute stroke trials.


COMBINED IV ABCIXIMAB/IA TPA TREATMENT OF ACUTE BASILARARTERY OCCLUSIONU. Becker, G. Urban, R. von Kummer, G. GahnUniversity Hospital Dresden, Dresden, Germany

Background: Acute basilar artery occlusion (ABAO) is a disease with highmortality and morbidity. The optimal therapeutic approach remains still unclear.We intended to analyze the rate of recanalization, complications and outcome inpatients with combined IV Abciximab/IA tPA treatment.Methods: Following a prospective protocol, patients, with ABAO received an IVbolus of Abciximab (0,25 mg/kg bodyweight), followed by a 12 hours infusion ofAbciximab (0,125 mg/kg bodyweight). Immediately after the bolus, we performedDSA and administered tPA locally (up to 40 mg or until recanalization wasachieved). We investigated the rate of recanalization, complications and the clinicalcourse by standardized telephone interview after 6 months. Good outcome wasdefined as a mRS < 3.Results: Since 2003, we included 16 patients with ABAO, mean age was 60.9years, 5 female, 11 male. Mean GCS ad admission was 9.4 (4-15), mean timewindow 8.2 hours (3-25).Recanalization was achieved in 13 patients (87%), symptomatic hemorrhage oc-curred in 2 patiens (12%). Survival after 6 months was achieved in 7 out of 16patients (44%), 3 of them had a good outcome (19%).Discussion: Regarding survival and good outcome we observed similar resultscompared to other treatment regimes, however, the rate of recanalization was higher.We are encouraged to treat more patients with the combination of Abciximab andtPa to find out whether morbidity can be reduced.


DYNAMICS OF MARKERS OF FIBRINOLYSIS AND ENDOTHELIALFUNCTION IN PATIENTS WITH ISCHEMIC STROKE TREATED WITHINTRAVENOUS RT-PAN.A. Shamalov, A.S. Kireev, N.A. Pryanikova, G.R. Ramazanov,A.G. Kobylyansky, I.A. Grivennikov, V.I. SkvortsovaFederal Stroke Institute, Moscow, Russian Federation

Background: The aim of this study was to assess the dynamics of the markers offibrinolysis and endothelial function in patients with ischemic stroke (IS) treatedwith intravenous (IV) rt-PA.

Methods: We prospectively included 14 patients treated with IV rt-PA within3 hours after stroke onset (mean age 62.9±12.2 years, 10 (71%) male, meanNIHSS score at admission was 12±5.8). A blood sample were obtained beforeand immediately after thrombolysis, in 24 hours, in 3 and 7 days after IS onsetto measure markers of fibrinolysis [Plasminogen Activator Inhibitor-1 – PAI-1(IU/ml), normal value 1-7] and endothelial function [Matrix Metalloproteinase-9 –MMP-9 (ng/ml), normal value 169-705].Results: Before thrombolysis increased plasma level of PAI-1 (14.1±17.9) anddecreased level of MMP-9 (102.9±106.8) was observed. After thrombolysis werevealed the tendency towards decreasing PAI-1 (6.8±11.2) and increasing MMP-9(136±166.1). In 24 hours, 3 and 7 days after IS onset level of PAI-1 insignificantlyincreased (7.2±5.1; 10.3±18.1 and 10.6±11.6 respectively) and MMP-9 leveldecreased (67.9±98.7; 35.1±7.5 and 54.3±26.6 respectively).Conclusion: IV rt-PA therapy in stroke patients may reduce the activity antifibri-nolytic system and cause short-term increase MMP-9 level.


ANCROD IN A SIX HOUR TIME WINDOW? EVIDENCE FOR AND AGAINSTEFFICACYD.G. Sherman, D.E. LevyUniversity of Texas Health Science Center at San Antonio, San Antonio, TX,USA

The publication in Lancet (2006;368:1871-78) of the unsuccessful European ancrodtrial (ESTAT), where the treatment began up to 6 hours after stroke onset, concludedthat compared with the positive results of STAT (JAMA 2000;283:2395-2403) withits 3-hour window, ESTAT’s results suggested that ancrod was ineffective if startedlater 3 hours. Although not published, ancrod’s efficacy was marginally better thanplacebo’s at 3-6 hours (43.2% vs. 42.3%) but was substantially worse at 0-3 hours(33.3% vs. 43.3%), inconsistent with attributing ESTAT’s poor results to its latertime window.Pooled analysis of the North American data from STAT and an earlier 6-hour study(Stroke 2004:25:291-97) shows that efficacy vs. placebo is virtually identical in 0-3hour (43.0% ancrod vs. 36.1% placebo) and 3-6 hour patients (43.4% vs. 34.1%)with no significant interaction of time-to-treat on ancrod response (p=0.69).Therefore, explanations other than enrollment beyond 3 hours must underlie theunfavorable ESTAT results. These include the fact that higher blood pressureswere permitted in ESTAT (up to 220/120) than STAT (185/105). Symptomaticintracranial hemorrhages occurred in 10.1% of ESTAT patients with systolic BPsof 185-220 vs. 6.9% of those with lower pressures. The ESTAT publication statesthat deaths rarely occurred in patients with symptomatic ICH (4/44 ancrod vs. 2/9placebo), but the actual data for 90-day mortality are 26/44 ancrod and 4/9 placebo.Other explanations for the unsuccessful ESTAT results include age imbalance (69.3ancrod vs. 67.6 placebo, p<0.03) and a 23% higher mean patient ancrod dose inESTAT than STAT.The final answer on ancrod’s efficacy beyond 3 hours awaits conclusion of the twocurrently-enrolling trials with 6-hour windows.


PERFUSION-CT GUIDED INTRAVENOUS THROMBOLYSIS AT 3 TO 6HOURS: FEASIBILITY AND SAFETY STUDY IN 15 PATIENTSP. Michel, M. Wintermark, M.D. Reichhart, F. Vingerhoets, P. Maeder,R.M. Meuli, J. BogousslavskyCentre Hospitalier Universitaire Vaudois, Switzerland

Objective: To show the feasibility, safety and possible effectiveness of applyingperfusion-CT (PCT) for patient selection for intravenous thrombolysis with rtPA 3to 6 hours after onset of acute ischemic stroke.Methods: Between 9/2002 and 1/2006, patients aged 18-80 with a NIHSS of 6-22arriving too late in our stroke center for treatment within 3 hours were eligible fortreatment. They had to have a minimal penumbra size for a given infarct (core) sizein the MCA-territory on PCT, according to a linear progressive cut-off table. Thismodel was designed for maximal potential benefit: the smaller the core size is, thesmaller the lower limit of the penumbra needs to be for inclusion. The maximalupper size for inclusion is a core of 30% of the MCA territory. The primaryoutcomes were symptomatic intracranial haemorrhage (ICH) and mortality at 90days. Secondary outcomes were independence at 3 months (modified Rankin scale0-2) and recanalisation rates at 24 hours measured by CT-angiography. The patientswere compared to 75 consecutive patients thrombolyzed in our center within 3hours based on NINDS-criteria (independently of their PCT results).Results (median ± standard deviation): 15 patients fulfilled the clinical and PCT-criteria. Age and NIHSS were similar in both groups. Median time to thrombolysis



was 255 (± 45) min. in the 3-6 hour group and 152 (±34) min. in the comparisongroup. Symptomatic intracranial hemorrage was similar (6.7% vs. 6.7%) and90-day mortality (6.7% vs. 10.7%) were similar. Good outcome was seen in 60%and 52% respectively (non-significant), and 24 hour recanalisation rates were 60%and 50% (non-significant).Conclusion: The time window for intravenous rtPA can safely be extended from 3to 6 hours if perfusion-CT shows salvageable tissue and no extensive core volume.Clinical outcome and recanalisation rates are comparable to 0-3 hour thrombolysisbased on plain CT.


SYSTEMIC THROMBOLYSIS WITH ABCIXIMAB IN ACUTE ISCHEMICSTROKE DUE TO ANTERIOR CIRCULATION INFARCTION BEYOND THE3-HOUR TIME WINDOWK. Barlinn, U. Becker, G. GahnUniversity of Dresden, Dresden, Germany

Background: Intra-arterial treatment of ischemic stroke using pro-urokinase isthe only recommended therapy within 6 h of onset of symptoms in patientswith occlusion of the middle cerebral artery (MCA). Results from other trialssuggested that the glycoprotein IIb/IIIa antagonist Abciximab might also be usefulin treatment of ischemic stroke in patients without MCA occlusion beyond the 3 htime window. We aim to evaluate the safety and efficacy of Abciximab at standarddose in patients with acute ischemic stroke without MCA occlusion beyond the 3 htime window.Methods: From May 2002 to January 2005, we retrospectively studied 7 patientswith acute ischemic stroke attributed to the anterior circulation without MCAocclusion who were treated with Abciximab (0.25mg/kg bolus, followed by 0.125μg/kg/min x 12h) beyond 3 hours from onset of symptoms. All patients hadwhether a progression or a fluctuation of focal neurological symptoms. MCAocclusion was excluded by computed tomography angiography (CTA). Intracranialbleeding complications were evaluated by CCT at day 1. National Institutes ofHealth Stroke Scale (NIHSS) was assessed at baseline and at discharge, modifiedRankin Scale (mRS) was obtained at discharge.Results: Mean onset to treatment time was 9 (range 4,5-48) hrs. No symptomaticintracranial haemorrhage occurred. There were no cases of extracranial bleedingcomplications. One patient had myocardial ischemia diagnosed. All patients (100%)improved clinically from mean NIHSS 7 (range 2-15) to 2 (range 0-10). The rateof favorable outcome (mRS 0-2) at discharge was 57% (4 of 7 patients).Discussion: Regarding clinical outcome, the absence of symptomatic ICH andother bleeding complications, Abciximab seems promising in late thrombolysis ofacute ischemic stroke in selected patients without MCA occlusion and progres-sion or fluctuation of symptoms. Nevertheless further prospective clinical trials arenecessary to select patients who benefit from treatment beyond the 3 h time window.


SETIS - SERBIAN EXPERIENCE WITH THROMBOLYSIS IN ISCHEMICSTROKED.R. Jovanovic, Lj. Beslac-Bumbaširevic, G. Toncev, M. Živkovic, for SETISGroupInstitute of Neurology, Clinical Center of Serbia, Belgrade, Yugoslavia

Background: First intravenous thrombolysis in IS in Serbia was carried out inFebruary 2006. We present our preliminary one year experience with intravenousthrombolysis in treating IS patients.Methods: All patients with IS treated with intravenous thrombolysis in Serbia wereincluded in the study. The time of stroke onset, first neurological exam, CT examand beginning of therapy were recorded. The early CT signs of ischemia weregraded by the ASPECTS score. Neurological deficit was assessed with NIHSSscore and functional outcome with modified Rankin Scale (mRS).Results: During one-year period 24 patients with IS were treated with intravenousthrombolysis in three tertiary care centers. Average age of patients was 50.5 years,ranging 18 to 78, with 62% of them younger than 55. Median time from symptomonset to hospital door was 52.5 minutes, median time door-to-CT was 37.5 minutes,and time from symptom onset to treatment was 165 minutes. Early CT signs ofischemia were present in 62% of patients with median ASPECTS score 9. Medianinitial NIHSS score was 14 with its decline during first 24 hours for at least 4points in 50% of patients. Symptomatic intracerebral hemorrhage was present intwo patients. After 30 days of follow-up, 33% of patients had favorable outcome(mRS < 1), 29% of patients had poor outcome (mRS 4-5) and two patients died,one with malignant MCA infarction with symptomatic parenchymal hemathoma,and the other patient with signs of heart insufficiency.

Conclusion: Based on our initial experience, we confirm that intravenous throm-bolysis carry a substantial benefit for the patients with acute IS. However, we alsouphold that one should strictly follow the thrombolysis protocol in order to avoidpossible complications.


INTRA-ARTERIAL THROMBOLYSIS IN ACUTE STROKE DUE TO MCAOCCLUSIONI. Henriques, A. Calado, R. Roque, A. Bandeira, I. Fragta, C. Ribeiro,J. Candido, J. ReisCentro Hospitalar de Lisboa - Zona Central, Lisboa, Portugal

Background: Standard treatment for acute ischemic stroke in eligible patients isintravenous (ev) thrombolysis, but efficacy is limited by low rates of recanalization.Intra-arterial (IA) thrombolysis, combined transcranial ultrasound, or clot retrievaldevices are being evaluated to improve thrombolysis efficacy. We prospectivelystudied consecutive patients with acute MCA occlusion submitted to IA rt-PAthrombolysis in the first 6 hours after symptoms onset.Methods: We included patients not eligible for ev thrombolysis according toECASS criteria. We excluded patients with vertebrobasilar stroke and with previousconcomitant ev thrombolysis. Symptomatic cerebral hemorrhage was considered ifassociated with clinical deterioration. Major neurological improvement was definedas a reduction of more than 7 points in NIHSS after procedure. Excellent or goodoutcome was considered if modified Rankin scale was 2 or less. We studied 15consecutive patients referred to our stroke unit (5 male) with median age of 72 (43–78) years.Results: Median NIHSS at admission was 19 (15-25), and 12 at discharge (2-25).MCA recanalization was total in 33% (TIMI grade 3) and partial (TIMI grade2)in 67%. In five patients, recanalization was achieved together with a mechanicalthrombectomy device. Hemorrhagic transformation was present in 7 patients (46%),and symptomatic in 5 (33%). Three patients died (20%) in the first 72h and noneafter. At discharge, excellent or good outcome was observed in 27% of the patients.Discussion: In this group of large MCA infarcts not eligible for ev thrombolysis,total recanalization was achieved in 33% and partial in 67%. Mechanical thrombec-tomy device was used together with IA rt-PA thrombolysis in 33% of our patients.Independency was achieved in 27% at discharge. Since patients not eligible forev thrombolysis with large MCA infarcts have a very high rate of disability andmortality, IA thrombolysis alone or together with mechanical device can be analternative to current treatment. Randomized trials may confirm safety and efficacyin larger series.


ARE POLYAMINE LEVELS USEFUL IN PREDICTING OUTCOMEFOLLOWING ACUTE STROKE?M.J. Macleod, P. Mahendra, M. Bruce, H.M. WallaceUniversity of Aberdeen, Aberdeen, United Kingdom

Background: It is difficult to predict stroke outcome on infarct size and strokeseverity alone. There is therefore interest in markers which might predict outcomefollowing acute stroke. The endogenous polyamines, putrescine, spermidine, andspermine, are found in high concentrations in the brain and have a regulatoryrole in apoptosis. When neuronal cells are damaged polyamines are ’mopped up’by erythrocytes. One small clinical study in acute stroke has suggested red cellpolyamine levels may correlate with infarct size and severity. The aim of this studywas to replicate and extend these findings in a group of patients admitted withan acute cortical infarct of less than 48 hours duration. Patients with significantcomorbidity or unable to give consent were excluded.Methods: Bloods and clinical assessment including NIHSS were performed atadmission, 12, 24 and 48 hours (depending on time of admission after onset ofsymptoms), 72 hours, 7 days, 14 days and 28 days. Samples were also availablefor 8 control subjects. Erythrocytes were separated and washed with isotonicNaCl, haemolysed with distilled water and extracted with HClO4. The extract wasneutralised with KOH and frozen at -40 oC. Analysis of polyamines was performedusing reverse phase HPLC and quantification by fluorescence detection.Results: Data from 10 patients is presented (6M/4F). Average age was 71.6 years(±8.9). Polyamine levels vary considerably between patients, but in all but onepatient (who had a normal CT scan) were higher than normal control subjects at 7days (p<0.008). Patients had a significant rise in polyamine levels between baseline(mean 8.7±5.4) and 7 days (mean 22.3±8.3), p<0.032. There was a correlationbetween total polyamine levels and NIH SS at day 7 post stroke (r=0.654, p=0.04).Discussion: These findings confirm that polyamine levels are elevated after anischaemic cortical stroke, and peak at 7-14 days. If the correlation with NIHSS is


confirmed in larger studies, polyamines may be a useful prognostic indicator afteran ischaemic stroke.

Long-term outcome of stroke

1 Long-term outcome of stroke

TIME SPENT AT HOME POST STROKE: “HOME-TIME” – A MEANINGFULAND ROBUST OUTCOME MEASURE FOR STROKE TRIALSJ. Dawson, T.J. Quinn, J.S. Lees, T.-P. Chang, M.R. Walters, K.R. LeesGardiner Institute of Cardiovascular and Medical Sciences, University ofGlasgow, Glasgow, United Kingdom

Background: Assessment of stroke outcome requires a measure of functional re-covery eg. modified Rankin scale. Such instruments are prone to bias and variationin clinical application. Approaches to improve rigour often increase complexitywithout similar increases in clinical utility. Length of stay in hospital and nursinghomes is strongly related to incremental increases in mRs score but is weakenedas an outcome measure because fatal outcomes tend to shorten stay. We examinedduration of stay in the patient’s own home or chosen environment – “Home-time”- as an alternative outcome more likely to show a graded response with lessconfounding by survival issues.Methods: We examined prospectively collected resource use data from the GAINInternational trial. We assumed Home-time if patients returned to their own orrelative’s home after stroke, restricting analysis to the first 90 days and usingANOVA with Bonferroni contrasts of adjacent mRS categories.Results: We had full outcome data from 1717 of 1788 intent to treat patients. Meanage ±SD was 70 ±12 years; 737 were female. Mean initial NIHSS was 13±6 and321 had primary intracerebral haemorrhage. Increasing Home time was associatedwith significantly improved mRs scores (p<0.0001; table).

mRs 0 mRs 1 mRs 2 mRs 3 mRs 4 mRs 5 mRs 6 (death)

N 197 268 205 214 366 143 324Home-time duration

(mean) 72.6 64.1 45.9 31.5 11.2 9.1 1.0– * * * * p=0.37 p=0.0003

95% CI 69.9–75.5 61.3–67.0 41.8–49.9 27.5–35.6 8.9–13.4 5.5–12.8 0.3–1.6

Home time = days spent living electively in existing setting, within the first 90d. *p<0.0001 comparedto preceding column

Conclusion: Recording of Home time offers a robust, useful and easily validatedoutcome measure for stroke, particularly across better recovery levels.


THE IMPACT OF OBESITY ON SHORT AND LONG-TERM OUTCOMEAFTER FIRST EVER ACUTE STROKE. THE STROKE OBESITY PARADOX?S. Scalidi, K. Xynos, T. Pappa, J. Zafeiriou, N. Mentis, N. Kokolakis,K. VemmosAcute Stroke Unit, Dept. of Therapeutics, Univ. of Athens Med. School,Alexandra Hosp., Halandri, Greece

Background: Obesity has long been implicated as a higher morbidity and mortalityrisk factor for cardiovascular disease. However, its potential role and pathophysio-logical significance on the outcome of patients after an acute stroke has not beenyet established. Our aim was to assess short and long term survival as well asrecurrence rate of obese patients suffering an acute stroke, compared to that of thenon-obese population.Methods: We prospectively studied 1998 patients with first-ever acute stroke. Apartfrom the stroke risk factors and body mass index (BMI), we scored consciousnesslevel and neurological deficit on admission by means of the Glasgow Coma Scale(GCS) and the Scandinavian Stroke Scale (SSS). Cox regression and Kaplan-Meiermethod was used in order to estimate the impact of obesity on survival andrecurrence.Results: In our study population, 473 patients (23.7%) were classified as obese(BMI>30) and 1525 (76.3%) as non-obese (BMI<30). Obese patients had higherproportion of hypertension, diabetes and cholesterol. No significant difference inage, GCS and SSS was observed between groups. After adjustment for age, sex,risk factors and stroke severity, obesity was an independent prognostic factor forsurvival Hazard Ratio=0.637 (95%CI 0.527-0.770), p=0.0001. Cumulative survivalfor obese and non-obese patients was: after 1 month 0.895 (95%CI 0,864-0.921)

and 0.831 (95%CI 0.812-0.850), and in the end of follow up at 10 years 0.464(95%CI 0.410 to 0.518) and 0.290 (95%CI 0.259 to 0.321) respectively, (log ranktest=24.23, p=0.0001). No differences were seen in recurrence rates (log ranktest=0.17, p=0.68).Conclusion: Obesity in patients with acute stroke is associated with better short andlong-term survival. The mechanisms involved in obesity-related neuroprotection inacute stroke demand further investigation.


BIOCHEMISTRY VERSUS CLINICAL SEVERITY OF ACUTE STROKE:SIGNIFICANCE OF NT PROBNP TO PREDICT ONE YEAR MORTALITYJ.C. Sharma, I.N. RossKings Mill Hospital, United Kingdom

Background: We have investigated the value of biochemical measurements topredict stroke mortality.Methods: Logistic regression was used to investigate significant variables of thebiochemical and clinical parameters in 125 previously independent (mRS <4) acutestroke patients for mortality at one year.Results: Age range 40 to 95, mean 73±12, Female:Male 63:62. 20 patients haddied at one year. Results are given for means ± SD between dead and alive patientsusing Student’s t test. There was no significant difference in mean creatinine111±32 vs 101±41, p=0.32, eGFR 54 vs 62, p=0.08; oxygen saturation–97.0 vs97.0, p=0.60; HbA1C 6.01±0.7 vs 6.3±1.3, p=0.38; glucose 6.6±1.4 vs 7.0±2.8,p=0.53; serum osmolality 306±11 vs 303±18, p=0.52; cholesterol 4.5±1.1 vs4.9±1.2, p=0.16 and HDL 1.4±0.7 vs 1.5±0.7, p=0.42. There was a significantdifference in urea 10.1±4.6 vs 7.5±4.4, p=0.01; ALT 133.3±459 vs 30±33,p=0.02; alk phosphatase 140±116 vs 91.6±45, p=0.002; NT proBNP 980±1249vs 125±244, p<0.001; Barthel Index 2.2±2.5 vs 7.1±4.9, p<0.001; Scandinavianstroke scale (SSS) 22±14 vs 35±12, p<0.001 and NT proBNP log 6±1.7 vs3.5±1.7, p<0.001.Logistic regression analysis using the significant variables from univariate analysisrevealed that NT proBNP was the most significant variable to predict mortality– Wald 17.9, p<0.001 followed by SSS – wald 5.7, p=0.02. Other biochemicalvariables and Barthel Index were not significant to predict mortality.Conclusion: NT proBNP, a measure of cardiac impairment, is the only significantbiochemical variable to predict one year mortality, more significant than themeasures of clinical stroke severity. This provides an opportunity for interventionto reduce stroke mortality.


INFLUENCE OF STROKE SECONDARY PREVENTION DRUGS ONMOLECULAR MARKERS OF INFLAMMATION. THE MITICO STUDYJ. Vivancos, J. Alvarez-Sabín, A. López-Farré, E. Martínez-Vila, J. Montaner,T. Sobrino, J. Castillo, on behalf of The MITICO Study investigatorsHospital Universitario de La Princesa, Stroke Unit, Neurology Service, Madrid,Spain

Background: The MITICO study primary objective is to determine the prognosticvalue of inflammation molecular markers (IMM) in vascular recurrence risk.As a secondary objective, we studied the influence of antiplatelet, statins andantihypertensive drugs on IMM profile.Methods: Multicenter prospective observational study, including non-anticoagulatedischemic stroke patients (within 1 to 3 months of stroke onset), with no inflam-matory processes. Four visits were performed during the one-year of follow-up toidentify vascular death (VD) or vascular event (VE). Blood samples were obtainedat baseline visit for further determination of fibrinogen, high-sensitive C-reactiveprotein, IL-6, IL-10, ICAM-1, VCAM-1, MMP-9 and cellular fibronectin.Results: From 965 included patients (recruited in 59 hospitals), 780 subjects(67.5±11.2 years, 33.6% female) were valid for the main analysis. One-hundredand three patients (13.2%) showed a new VE and 116 patients (14.9%) either a VEor VD (66.4% stroke, 21.5% coronary and 12.1% peripheral).Only 21 patients were not taken antiplatelet drugs. Statins treatment was associatedwith reduction of VE and VD (47.4% vs 28.2%. p=0.001). None of them modifiedthe one-year functional outcome. There were significant changes between baselineand final plasma levels of IL6, MMP-9 and cellular fibronectin in statins-treatedpatients in comparison with non-treated patients (IL6: 0.9 [-1.5, 6.5] vs 0.3[-1.9,1.4] p<0.0001; MMP-9: 23.5 [-27.6, 119.5] vs -2.3 [-48.1, 34.4] p<0.0001;cFn: 4.4 [-4.8, 14.7] vs -6.1 [-10.8, 3.6] p<0.0001).Discussion: Statins treatment is associated with a significant reduction of VE andVD. Statins treatment influences IMM by lowering IL6, MMP-9 and cFn plasmalevels along follow-up period.

Poster SessionLong-term outcome of stroke



MORTALITY AND VASCULAR MORBIDITY AFTER ISCHEMIC STROKE ATA YOUNG AGE. A CROSS-SECTIONAL STUDY IN WESTERN NORWAYU. Waje-Andreassen, H. Naess, L. Thomassen, G.E. Eide, C.A. VedelerHaukeland University Hospital, Bergen, Norway

Background: The aims of this population-based long-term follow-up study were toobtain data on cardiovascular mortality and recurrent stroke, coronary disease (CD)and peripheral artery disease (PAD) for long-term survivors of ischemic stroke andcontrols.Methods: After a median observation time of 11.1 years we evaluated all 232patients aged 15-49 years with first-ever cerebral infarction in 1988-1997 and 453birthday- and sex-matched controls for causes of death. We used information fromthe Norwegian national death register, from hospital records and autopsy reportswhen available. Among long-term survivors we evaluated 144 patients and 167controls for cardiovascular events by questionnaires, hospital records and clinicalexamination including an electrocardiogram (ECG) for patients.Results: 45/232 (19%) patients and 9/453 (2%) controls died during follow-up(p < 0.0005). Causes of death among patients were: stroke (9), acute myocardialinfarction (9), other heart disease (4), sudden death (5), unknown (5), cancer (7) andothers (7). One patient had 2 causes of death by autopsy. Controls died from acutemyocardial infarction (1), cancer (6) and others (2). Among long-term survivorsrecurrent stroke was registered for 38 (26.4%) patients versus stroke in 5 (3%)controls (p < 0.0005), CD occurred in 19 (13.2%) patients versus 9 (5.4%) controls(p = 0.018) and PAD occurred in 17 (11.8%) patients versus 2 (1.2%) controls (p< 0.0005).Discussion: Mortality is increased after ischemic stroke at a young age andcardiovascular death is dominating among patients. Cardiovascular morbidity ishigher on cerebral, coronary and peripheral level in long-term surviving patientscompared with controls.Key-words: stroke, mortality, cardiovascular morbidity.


EFFECTIVENESS OF THROMBOLYTIC THERAPY ON OUTCOME WITHIN3 MONTHS AFTER ISCHEMIC STROKE: THE TELEMEDICAL PILOTPROJECT FOR INTEGRATIVE STROKE CARE (TEMPIS)A.M. Toschke, P.U. Heuschmann, J. Schenkel, H. AudebertKing’s College London, London, United Kingdom

Background: Randomized trials showed a benefit of intravenous application oftissue-type plasminogen activator (tPA) for ischemic stroke patients after threemonths. Observational studies reported inconsistent results regarding effective-ness of this treatment in terms of early mortality. Data on mortality after tPAadministration after 3m outcome from unselected community hospitals are scarce.Methods: Data were collected from the Telemedical Pilot Project for IntegrativeStroke Care (TEMPiS) in Southern Germany including comprehensive strokecentres and community hospitals. Patients were followed three months afterstroke onset. Propensity score analysis was used for adjusting differences insociodemographics, clinical characteristic, stroke severity, and comorbidities bytPA treatment.Results: Between July 2003 and March 2005 1710 patients after ischemic strokewere observed with a mean age of 74y (SE 0.3y); 48% were male. 76 (4.4%)patients were treated with tPA. Patients receiving tPA were younger, more oftenmale, had less often recurrent strokes; prevalence of diabetes, atrial fibrillation,dyslipidemia and hypertension did not differ compared to non-tPA patients. Theproportion of tPA patients who died in-hospital or after 3 month was similar(7.9% or 14.5%) compared to patients not receiving tPA (6.4% or 15.4%; p=0.61and p=0.83). After adjusting for baseline differences by propensity score, patientsreceiving tPA tended to have a lower probability of death in hospital (odds ratio(OR) 0.58, 95%CI 0.21-1.63) and at 3 months (OR 0.69; 95%CI 0.32 to 1.50).Conclusions: The tPA treatment within the TEMPiS community hospitals had noadverse effect on in-hospital and 3months mortality after stroke.


LONG-TERM PROGNOSIS OF STROKE IN YOUNG ADULTS: RESULTSFROM THE FIRST 227 CONSECUTIVE CASES ENROLLED DURING5-YEARS IN THE “ATHENIAN REGISTRY OF STROKE IN YOUNG ADULTS-ARSYA”K. Spengos, S. Vassilopoulou, M. Papadopoulou, A. Papapostolou,G. Papadimas, E. Manios, G. TsivgoulisEginition Hospital, University of Athens, Athens, Greece

Background: California- and ABCD-scores reliably predict short-term risk ofstroke after TIA. Both scores contain similar components. However, diabetesmellitus (DM) is only included in the California Score. Aim of the present studywas to evaluate the potential relationship of DM with the early risk of stroke in acohort of hospitalised TIA patients.Methods: All patients hospitalised in our Department with definite TIA duringa 5-year period were identified and their medical charts as well as their Emer-gency Room records were retrospectively reviewed by two investigators blindedto follow-up. Patients with previous history of stroke and those who missed theirfollow-up evaluations at the outpatient clinic of our Department at 1 month afteradmission were excluded. DM was specified as fasting serum glucose 7.0 mmol/L,nonfasting serum glucose 11.1 mmol/L, or use of oral blood sugar-lowering drugsor insulin. The outcome events of interest in all TIA patients were subsequentstrokes during the 1-month follow-up period. Statistical analyses were performedusing the Kaplan-Meier product-limit method and stepwise Cox’s proportionalhazards model.Results: The 30-day risk of stroke in the present case series (n=226) was 9.7%(95%CI:5.8-13.6%; 22 events). The 30-day risk of stroke was higher in patients withDM (17.3%; 95%CI:7.6-27.0%) than in non-diabetic patients [(7.1%; 95%CI:3.2-11.0%); log-rank test=5.20; df=1; p=0.0225]. After adjustment for demographiccharacteristics, stroke risk factors, history and number of prior TIAs, duration andsymptoms of TIAs, as well as secondary prevention treatment strategies duringhospitalisation, DM was independently (p=0.015) associated with a three-foldgreater 30-day risk of stroke (HR:2.98; 95%CI:1.28-6.94).Discussion: DM is an independent predictor of subsequent stroke in patientspresenting with TIA. It should be taken into account by prognostic scores thatstratify the risk of early stroke in TIA patients.


LONG-TERM SURVIVAL AFTER FIRST-EVER STROKE IN THE BESANÇONSTROKE REGISTRY: IMPACT OF STROKE UNIT ORGANISATIONP. Decavel, E. Medeiros, E. Vidry, F. Vuillier, E. Revenco, M. Pellicier,T. MoulinUniversity Hospital Besançon, Besançon, France

Background: Development of stroke management over a number of years haschanged the vital and functional prognosis of patients. The aim was to test theinfluence of stroke management on the long-term survival of patients admitted withfirst-ever stroke in the Besançon stroke registry.Method: To test long-term outcome of patients with first-ever stroke over 3different periods (period 1: 1987-1994; period 2: 1998-2002; period 3: 2003-2006)corresponding respectively to “basic stroke unit” period, stroke unit organisationperiod and network organisation period in 3 cohorts of unselected stroke patients.Results were statistically adjusted according to age, gender and stroke character-istics. During the different periods, all patients with a first-ever stroke (infarction,haematoma and TIA) admitted to the Besançon university hospital were registeredprospectively and assessed according to standardised diagnostic criteria. Patientswere followed up over several years and the outcome was analysed during the firstyear.Results: There were 6103 patients (55% male) which included 4250 (70%) infarc-tions, 678 (11%) haematoma and 1175 (19%) TIA. The median age of the cohortswas 71 years. The cohorts were different for each period in terms of recruitment(gender - p = 0.035 - and age - p < 0.000) and stroke subtypes (p < 0.000). In thefirst year after stroke, the overall survival rate was 75% (period 1), 80% (period 2)and 82% (period 3). Although there was no difference in survival rates for patientswith TIA in any period, there was continuous improvement in survival rates forpatients with haematoma (p = 0,041) or infarction (p = 0,022). Adjustments to ageand gender amplified these results.Conclusion: This study shows the strong impact of stroke management organisationon long-term patient outcome in the Besançon area.



DOES ADMISSION C-REACTIVE PROTEIN PREDICT OUTCOME INSTROKE PATIENTS UNDERGOING THROMBOLYSIS?R. Topakian, A.M. Strasak, H.-P. Haring, K. Nussbaumer, F.T. AichnerAcademic Teaching Hospital Wagner-Jauregg, Linz, Austria

Objective: After acute stroke, increased levels of C-reactive protein (CRP) mea-sured at discharge are associated with unfavourable outcome. We tested thehypothesis that admission CRP may predict outcome in stroke patients undergoingintravenous thrombolysis treatment (IVT).Methods: From January 2003 to June 2006, 129 patients underwent IVT for acuteischemic stroke in our centre. 111 patients were valid for analyses after exclusionof those with stroke involving a territory other than the middle cerebral artery’s andthose probably infected (admission CRP >6 mg/dl). Patient data were collectedin a prospective local registry. CRP was measured by turbidimetry (Cobas Integra700, Roche).Results: 52 (46.8%) patients were independent after 3 months, defined by a modi-fied Rankin Scale (mRS) score <3. Admission CRP levels were non-significantlyhigher in 1) patients independent after 3 months compared to patients with mRS>2 [median (range): 0.4 (0-5.7) mg/dl vs. 0.3 (0-5.9) mg/dl, p=0.131], 2) patientswho survived (87.4%) vs. patients who died within 3 months [median (range):0.4 (0-5.9) mg/dl vs. 0.2 (0.1-1.5) mg/dl, p=0.275], and 3) patients who didnot deteriorate neurologically within 24 hours (91.9%) vs. those who developeddeterioration defined by an increase of the National Institute of Health Stroke Scale(NIHSS) score of at least 4 points compared to baseline [median (range): 0.4 (0-5.9)mg/dl vs. 0.2 (0.1-0.7) mg/dl, p=0.091]. In multivariate logistic regression analyses,baseline NIHSS was the only variable significantly associated with independencyafter 3 months (OR 1.235, 95%CI 1.118-1.363, p<0.001).Conclusion: Our findings suggest that admission C-reactive protein is not useful inpredicting outcome in stroke patients with thrombolysis treatment.


INITIAL EXPERIENCE WITH VIDEO BASED MODIFIED RANKINASSESSMENTT.J. Quinn, J. Dawson, M.R. Walters, K.R. LeesGardiner Institute of Cardiovascular and Medical Sciences, Glasgow, UnitedKingdom

Introduction: Modified Rankin Scale (mRs) is the preferred outcome measure instroke trials. Despite availability of training and structured interview interobservervariability remains apparent - kappa=0.75 among UK SAINT trial investigators.We hypothesised that “off-line” assessment of video recorded interviews wouldoffer the means to improve reliability.Methods: 102 consenting patients were graded independently by two assessors.Patients were randomised to undergo of structured interview or standard assess-ment. One assessment from each pair was further randomised to video recording.Videos were assessed by four experienced researchers, blind to interviewer’s andother panellists’ gradings.Results: 100 videos were of technical quality to allow assessment. A range of ages(median:70 range:30-96) and stroke subtypes (Cortical 44; Lacunar 41; Posterior9) representative of a trial population were included. Initial mRs scores agreedin 66.7% of cases. Use of the structured interview did not significantly improvereliability. At video assessment there was consensus amongst scorers for 55%of cases (kappa 0.70). Greatest variability was seen for mRs grades 2 and 4(kappa=0.60, 0.65).

Paired mRs Assessment Total Structured (n=49) Standard (n=53)

Agreement 68 30 38Disagree = 1 mRs grade 29 14 15Disagree > 1 mRs grade 5 5 0

Video Assessment (4 reviewers) Total Structured (n=48) Standard (n=51)

Consensus 54 24 303/4 Agree 32 17 152/4 Agree 13 7 6

Discussion: We have shown that “off-line” assessment of mRs is possible in a mockclinical trial setting. Video assessment did not alter interobserver variability butoffers potential for central endpoint committee review with resultant improvementsin precision: pilot work to assess effect on reproducibility is now underway.


CHANGES IN QUALITY OF LIFE FROM ONE TO SIX MONTHSFOLLOWING ACUTE STROKE AND ITS DETERMINANTSO.M. Ronning, K. StavemStroke Unit, Akershus University Hospital., Lørenskog, Norway

Background: There is little information available about change in health-relatedquality of life (HRQoL) during the first few months following acute stroke, andwhether baseline variables can predict who will have the largest improvementin HRQoL. This study assessed the change in HRQoL from one to six monthsfollowing acute stroke and the determinants of these changes.Methods: Patients > 60 years of age, who experienced an acute stroke and wereadmitted to hospital within 24 hours of onset, were followed prospectively. HRQoLwas measured with the SF-36 health status questionnaire.Results: Of the 550 eligible patients, 315 fulfilled the inclusion criteria andwere alive after 30 days. At one month 174 responded to the questionnaire ofwhom 140 also completed the second questionnaire. The changes in HRQoL werestatistically significant on all the SF-36 scales (Physical functioning: p<0.001,Role physical: p<0.001, Vitality: p<0.001, Social functioning: p<0.001, Roleemotional: p<0.001, Bodily pain: p=0.016, General health: p=0.002, mental healthp=0.02). Mean scores for the two summary scales, physical component summary(PCS) and mental component summary (MCS), increased from one to six monthsfrom 36 to 42 (p < 0.001) and 43 to 53 (p < 0.001). Higher baseline scores atone month were associated with lower changes in PCS and MCS in multivariateanalysis. Less severe stroke related to large improvement in PCS. Treatment instroke unit was associated with a larger improvement in MCS.Discussion: In the present study we show a marked improvement in HRQoL fromone to six months. There was a favourable change in all domains assessed.


LONG TERM SURVIVAL OF STROKE PATIENTS FOLLOWING ANINPATIENT REHABILITATION ADMISSSIONC.E. Connolly, J. Estell, F. Kohler, R. RentonBraeside Hospital, Prairiewood, NSW, Australia

Objective: Patients admitted for stroke rehabilitation generally have numerouscomorbidities and a significantly increased mortality rate compared to the restof the population. The pioneers of Rehabilitation medicine demonstrated thatrehabilitation improved quality of life and minimised dependency. Limited researchhas however been done to determine the survival period of patients after inpatientrehabilitation for stroke. The aim of this study is to determine the survival periodof stroke patients following an episode of rehabilitation in our unit.Method: All patients admitted to the Braeside Rehabilitation Unit (NSW Australia)for stroke rehabilitation in the two years from 1st January 1997 to 31st December1998 were identified. The hospital databases were checked to determine; the lastdate of patient contact with a health service, or any indication that the patient haddied, and if so, the date of death. If there was no recent contact or confirmeddate of death a search was performed on the National Death Register kept by theAustralian Institute of Health and Welfare to establish if death had occurredResults: 253 patients were admitted for stroke rehabilitation during the referenceperiod. Of these 7 patients died within 28 days of discharge, 20 patients died within1 year and 30 within 2 years of discharge. By the end of eight years 72 patients haddied.Discussion: The study shows that 71.5% of patients who were discharged followinginpatient stroke rehabilitation remained alive 8 years later. Given the age of thispopulation and presence of multiple comorbidities this survival rate is high andunderlines the importance of maximising patient function and outcomes for thispatient group. Further investigation with regards the survival periods in differentstroke subtypes is warranted.


ASSESSMENT OF QUALITY OF LIFE IN STROKE PATIENTS CAREGIVERS.HOW TO PREDICT CAREGIVERS AT RISKE. Marco, E. Duarte, M. Tejero, J.M. Muniesa, R. Belmonte, A. Aguirrezabal,M. Pou, C.B. Samitier, F. EscaladaPhysical Medicine and Rehabilitation Department, Hospital de l’Esperança,Barcelona, Spain

Background: The increased proportion of stroke survivors has led to more impairedand disabled subjects. The patient disability condition and the provision of carehave negative consequences for caregivers. Objectives: To detect the most affected

Poster SessionLong-term outcome of stroke


dimensions in quality of life of stroke patients caregivers, and predict caregivers atrisk of presenting a worse health-related quality of life.Methods: Cross-sectional study of 215 family caregivers of stroke patients. Mainvariables were assessed with: the Short Form 36 (SF-36), the Geriatric DepressionScale and the FIM-instrument. Statistical tests: Chi-squared and Fisher test, Student“t” test, ANOVA, Mann-Whitney “U” test and Spearman correlation.Results: The SF-36 subscales of vitality, social function, emotional role and mentalhealth are the most affected. There is a positive correlation between caregivershealth perception and patients health perception in these subscales. Motor disabilityand assistence time required are the patient characteristic with a major influencein caregiver health perception. The presence of depression and osteoarthritis incaregivers have a significant effect on their perception of health. A model to beused to detect the caregiver at risk of presenting a bad perception of health statusis constructed. This predictive model contemplates 8 different probability groupsranging from 8.1 to 88.4%.Discussion: Health-related quality of life of caregivers is affected in differentdimensions. To detect which caregivers are at most risk of presenting a badperception might be useful to provide resources and services to help caregivers incharge of stroke patients.


POSTSTROKE EPILEPSY IN PATIENTS TREATED WITH SYSTEMICTHROMBOLYSISB. Dimitrijeski, A. Villringer, A. HartmannCharité-Campus Benjamin Franklin, Berlin, Germany

Objectives: Poststroke Epilepsy (PSE) is an important complication after ischemicstroke. We aimed to asses the occurrence of poststroke epilepsy, and to identifypossible predictors in patients with ischemic stroke treated systemically withthrombolysis.Methods: Patients with supratentorial stroke treated between 1998 and 2003systemically with rt-PA according to the NINDS trial-protocol were followedprospectively. Neurological status was measured at admission, discharge, 3-months-follow-up and long-term follow-up. Occurrence of poststroke epilepsy wasnoted. Predictors for PSE were identified using univariate analysis.Results: A total of 134 Patients were treated. Long-term follow-up (mean follow-up–time for the survivors was 39 months, range 14-66 months) was completed for132 patients, 2 patients were lost to follow-up. Good clinical outcome (Rankin<2)and mortality at 3 months was 50% and 13%, and at long-term follow-up 45% and28% respectively.Poststroke epilepsy occurred in 14 patients (11%). Factors univariately associatedwith poststroke epilepsy were NIH at discharge (p= 0.001), cardio-embolic stroke(p=0.018) and increasing infarct size (p=0.003).Conclusions: The prevalence of poststroke epilepsy in patients treated with sys-temic thrombolysis after a mean follow-up of 39 months was 11%. NIH-Score atdischarge, cardioembolic stroke and increasing infarct size are associated with ahigher risk for developing PSE.


INCIDENCE OF ASPIRATION PNEUMONIA AFTER SYSTEMATICAPPLICATION OF AN EXPERT BEDSIDE SWALLOWING EVALUATION ANDFEEDING PROCEDURE IN A STROKE UNITV. Palumbo, S. Centorrino, P. Nencini, G. Pracucci, D. InzitariUniversity of Florence, Florence, Italy

Background: Aspiration pneumonia is one of the most common and severecomplications of stroke-related dysphagia. We aimed to evaluate the variation overtime of the incidence of pneumonia in acute stroke patients admitted to a StrokeUnit, who underwent an expert swallowing assessment and standardized feedingprocedures.Methods: We evaluated after 2 years a second sample of 102 patients withacute stroke (83% ischemic, 17% hemorrhagic) admitted consecutively in a StrokeUnit over 6 months. All patients were systematically screened for dysphagia onadmission using the Standardized Bedside Swallowing Assessment, which involvesthree stages: general evaluation (level of consciousness, postural control, voluntarycough, voice quality and ability to swallow saliva), sipping water from a spoon,and, if safe, proceeding to drink water from a glass. If difficulty on swallowing waspresent, non-oral nutrition was considered, and the patient was referred to a speechtherapist for expert assessment.Results: Median NIHSS of the cohort was 9, mean age 66.2 years, 63% of patientswere male.Dysphagia was diagnosed in 18 (17.6%) patients; 6 (5.8%) patients developed

pneumonia during hospitalization. Data from a sample of 100 patients examined 2years before showed similar rates (dysphagia 27.0%, pneumonia 6.0%). Predictorsof increased risk of pneumonia were stroke severity, hemorrhagic (rather thanischemic) stroke, and the presence of a total anterior circulation syndrome.Discussion: The low incidence of aspiration pneumonia in our Institution supportsthe effectiveness of a standardized swallowing assessment and feeding procedure,and continuous training of the nursing staff. The incidence of pneumonia remainsin fact persistently low over time. However, there is a small group of patients athigher risk of developing pneumonia. These patients, who are likely to have anhemorrhagic stroke, severe symptoms and larger lesions, could benefit of moreselected strategies, to be eventually tested in this particular stroke subgroup.


QUALITY OF LIFE AND LONG-TERM FUNCTIONAL OUTCOME IN YOUNGPATIENTS AFTER AN ISCHEMIC STROKEA.A. Gongora, C. Mader, J.P. Arroyo, R. Garcia, A. LeyvaInstitute National of Neurology and Neurosurgery, Mexico City, Mexico

Background and purpose: The purpose was to evaluate quality of life and long-term functional outcome in patients under 45 years, after ischemic stroke to identifyvariables that will accurately predict quality of life and long-term functionaloutcome.Methods: This was a cross-sectional, descriptive correlational design. The modifiedRankin scale, Barthel index and Short From -36 (SF 36) were administered to 192stroke patients under 45 years. Subjects were interviewed 1 to 5 years after thestroke. Independent variables were age, sex, comorbidity, cause, and location ofstroke. Multiple regression analysis was used to predict quality of life.Results: 192 patients with a previous history of ischemic stroke <45 years of agewere included. The cause of the stroke was cardioembolism 20%, major bloodvessel atherosclerosis 9%, arterial dissection 28%, hypercoagulable state 15%, andnot determined 28%.The most important risk factors associated with the cause of the stroke werehypertension and major blood vessel atherosclerosis with 82% (p < 0.0005),diabetes mellitus and major blood vessel atherosclerosis 47% (p < 0.0005). Themean Barthel index was 95 with Barthel > 85 in 98% of cases. Rankin score was <

2 in 48%. (Rankin 3 – 4 = 52%). There was no significant difference between thecause of the stroke and the SF – 36 profile category results. The SF – 36 categorywhich was affected the most, was emotional role with over 30% of patients with anaverage value of less than 33.Conclusions: There seems to be no difference between the long term functionaloutcome and quality of life and the cause of the ischemic stroke. The identificationof alterations in the emotional role of patients furthers the need for supportprograms.


PROGNOSIS IN FIRST-EVER ISCHEMIC STROKE/TRANSIENT ISCHEMICATTACK PATIENTS WITH SIGNIFICANT EXTRACRANIAL CAROTIDARTERY DISEASEH.J. Lin, P.S. YehChi-Mei Medical Center, Yong Kang, Tainan, Taiwan

Background: The impact of significant extracranial carotid artery disease on theprognosis in patients with ischemic stroke or transient ischemic attack (TIA) isunclear in Taiwanese people, who have lower prevalence of such artery diseasethan Western people.Methods: From a prospective hospital-based registry of patients with first-everischemic stroke or TIA, we investigated the outcomes among those with newlyfound extracranial carotid artery disease > 50% stenosis. Data were collectedaccording to predetermined evaluation systems and diagnostic criteria, and thesubjects received regular follow-up. The composite outcome endpoint was subse-quent stroke, myocardial infarction, or vascular death after the index event. TheKaplan-Meier product-limit method was used to estimate the cumulative risk of theendpoint, and the Cox regression model for evaluating prognostic factors.Results: We enrolled 109 patients with a mean age of 69 years and 72% of men. Themechanisms of the ischemic events included 64% of large artery atherosclerosis,13% of small vessel disease, 9% of cardioembolism, and 12% of undeterminedetiology. All patients received medical treatments only. After a median follow-upduration of 21.1 months, 33 events developed, including 14 strokes, 2 acutemyocardial infarcts, and 17 vascular deaths. The cumulative risks of the endpointwere 7% in 1 month, 24% in 1 year, and 31% in 2 years. The Cox model analysesrevealed prior ischemic heart disease as a significant prognostic factor (hazard ratio,2.6; 95% confidence interval, 1.0-6.8)


Conclusions: Patients with first-ever ischemic stroke/TIA and newly found sig-nificant extracranial artery disease are predisposed to grave vascular outcomes, inparticular those with concomitant ischemic heart disease.


PLASMA GLUTAMATE LEVELS PREDICT FATIGUE AFTER TIA ANDMINOR STROKEA.B. Syed, L.M. Castell, A. Ng, C. Winward, P.M. RothwellStroke Prevention Research Unit, Department of Clinical Neurology, Universityof Oxford, Oxford, United Kingdom

Background: Post-stroke fatigue (PSF) is common and can be severe, even afterminor cerebral ischaemic events, but the mechanism(s) are poorly understood. Highlevels of plasma glutamate are associated with fatigue in multiple sclerosis and inchronic fatigue syndrome. We aimed to determine whether plasma glutamate andglutamine levels predicted PSF after transient ischaemic attack (TIA) and minorstroke.Methods: We studied consecutive patients with TIA and minor stroke from theOxford Vascular Study (OXVASC). Participants were excluded if they had a BarthelScore less than 18/20, a Rivermead Mobility Index (RMI) of less than 10/15, orMini Mental State Examination (MMSE) less than 23/30. Blood samples weretaken within one week of the presenting event and fatigue was assessed at home bya research nurse (blind to all biochemical analyses) 6 months after the event withthe Chalder Fatigue Scale. Enzymatic spectrophotometric techniques were used toassay plasma glutamine and glutamate levels.Results: Among 38 patients (28 stroke, 10 TIA, 15 male, mean age 74 yrs), 64%of stroke patients and 25% of TIA patients reported PSF. PSF was independentof age, sex, depression, smoking, medication and (among stroke patients) residualneurological deficit. However, there was a negative correlation between plasmaglutamine/glutamate ratio and PSF (r=-0.38. p=0.02). Plasma glutamate correlatedpositively with PSF (r=0.35, p=0.02). Discussion: Plasma glutamate and the glu-tamine/glutamate ratio appear to predict PSF 6 months after TIA or minor stroke.Further research is warranted to confirm the predictive value of these tests and todetermine whether the relationship is causal.

Brain imaging -- new developments

1 Brain imaging – new developments

THE USE OF [11C]-(R)-PK11195 LIGAND AND POSITRON EMISSIONTOMOGRAPHY IN ACUTE ISCHAEMIC STROKE: INSIGHTS INTO THEINFLAMMATORY PROCESSJ.A. Zavala, M.N. Perera, H.H. Ma, G. O’Keefe, H. Tochon-Danguy,U. Akermann, J. Ly, D. Reutens, C. Rowe, G.A. DonnanNational Stroke Research Institute, Australia; Centre for PET Austin Health,Australia

Inflammation after an ischaemic insult to brain tissue may have a key role inthe survival of viable hypoperfused tissue. Microglial and other inflammatorycells are rapidly activated (within hours) after pathological insults to the CNS.Macrophages accumulate after a period of days. Peripheral benzodiazepine bindingsites (PBBS) are mitochondrial membrane receptors in microglia and macrophages.(R)-PK11195 is a ligand that binds to PBBS. [11C] PET scan can be used as anon-invasive method of inflammatory response imaging. In this study we aimedto understand the spatial and temporal changes of inflammatory response afterischaemic stroke. We hypothesized that [11C]-(R)-PK11195 uptake levels increasenot only within the core of the ischaemic lesion but also in hypoperfused regions.Methods: We studied ischaemic stroke patients within 1 month of onset. Imagingwas performed including CT, CT perfusion or MRI/A. [11C]-(R)-PK11195 wasgiven intravenously followed by 3-dimendional dynamic acquisition (PET) over 60minutes. Distribution volume ratio maps were created using ipsilateral cerebellumas reference tissue. The PET images were then coregistered to DWI and perfusionmaps.Results: Fourteen patients were studied (median age 72 years, range 52 to 89years). PET scans were performed from 48 hours to 20 days after stroke onset. Ninepatients had corresponding perfusion scans. There was no [11C]-(R)-PK11195uptake within the first 5 days (5 patients) but increased uptake in 7 of 9 scansperformed from day 6 to day 20 after stroke onset. Increased uptake was found atleast as frequently beyond the infarct as within its core. For patients submitted toperfusion scans, there was little uptake in chronically hypoperfused areas comparedto elsewhere outside the infarct core.

Conclusions: More likely, [11C]-(R)-PK11195 uptake relates to macrophage (lateinflammatory response) rather than microglial activity. Because the uptake washigher in the normally perfused rather than chronically hyposperfused areas it maybe that reperfusion promotes the inflammatory response.


TC PERFUSION PREDICTS EARLY CLINICAL RESPONSE TOINTRAVENOUS THROMBOLYTIC THERAPY OF ISCHEMIC STROKE INTHE FIRST 3 HOURSM. Revilla, E. Palacio, F. González, C. Ramón, P. Sánchez-Juan,A. González-Mandly, E. Marco de Lucas, A. Gutiérrez, M. Rebollo, J. BercianoHospital Universitario Marqués de Valdecilla, Santander, Spain

Background and purpose: CT perfusion (CTP) and CT angiography (CTA) imag-ing techniques identify tissue in penumbra and intravascular thrombus in acuteischemic stroke, but their utility in the stroke therapy in the first three hours hasnot been defined. We evaluated if CTP and CTA findings conditioned early clinicalresponse to intravenous (iv) thrombolytic therapy in the first three hours.Methods: Forty-seven consecutive patients were treated with iv tPA according toSITS criteria. Additionally, 33 of them were studied with CTP+/-CTA before tPAadministration. We evaluated cerebral blood flow (CBF), mean transit time (MTT)and cerebral blood volume (CBV) images, CBF/CBV mismatch and arterial oc-clusion. Early recanalization was detected with control CTA+/-transcranial dopplerexamination. Early clinical response was measured by NIHSS at 0, 1, 24 and 72hours post-treatment.Results: Clinical evolution of patients by mean NIHSS was 13.5 (0 h), 10.8 (1 h),9.4 (24h) and 7.6 (72 h). Between 0 and 72 hours NIHSS differences, measured bymean rank, were as follows: 16.4 vs 19.4 if there was CBF abnormality (n=26);14.2 vs 19.3 if there was CBV abnormality (n=15); 15.2 vs 6.2 if there wasCBF/CBV mismatching (n=21) (p=0.01, Mann-Whitney U); 12.9 vs 8.5 if therewas arterial occlusion (n=15); and 13.2 vs 4.5 if there was early recanalization(n=12) (p=0.001, Mann-Whitney U).Conclusion: Findings of mismatch and early recanalization of arterial occlusionin CTP/CTA images predict a better early clinical response and may help in theselection of patients for iv thrombolysis of stroke in the first three hours.


INTRAPLAQUE HAEMORRHAGE IS ASSOCIATED WITH MULTIPLEDIFFUSION WEIGHTED IMAGING LESIONS IN SYMPTOMATIC PATIENTSWITH HIGH GRADE CAROTID STENOSISN. Altaf, S. Goode, P.S. Goode, J.R. Gladman, S.T. MacSweeney, D.P. AuerUniversity of Nottingham, Nottingham, United Kingdom

Introduction: Magnetic Resonance Imaging sensitively detects carotid intraplaquehaemorrhage (IPH) that is increasingly accepted as surrogate marker of plaqueinstability. The aim of this study was to investigate an association between IPH andthe presence, pattern and extent of cerebral acute and sub-acute ischaemic lesionsidentified by diffusion weighted imaging (DWI).Methods: 46 patients (18 females, mean age 71.5 years ± 10.7) with high gradecarotid stenosis (>70% stenosis) presenting with stroke, TIA or amaurosis fugaxwere prospectively recruited. All patients underwent MRI assessment of IPH inthe carotid artery and DWI of the brain. The presence, extent and pattern of DWIhyperintense lesions were compared with the IPH status of the presenting carotidartery.Results: 32 (69.6%) patients had evidence of IPH in the presenting carotid arteryand 26 (56.5%) had DWI lesions. The mean delay from the presenting symptomto MRI was 21.8 ± 18.5 days. Patients with carotid IPH were more likely tohave ipsilateral DWI lesions than those without IPH (22/32 [69%] vs. 4/14 [29%],P<0.05); had more lesions (mean 2.8 ± 3.3 vs. 0.7 ± 1.2, P<0.05); multiple lesionswere strongly associated with IPH (17/32 vs. 1/14), whereas the DWI hyperintenselesion volume was similar (mean 2.1 ± 3.4 ml vs. 1.7 ± 4.0 ml, P=0.6).18/32 (56.3%) of patients with IPH had multiple DWI lesions and 3/14 (21.4%)patients without IPH had multiple lesions (P<0.05).The association between carotid IPH and the presence of DWI lesions wasmarginally significant (Odds ratio = 3.8; 95% C.I. 0.8 – 18.2, P=0.09) aftercontrolling for stroke and time between symptom and MRI.Conclusion: Intraplaque haemorrhage in the presenting carotid artery is moder-ately associated with DWI ischaemic lesions and strongly with multiple lesionssupporting a thromboembolic pattern.

Poster SessionBrain imaging – new developments



TISSUE SWELLING WITHOUT HYPOATTENUATION ON NON-CONTRASTCT IS RARE BUT POTENTIALLY REVERSIBLE IN ACUTE ISCHEMICSTROKEI. Dzialowski, S. Subramaniam, V. Puetz, A. Krol, J.M. Boulanger, P.A. Barber,M.D. Hill, S.B. Coutts, T. Watson, A.M. Demchuk, for the Calgary CTA StudyGroupUniversity of Dresden, Dresden, Germany

Background: Early ischemic changes (EIC) on non-contrast CT (NCCT) canrepresent tissue hypoattenuation or cortical swelling and are both scored tradi-tionally in the Alberta Stroke Program Early CT Score (ASPECTS). However,only hypoattenuated brain tissue seems to be specific for ischemic core whereasisolated cortical swelling (ICS) might be reversible. We sought to assess ASPECTSignoring isolated cortical swelling (ASPECTS - ISC) and compare the incidence ofdeviation from classic ASPECTS.Methods: We studied ischemic stroke patients from 04/2002-07/2005 presentingwithin 24 hrs of onset in whom a NCCT was performed. We prospectively appliedASPECTS to all baseline NCCT scans by 3-reader consensus scoring any EIC.A normal scan scores 10, a complete middle cerebral artery infarction 0. In thesame session, we interpreted ASPECTS - ISC. We independently assessed follow-up ASPECTS at day 1-7. We determined the incidence of differences betweenASPECTS-ICS and ASPECTS and proportion of reversibility at follow-up.Results: We identified 335 patients with a mean age of 67 years, mean onset-to-NCCT-time of 286 min, median baseline NIHSS score of 7 and median ASPECTSof 10. We found ASPECTS - ICS > ASPECT scores in 7/335 (2.1%) of patients.In this group, mean onset–to-NCCT-time was 399 min and median ASPECTS-difference was 2 (range 1-4). In 5/7 patients follow-up imaging was availableand 2/5 (40%) showed higher final ASPECTS (> 1-point increase) than baselineclassic ASPECTS consistent with reversibility. In the population without baselineASPECTS difference, 16/247 (6%) of available follow-up images showed higherfinal ASPECTS than baseline classic ASPECTS.Conclusion: In our study, isolated tissue swelling was rare but likely to reverse.ASPECTS interpretation should ignore isolated cortical swelling to better representirreversible ischemic core.


LACTATE DOES NOT PREDICT INFARCT GROWTHV. Cvoro, J.M. Wardlaw, S. Muñoz Maniega, I. Marshall, P.A. Armitage,C.S. Rivers, M.S. DennisDivision of Clinical Neurosciences, University of Edinburgh, Edinburgh, UnitedKingdom

Background: In patients with acute ischaemic stroke, the mismatch between mag-netic resonance (MR) diffusion- and perfusion weighted imaging (DWI and PWI)was initially thought to predict infarct growth, but recent studies have questioned thestrength of this association. Lactate is a marker of early ischaemia and is elevatedin acute stroke lesions. N acetyl aspartate (NAA) which represents neuronal lossfalls more gradually. We examined whether elevated lactate or decreased NAA inmismatch tissue predicted infarct expansion.Methods: Patients with acute ischaemic stroke underwent diffusion tensor imaging(DTI), dynamic susceptibility contrast PWI, T2W and MR spectroscopic imaging(SI) at admission, days 5 and 14, and 1 and 3 months. A 0.5 cm diam. voxel gridwas superimposed on the baseline DTI and metabolite data were extracted from thenormal, mismatch and DTI lesion tissue. Infarcts were categorized into those withor without lesion growth.Results: 21 patients had DTI/PWI mismatch; 7 developed infarct expansion, 10 didnot (4 patients did not have follow up scans, and could not be included). Mean agewas 77 years (range 37-95), NIHSS 16 (range 7-29); 30% were first imaged <6hrs, 40% from 6-12 hrs and 30% from 12-24 hrs. Lactate (34.4±21.2 vs 18.9±9.7p<0.01) but not NAA (122.1±23.9 vs 115±30.6 p=NS) was elevated in mismatchtissue compared with normal brain. However, there was no difference in mismatchtissue at baseline in lactate or NAA between infarcts that expanded versus those thatdid not (lactate 39.3±20.5 vs. 23.9±27.3, p=NS; NAA 118.7±23.4 vs. 137±18,p=NS respectively). Furthermore there was no difference in mismatch tissue overthe first five days in lactate or NAA between lesions that grew and not.Summary: Lactate may be a marker of ischaemia, but its presence in mismatchtissue does not predict infarct growth. Infarct growth must be related to otherindividual factors.


MRI ON DAY 1 IDENTIFIES PATIENTS AT RISK FOR DELAYED STROKEPROGRESSION AFTER I.V. THROMBOLYSISR. Kern, K. Szabo, S. Bukow, M. Griebe, A. Förster, M.G. Hennerici, A. GassUniversitätsklinikum Mannheim, University of Heidelberg, Mannheim,Germany

Objective: Deterioration of clinical status after treatment with tPA for acute strokeis a possible outcome but difficult to predict on an individual basis. Besides anearly malignant course a delayed symptom progression can occur. In an approach tocharacterize the stability in the post-tPA phase we performed systematic follow-upMRI in this patient group.Methods: MRI (T2w, T1w, T2*w, DWI, TOF-MRA, PWI) was performed in 45acute stroke patients on the first day after CT-based tPA therapy (3h time-window).8/45 patients had an early malignant course and in 12/45 there was marked clinicalimprovement with MRI demonstrating successful therapy on day 1 and no furtherMRI was performed. In 25/45 patients MRI on the 1st and 7th day were compared.MRI findings on day 7 were either considered “improved” (= vessel recanalisation;resolution of hypoperfusion), “stable” (= no progress of infarct size, of hemorrhagictransformation [HT], or of vessel pathology) or “progressive” (= progression ofinfarcted or hypoperfused tissue size, HT or vessel pathology).Results: In 19/25 (76%) MRI was stable or improved on day 7, whereas 6/25(24%) were progressive: 6 patients showed new DWI lesions - 4 in the same, 2 in adifferent vascular territory. One patient developed HT between day 1 and 7. 67% ofpatients with persistent vascular obstruction and persistent hypoperfusion on day 1had progressive MRI findings on day 7 and did not improve clinically.

Fig. 1. MRI performed on day 1 and day 7 of a 68-year-old patient who underwent intravenousthrombolysis 2.5 hours after symptom onset DWI on day 7 showed multiple new acute lesions in theright PCA territory with a persistent perfusion deficit on TTP maps. On MRA, vascular obstruction ofthe left PCA even became more prominent. MRI characteristics were considered progressive.

Discussion: MRI on day 1 can identify those patients with a persisting unstablesituation at risk for stroke progression as indicated by hypoperfusion due toincomplete vessel recanalisation after thrombolysis. Identifying these patients isimportant for clinical management with close monitoring and blood pressuremanagement.


TRANSCRANIAL SONOGRAPHIC DELINEATION OF INTRACEREBRALHEMORRHAGE –A PROSPECTIVE MULTICENTER STUDYK. Meyer-Wiethe, R. Kern, S. Meairs, G. SeidelUniversity Hospital Schleswig-Holstein, Campus Lübeck, Lübeck, Germany

Background: A prospective study was performed in patients suffering from acuteintracerebral hemorrhage (ICH) in two German stroke centers to determine sensi-tivity, extent of midline shift (MLS) and lesion volume determined by transcranialultrasound (US).Materials and methods: US was performed with two systems (Philips SONOS5500 and HDI 5000) via the temporal acoustic bone window. We used sector trans-ducers at 2 MHz obtaining axial and coronary imaging planes. The sonographerswere blinded to the results of computed tomography (CT) performed in each patientas a reference.Results: 33 consecutive patients suffering from acute ICH (mean age 65 years,range 37 -84, median NIHSS 8/34) were investigated within 48h of symptom onset.


There was no difference in baseline characteristics between the patients from thedifferent centres. The localisations of the lesions were as follows: 23 basal ganglia,2 frontal, 5 parietal, 2 temporal, 1 occipital lobe. In 30 of 33 patients (91%), USidentified the lesion correctly. In three patients ICH could not be detected due toinadequate insonation conditions. Both US and CT showed no case of significantmidline shift of > 2 mm. CT depicted ventricular hemorrhage in 12 patients (US: 7patients – sensitivity 0.58, specificity 1.0). There was a close correlation betweenblood clot volume measured in CT and US (r = 0.85, P < 0.001, n = 30).Conclusions: In this prospective multicenter study US correctly diagnosed, lo-calized and measured intracerebral hemorrhage in patients with adequate bonewindows. In contrast, US depiction of ventricular hemorrhage showed highspecificity, but low sensitivity.This study is part of the UMEDS project (Ultrasonographic Monitoring and EarlyDiagnosis of Stroke) funded by the European Commission (QLG1-CT-2002-01518).


DIAGNOSTIC VALUE OF COMBINED ANALYSIS OF T2-WEIGHTEDGRADIENT ECHO IMAGING AND POSTCONTRAST TIME-OF-FLIGHT MRANGIOGRAPHY IN HYPERACUTE ISCHEMIC STROKES.I. Sohn, C.H. Sohn, H.W. Chang, S.H. Choi, S.R. Kim, H.C. ParkKeimyung University, DongKang Hospital, Andong General Hospital, Daegu,South Korea

Background: Identifying the composition and the length of clot may be importantin choosing the optimal treatment on acute thrombolysis. We assessed the diagnos-tic value of combined analysis of T2-weighted gradient echo imaging (GRE) andpostcontrast MR angiography (PC-MRA) in patients with acute middle cerebralartery (MCA) occlusion.Methods: From May 2004 and December 2006, consecutive 49 patients withocclusion of the MCA M1 segment within the first 6 hours from the onset ofsymptoms admitted to our emergency department were enrolled. Then all patientswere imaged using acute stroke MR protocol included GRE for susceptibilityvessel sign (SVS) and pc-MRA for the length of occlusion and had conventionalangiography. We classified into 4 groups as the visibility and length of GRE SVSand the signal gap of PC-MRA: longer clot length of GRE SVS than the signal gapof PC-MRA (group A), longer signal gap of PC-MRA then the length of GRE SVS(group B), the signal gap of PC-MRA without GRE SVS (group C), non-visible ofthe signal gap of PC-MRA and negative GRE SVS (group D). MR findings werecompared with findings of conventional angiography.Results: Among 49 patients, 42 (85.7%) patients with good MR imaging wereselected (23 men, mean age: 66.5). Group A was the most common type. Thedifference of the length and the gap in group A and B may be associated with theshape of thrombus. Long difference assumed oval shape clot and short differencetended to be barrel shape clot. Group C showed focal occlusion of the MCA M1segment by atherosclerosis. Group D had poor collateral circulation on conventionalangiography.Conclusions: Compared to independent analysis of GRE SVS or PC-MRA inpatients with acute arterial occlusion, combined analysis was showed more accurateinformation for the clot property and occlusion status.


GLOBAL CHANGES ON DIFFUSION WEIGHTED IMAGING (DWI) ANDMAGNETIZATION TRANSFER (MT) IN RELATION TO WHITE MATTERHYPERINTENSITIES: THE LADIS STUDYS. Ropele, A. Seewann, W. van der Flier, L. Pantoni, E. Rostrup, T. Erkinjuntti,L.-O. Wahlund, R. Schmidt, F. Barkhof, F. FazekasMedical University Graz, Graz, Austria

Objective: DWI and MT imaging should improve the detection and quantificationof cerebral tissue changes associated with white matter hyperintensities (WMH).Supportive data come mostly from single centres which studied only one modalityin small and selective groups of individuals. We therefore aimed to investigateand compare the sensitivity of these techniques for describing changes in normalappearing brain tissue (NABT) and WMH in a multi-centre setting.Subjects and methods: Within the LADIS study investigating the impact of WMHon 65 to 85 year olds without prior disability we obtained DWI and MT in 9centres with 1.5T whole body systems from different manufacturers. Lesions weredelineated on the FLAIR images; apparent diffusion coefficient (ADC) and MTratio (MTR) maps were calculated, co-registered and the respective values assessedglobally for WMH and NABT by means of histogram analysis. The mean value,the peak position (PP), and the relative peak height (rPH) were related to subjects’age and WHM severity.

Results: We analyzed 340 DWI and 177 MT scans and both modalities in 124subjects. ADC and MTR values showed a significant inter-site variation which wasstronger for the MTR. After z-score transformation multiple regression analysisshowed WMH severity and age as significant predictors for all ADC and MTRhistogram metrics of NABT. Only lesional ADC was increasing with WMH severitywhile such correlation was not seen with MTR.Conclusions: Despite some variation from a multi-centric collection of ADC andMTR data both modalities appear sensitive for changes in NABT which appearto occur with ageing and increase with the severity of WMH. However, the ADCwas more sensitive for discerning tissue changes within WMH and their relation tolesion size.


CHANGES IN BRAIN VOLUME 2 YEARS AFTER EC-IC BYPASS SURGERY:A PRELIMINARY SUBANALYSIS OF THE JAPANESE EC-IC BYPASS TRIALJ. Jinnouchi, K. Toyoda, T. Inoue, S. Fujimoto, S. Gotoh, K. Yasumori,S. Ibayashi, M. Iida, Y. OkadaNational Hospital Organization Kyushu Medical Center, Nippon Steel YawataMemorial Hospital, Kitakyushu, Japan

Background: Changes in cerebral blood flow (CBF) may be associated with brainatrophy, especially in patients with cerebral artery occlusive disease. However,previous studies have failed to find a significant relationship between CBF andbrain atrophy. Recently, Japanese extracranial-intracranial (EC-IC) bypass trial(JET) revealed that EC-IC bypass was effective for stroke prevention. JET is amulticenter, randomized, prospective study of patients with hemodynamic brainischemia due to cerebral artery occlusive disease. Here, we compared the changesin brain volume and cerebral hemodynamics in patients with and without EC-ICbypass surgery.Methods: We registered 10 Japanese patients with mild ischemic stroke for theJET. Six patients successfully underwent EC-IC bypass surgery and 4 were treatedmedically. We studied changes in brain volume on magnetic resonance imaging.We also examined the association of cerebral hemodynamics on single photonemission computed tomography with the changes in brain volume. The differencesbetween patients with and without EC-IC bypass were investigated.Results: The affected/unaffected ratio of the % brain volume declined in patientswithout EC-IC bypass surgery (p<0.02, n=4), and the affected/unaffected % rCBFratio increased in patients with the surgery (p<0.03, n=6). Acetazolamide reactivityincreased in the affected hemisphere of patients with surgery (p<0.01). Two-yearincrease (decrease) in acetazolamide reactivity of the affected hemisphere showeda significant positive correlation with 2-year changes in the affected/unaffected %brain volume ratio (R2 = 0.737, p=0.0007).Conclusions: Change in acetazolamide reactivity might be a good predictor forbrain atrophy in cerebral artery occlusive disease.


PERFUSION PATTERNS IN PATIENTS WITH SEVERE INTERNAL CAROTIDARTERY DISEASE USING PERFUSION-CTM.G. Delgado, V. Mateos, S. Calleja, R.L. Roger, P. Vega, C.H. LahozHospital Universitario Central de Asturias, Oviedo, Spain

Introduction: Cerebral perfusion profile of patients with chronic internal carotidartery disease has not been well studied.Material/Methods: Between January 2006 and January 2007, we studied consec-utive patients with severe internal carotid artery (ICA) disease by CT-Angiographyand Perfusion-CT. Hypoperfusion was defined as increased MTT, decreased CBFand CBV.Five perfusion patterns are described: cerebral hemisphere hypoperfusion (type 1),middle and anterior cerebral arteries territory (MCA and ACA) hypoperfusion (type2), MCA territory hypoperfusion (type 3), watershed territory hypoperfusion (type4) and normal pattern (type 5).Results: We identified 26 patients, 24 males and 2 females. 73% of patients hadcritical ICA stenosis and 27% of patients had ICA occlusion. Perfusion patternswere: 38% type 1, 31% type 2, 15% type 3, 8% type 4, 11% type 5. In ICAocclusion we only found 3 patterns: 43% type 1, 43% type 2 and 14% type 3.In critical ICA stenosis we found: 37% type 1, 21% type 2, 16% type 3, 10%type 4 and 3 patients (16%) had a normal perfusion study (type 5) with anteriorcollateral circulation preserved and ipsilateral posterior communicating cerebralartery absence. Six patients (26%) with abnormal perfusion study had a completecircle of Willis.Conclusions: The majority of patients with critical ICA stenosis or occlusion hadcerebral perfusion deficit. There were more patients with critical ICA stenosis than

Poster SessionBrain imaging – new developments


occlusion with cerebral hemisphere perfusion deficit (type 1). A normal perfusionstudy can be found in symptomatic patients.


INTER-OBSERVER AGREEMENT ABOUT THE PRESENCE ANDDISTRIBUTION OF BRAIN MICROBLEEDS IN ADULTS WITH STROKEC. Cordonnier, G. Potter, C. Jackson, C.L.M. Sudlow, J.M. Wardlaw,R. Al-Shahi SalmanDivision of Clinical Neurosciences, University of Edinburgh, Edinburgh, UnitedKingdom

Background: The increasing use of haem-sensitive gradient echo (GRE, T2*)sequences in magnetic resonance (MR) imaging of stroke has lead to frequentdetection of brain microbleeds (BMBs). If BMBs are found to be of diagnosticor prognostic significance, and are used for these purposes in clinical practice,observer variation in their assessment must be known.Methods: Two doctors assessed the MR imaging of 264 adults with stroke.BMBs were defined as small, homogeneous, round foci of low signal intensity onT2*-weighted images of less than 10 mm in diameter. Reviewers were blinded,and quantified BMBs on each side of the brain in the following locations: lobar(cortex, grey-white junction, deep white matter), deep (basal ganglia grey matter,internal capsule, external capsule, and thalamus), and posterior fossa (brainstemand cerebellum).Results: Thirty percent (95% confidence interval [CI] 26-34) of patients had 1BMB or more. Agreement about the presence/absence of BMBs at any location wasmoderate (75%, 95% CI 70 to 80; kappa 0.44, 95% CI 0.32 to 0.56). Agreementwas worse in lobar locations (81%, 95% CI 76 to 85; kappa 0.44, 95% CI 0.30 to0.58) than in deep locations (90%, 95% CI 86 to 93; kappa 0.62, 95% CI 0.48 to0.76) or the posterior fossa (95%, 95% CI 92 to 97; kappa 0.66, 95% CI 0.47 to0.84).Discussion: This study provides insight into one of the reasons why inter-observeragreement about the presence of BMBs is only moderate. Agreement was moderatein lobar locations, but substantial in deep areas and the posterior fossa. This may bedue to the existence of BMB mimics in lobar locations, especially vessel flow voids.We will explore agreement about BMB size and number, and ways of increasingagreement about lobar BMBs, in an effort to develop a BMB grading scale.


STUDIED ON BROCA’S APHASIA BY DIFFUSION TENSOR IMAGINGY. Zhang, S. Wang, C. Wang, X. Zhao, Y. WangBeijing Tiantan Hospital, affiliated with Capital University of Medical Sciences,Beijing, China

Background and purposes: Diffusion tensor imaging (DTI) is sensitive to the rateand direction of water diffusion, The fibers distributing of language functionalareas exhibit that extensive and complicated relationship between language areasand other areas. We studied Broca’s aphasia cases by the technique in order tocomprehend clinic symptom of the aphasia type.Methods: DTI in axial covering the entire brain volume were obtained in thirtyvolunteers and thirty Broca’s aphasia patients who suffered from left hemispheredamaged after stroke. Used SIEMENS DTI software to post process and to measurefractional anisotropy (FA) value and display the course of Broca’s area and themirror side.Results: The results showed that the left Broca’s area FA of volunteers was 0.3081± 0.0325, the mirror side was 0.3069 ± 0.0630, and there were no significantbetween them (p>0.05). On the other hand, the left Broca’s area FA of Broca’saphasia patients was 0.2578 ± 0.05260, right corresponding area was 0.3063 ±0.0562, there were significant between them (p<0.05).Conclusions: The Broca’s area fibers of Broca’s aphsia were damaged. and usingDTI can analyse the fibers distributing of language functional areas, offer anatomyinformation for clinic and explain the baffling of neurology of widen activatingsignal language areas on cortex. Key words: DTI, Broca’s Aphasia, Broca’s area


CT PERFUSION IN ASSESSMENT OF BRAIN CIRCULATION IN PATIENTSWITH STENOSIS OR OCCLUSION OF INTERNAL CAROTID ARTERYG. Witkowski, P. Richter, A. Rozenfeld, R. Poniatowska, A. Dowzenko,H. Jarosz, D. RyglewiczInstitute of Psychiatry and Neurology, Warsaw, Poland

Background and purpose: The risk of stroke due to severe stenosis or occlusionof internal carotid artery (ICA) is higher in patients with insufficient collateralbrain circulation. Transcranial Doppler Ultrasonography (TCD) is routinely appliedfor assessment of collateral circulation. Computer tomography perfusion (CTP)imaging is used in the clinical practice especially in the evaluation of brain bloodflow during acute stroke, transient ischemic attacks (TIA), in epileptogenic foci anddifferential diagnosis of brain tumors. CTP also can be applied in assessment ofbrain circulation in patients with unilateral stenosis of ICA.The aim of the present study was to correlate the signs of collateral circulation inTCD with the results of CTP in patient with symptomatic carotid arterial occlusionor stenosis.Methods: 17 patients hospitalized in First Department of Neurology, Institute ofPsychiatry and Neurology with TIA due to stenosis or occlusion of ICA wereintroduced to the study. 4 patients were previously treated with the intravas-cular occlusion (Gold Baloon) because of carotid cavernous fistula and brainaneurysm. In Doppler examination blood flow through ophthalmic artery andanterior communicant artery was estimated. CTP was routinely applied in all cases.Results: Unilateral cerebral hypoperfusion was more pronounced in case of insuffi-cient cerebral collateral circulation. In these patients Mean Transit Time parameter(MTT) was prolonged for about 30-40%.Conclusion: CT perfusion can be considered as a complementary method to TCD.This examination helps to estimate the influence of arterial stenosis or occlusionon cerebral blood flow. It can also predict the increased risk of ischemic stroke inpatients with carotid stenosis.

Interesting cases

1 Interesting cases

MULTIPLE LOCALISED CERVICOCRANIAL ARTERY DISSECTIONSDEVELOPING AFTER AORTIC ARCH DISSECTIONA. Lovrencic-Huzjan, V. Vukovic, S. Antic, I. Galinovic, V. DemarinUniversity Hospital, Zagreb, Croatia

A 56-year old male without vascular risk factors, presented with syncope, shockand encephalopathy afterwards. The diagnosis of aortic arch dissection was set.Aortic arch was operated, aortic valve was replaced and warfarin was introduced.During the first year of follow up, he was symptom-free. However carotid colorDoppler (CCD) showed a newly developed localized distal right common carotidand proximal internal carotid artery dissection. During the second year of followup, he was ten months symptom-free. Afterwards, he had two attacks of vertigoand nausea, without vomiting. No other neurological symptoms or signs developed.He started complaining on the pain in the right shoulder. CCD confirmed persistentright carotid dissection, and a new right subclavian artery dissection was displayed.This is the first report of a previously healthy patient who developed consecutivelocalized carotid and subclavian dissections during two-year follow-up of aorticarch dissection.

2 Interesting cases

BILATERAL ISCHAEMIC STROKES IN A 33 YEAR OLD WOMAN WITHIN AFEW MONTHS OF PITUITARY IRRADIATIONG. Beamond, K. Murray, P. Keston, C. SudlowUniversity of Edinburgh, Edinburgh, United Kingdom

Cranial irradiation is associated with increased stroke risk, usually years afterexposure. We describe a 33 year old female with recurrent ischaemic strokes,starting only 4 months after cranial irradiation.She presented initially with acromegaly and a pituitary macroadenoma on MRbrain scan. After transphenoidal hypophysectomy, she received octreotide and45Gy of radiotherapy to the pituitary bed in 25 fractions over 5 weeks. Subsequenthypopituitarism was treated with DDAVP, hydrocortisone, thyroxine and the oralcontraceptive.4 months later she developed sudden onset apraxia, agnosia, acalculia and agraphia,


transient episodes of right arm jerking, right-sided weakness and dysphasia. MRbrain scan, MR angiography and CT angiography showed bilateral hemisphericwatershed infarction, complete occlusion of the cavernous sinus portion of the rightinternal carotid artery (ICA) and severe stenosis of the intracavernous left ICA. Shehad no vascular risk factors, and investigations for causes other than radiotherapywere negative. She received aspirin and then warfarin. Her neurological deficitalmost completely resolved over the next few weeks, and subsequent CT perfusionscan showed good perfusion bilaterally. 11 months after her initial strokes, warfarinwas discontinued and aspirin re-started. She re-presented 2 weeks later with furtherleft hemisphere ischaemic strokes and radiological evidence of complete occlusionof both intracavernous ICAs. Warfarin was re-started. Several months later she hadresidual language difficulties and mild right upper limb dysfunction.We assume that her recurrent strokes are due to large artery radiation-induced arteri-tis with associated thromboembolism. Such early development of this complicationhas not been described previously.

3 Interesting cases

HEAD POSITION DEPENDENT VERTEBROBASILAR TIA’SA.A. Weck, H. Hungerbühler, A. Mironov, G. SchweglerCantonal Hospital of Aarau, Aarau, Switzerland

Based on studies using duplex ultrasonography mechanical compression of theextracranial vertebral artery (VA) during rotation of the head is not very rare, buthardly ever symptomatic with signs of vertebrobasilar ischemia. As a general rule, tobecome symptomatic prearranged anatomical conditions in the vertebrobasilar cir-culation must be present: flow restriction in the contralateral VA (occlusion/severestenosis/hypoplasia) and lack of functioning posterior communicating arteries.We report a 62 years old man who suffered from blurred vision after head rotationto the right side. Back in the neutral position his visual symptoms resolved rapidly.Driving car was a major problem for him due to impaired sight by turning his headto watch the traffic on the right side.Ultrasonography showed a normal right VA and a severe hypoplasia of the left VA(1.2 mm diameter). Both posterior communicating arteries were absent. Transcra-nial colour coded sonography documented a massive decrease of flow velocities inthe posterior cerebral artery during head turning to the right reproducing his typicalsymptoms (video demonstration) and a transient reactive hyperemia of BFV abovebaseline values by return to the neutral position. Angiography of the right VA wasnormal in neutral position. By turning the head to the right, the VA became stenoticat level C6-C7 (V2 entrance zone). After maximal head rotation the bloodflowdistal of the C6/7 segment stopped due to mechanical occlusion at that level. Thecervical spine CT revealed no relevant osteophyte formation at C5/6 and C6-C7.We suggest a rotational obstruction due to extraluminal cervical fascial bands ofthe longus coli muscle. Surgical evaluation is planned.The TCCD monitoring of the posterior cerebral artery is a reliable and reproduciblemethod to detect cases of vertebrobasilar insufficiency dependent on head rotationand mechanical compression.

4 Interesting cases

INTERNAL CAROTID ARTERY OCCLUSION DUE TO WEGENER’SGRANULOMATOSIST. Das, W. Sunman, R.H. Harwood, J. Beavan, S. MunshiNottingham University Hospitals NHS Trust, Nottingham, United Kingdom

A 44 year-old male presented to the Emergency Department (ED) with profuseepistaxis. Three months previously he was seen by otorhinolanrygologists forleft sided otalgia, tinnitus and hearing loss. They noted left middle ear effusionand a polyp in the post-nasal space. Computed Tomography (CT) of the neckdemonstrated an ulcerated pharyngeal mass. A nasopharyngeal biopsy showeda granulomatous lesion with central necrosis. A diagnosis of tuberculosis wasconsidered as he had severe ankylosing spondylitis and was being assessed foranti-TNF therapy. He had no vascular risk factors.In the ED, he developed sudden-onset aphasia and right hemiplegia, with a leftHorner’s syndrome. CT head scan showed a hyperdense left middle cerebral arteryand early signs of cerebral infarction. He had greatly raised inflammatory markers,a mild anaemia and normal renal function. Electrocardiography, urinalysis andchest X-ray were normal. His cANCA (PR3) level was high and indicative ofWegener’s granulomatosis.Magnetic Resonance Imaging showed a nasopharyngeal mass involving the hor-izontal petrous carotid canal and occlusion of the left internal carotid arteryfrom its origin to the middle cerebral artery M1 segment. Catheter angiogramshowed no evidence of pseudoaneurysms in external carotid artery branches as thecause of his epistaxis. He was treated with intravenous methylprednisolone and

cyclophosphamide. However his aphasia and weakness persisted. Retrospectiveanalysis of his initial CT neck scan showed signs of inflammation within the leftcarotid sheath.Wegener’s Granulomatosis is a rare cause of central nervous system infarction,usually due to small vessel vasculitis. We present a case of carotid artery thrombosisrelated to extravascular granulomatous involvement of a large vessel. There is noreported association between ankylosing spondylitis and Wegener’s granulomatosis.

5 Interesting cases

CLINICAL PRESENTATION OF INTERNAL CAROTID ARTERYDISSECTION: REPORT OF 10 CASESI. Divjak, M. Jovicevic, A. JovanovicInstitute of Neurology, Clinical Centre Novi Sad, University of Novi Sad, NoviSad, Yugoslavia

Background: Internal carotid artery dissection (ICAD) is a recognized cause ofstroke, particularly in young adults. It may occur spontaneously or result from localtrauma. Clinical diagnosis may be difficult and the classical triad of symptoms isuncommon. Imaging plays a pivotal role in the diagnosis of ICAD. The aim was toanalyze the spectrum of clinical presentation in 10 ICAD patients, with a specialemphasis put on a patient presenting with Horner’s syndrome and facial and neckpain as the only symptoms of ICAD.Methods: Ten patients with ICAD aged 35-45 (mean age 42.1 years) wereevaluated in the period January 2001 – December 2006. The ICAD diagnosis wasestablished using MRI, MRA and duplex sonography in all cases. In one case CTangiography was additionally performed.Results: Four patients presented with facial pain, Horner’s syndrome and con-tralateral sensorimotor deficit. One patient presented with facial and neck pain andHorner’s syndrome only. Five patients presented with contralateral sensorimotordeficit, with or without speech impairment. Two patients had traumatic ICAD (onewhile unloading sacks of corn and the other after sudden head turning) and othereight patients had spontaneous ICAD. MRI revealed infarction in 9 patients, whilein the patient presenting with facial and neck pain and Horner’s syndrome diffusionMRI did not show evidence of infarction. Good outcome (defined as modifiedRankin score of 0-2) was seen in all patients. Complete recanalization of ICADwas associated with favorable prognosis.Discussion: The spectrum of clinical presentation of ICAD is variable. ICAD isnot necessarily accompanied by infarction on diffusion MRI.

6 Interesting cases

BILATERAL SPONTANEOUS CAROTID ARTERY DISSECTIONH. Weitenberg, M. Uyttenboogaart, J. De Keyser, G.J. LuijckxUniversity Medical Centre Groningen, Groningen, The Netherlands

Background: Spontaneous carotid artery dissection is a cause of ischemic strokein the young. In this case we report a patient with spontaneous bilateral carotidartery dissection.Case: A previous healthy 40 year old man presented with an isolated Hornersyndrome on the right side. Several weeks before patient had an upper airwayinfection. This was followed by a period of right sided headache. Besides theHorner syndrome neurological examination was normal. MR angiography (MRA)revealed a dissection of the right carotid artery from bifurcation to skull base witha fresh trombus and a dissection with a double lumen of the left carotid artery.To prevent tromboembolic complications patient was treated with antiplatelettherapy for a year. Control MRA showed normalisation of the right and a slightpseudo-aneurysm of the left carotid artery.Discussion: The pathogenesis of spontaneous carotid dissection is at presentuncertain. If multiple vessels are involved fibromuscular dysplasia, Ehlers-Danlos,Marfan syndrome, osteogenesis imperfecta and alpha-1-antitrypsin deficiencyshould be considered. These conditions were ruled out in this patient. Recentlyan association between spontaneous carotid artery dissection and upper airwayinfection has been reported. Suggested pathophysiological mechanisms are localinfection of the arteries, or mechanical by sneezing. The higher incidence of carotidartery dissections in autumn is an argument for the possible relationship withupper airway infection. This case demonstrates that after excluding an underlyingconnective-tissue disorder upper airway infection should be considered as a causeof spontaneous bilateral carotid artery dissection.

Poster SessionInteresting cases


7 Interesting cases

TRANSIENT ISCHEMIC ATTACK AND LIVEDO RACEMOSA –ANTIPHOSPHOLIPID SYNDROMEL. Valdemar, A. Marinho, G. LopesDept. Neurology, Hospital Geral de Santo António, Porto, Portugal

Background: The Antiphospholipid Syndrome (APS) is an autoimmune disor-der characterized by persistently elevated titters of antiphospholipid antibodies,associated to thrombotic events, without vasculitis, typically affecting females(82%). Ischemic stroke is reported in 30% of the patients with APS, with transientischemic attacks (TIA) representing one third of them, but only 7% at diseaseonset. Livedo Racemosa (LR) is a rare pathologic skin condition occurring in someimmunologic disorders. When associated to cerebrovascular disease it is calledSneddon’s Syndrome (SS).Case presentation: A fifty three years old male patient, with hypertension,suddenly became nauseated, vomited, and had vertigo and disequilibrium. Theneurological examination showed dysarthria, right dysmetria and ataxic gait. Healso had an exuberant LR. He completely recovered from all symptoms andsigns in less than 24 hours. The cerebral MRI showed bihemispheric ischemicleukoencephalopathy and no signs of acute ischemia with diffusion technique. Highand persistent titters of autoantibodies were found (anticardiolipin, antiB2GPI andlupus anticoagulant). AngioMRI, transcranial Doppler, ultrasonography of cervicaland renal arteries, ECG, transesophagic echocardiogram and serologic studies werenormal. Anticoagulation was started.Discussion: This patient, with a cerebellar TIA, presented as a SS and fulfilsclinical, imagiological and laboratorial criteria for APS with LR. Ischemic cere-brovascular disease can be a manifestation of APS, but its association with LR israre. This clinical presentation, in a male patient, is even less frequent. Relationshipbetween SS and APS with LR is not clear, as these two entities are clinicallyindistinguishable and classified as a continuous spectrum of a disease.

8 Interesting cases

AN UNRECOGNIZED CAUSE OF THUNDERCLAP HEADACHE:REVERSIBLE CEREBRAL VASOCONTRICTION SYNDROMEK. Koopman, M. Uyttenboogaart, G.J. Luijckx, J. De Keyser, P.C. VroomenUniversity Medical Centre Groningen, Groningen, The Netherlands

Background: An unrecognized cause of thunderclap headache (TCH) is ReversibleCerebral Vasoconstriction Syndrome (RCVS). We describe 3 patients with RCVS.Cases: Three women, aged between 40-55 yrs, presented with TCH. One patienthad a history of migraine with sumatriptan abuse and one had an exacerbationof Crohn’s disease. SAH was ruled out by CT scan and CSF examination. MRvenography was normal. Brain MRI in 2 patients showed infarction in the posteriorregions. Cerebral angiography (DSA) showed diffuse beading in one patient. Shewas suspected of having primary angiitis of CNS (PACNS) but did not respond totreatment with immunosuppressants. All had increased velocities on TCD. RCVSwas diagnosed and they were treated with calcium channel inhibitors. This led toclinical improvement and normalisation of TCD within weeks.Discussion: RCVS is characterised by a reversible segmental vasoconstriction ofthe cerebral vessels, most commonly occurring in women aged 20-50 yrs. It is asso-ciated with conditions such as migraine, certain drugs and pregnancy. The strikingpresenting feature is TCH, with or without focal signs. CSF is (near) normal, incontrast to PACNS and SAH. Brain imaging findings vary between normal andinfarction, particularly in the posterior circulation. Segmental vasoconstriction onDSA does not differentiate between vasculitis and RCVS. One of the hallmarks ofRCVS is the complete reversibility of vasoconstriction on TCD. Treatment is withcalcium channel blockers, in severe cases combined with prednisone.These patient cases underline that (1) TCH, normal CSF, and MRI or angiographicabnormalities may point to RCVS, (2) TCD is helpful in diagnosing RCVS and (3)proper diagnosis of RCVS has important therapeutic consequences.

9 Interesting cases

BROCA’S APHASIA ELICITED BY WERNICKE’S AREA DAMAGEDY. Zhang, N. Wie, H. Chen, N. Zhang, Y. WangBeijing Tiantan Hospital, affiliated with Capital University of Medical Sciences,Beijing, China

Aphasia is one of the common symptoms in acute and chronic stroke patients, manypostmortem and radiologic studies have documented the pattern of associationsbetween brain lesions and aphasic syndromes, such as Broca’s aphasia is mainlydue to a lesion damaged of the left inferior frontal area, namely Broca’s area,

Wernicke’s aphasia is usually associated with a lesion of the posterior part of thelateral temporal areas, namely Wernicke’s area, conduction aphasia is associatedwith lesion of the left arcuate fasciculus or of the left supramarginal gyrus, and soon. However, we found that not all aphasia types met with the pattern.A fifty-six man suffered from aphasia after stroke, he understanded what wordsmean, but had trouble performing the motor or output aspects of speech, and hecouldn’t communicate through writing. The results of Western Aphasia Batteryshowed he was Broca’s aphasia, but we found the damaged lesion was Wernicke’sarea instead of Broca’s area. On the other hand, we studied regional blood volume(rCBV) and regional cerebral blood flow (rCBF), mean transit time (MTT), andtime to peak (TTP) of Broca’s area of the patient by perfusion-weighted imag-ing, and compared with that of the contralateral hemisphere, we also measuredthe metabolic rate of N-acetylaspartate (NAA), choline (Cho), and creatine (Cr)by magnetic resonance spectroscopy, and compared the results with that of thecontralateral hemisphere. We found the Broca’s area were in a hypoperfusion andhypometabolism state compared with the contralateral hemisphere, maybe this canexplain why the type of this case was Broca’s aphasia while damaged lesion wasWernicke’s area. The aphasia case challenged the anatomy of aphasia theory.

10 Interesting cases

RECURRENT TRANSIENT ATAXIC HEMIPARESIS REVEALING AHYPOGLYCAEMIC PARANEOPLASIC SYNDROMEP. Olivier, J. Zapf, P. MichelCentre Hospitalier Universitaire Vaudois, Switzerland

Aim: We describe a patient presenting with recurrent ataxic hemiparesis as aconsequence of hypoglycaemia due to a benign pleural fibrous tumour.Case description: A 80 year old hypertensive women with a history of a pleuraltumor resected 11 years earlier presented four episodes of mild transient righthemiparesis and mild confusion upon awakening on four consecutive mornings.On admission, mild right ataxic hemiparesis was present. At the end of her firstnight in the stroke unit, worsening of the right ataxic hemiparesis and decreasedvigilance occurred.Investigations: Head CT and CT-angiography were performed. Blood sugar, in-suline, cortisol, C-peptide, insuline-like growth factor-I (IGF-I) and IGF-II weremeasured. Thoraco-abdominal CT and PET were done.Results: Cerebral CT with angiographic sequences didn’t reveal any abnormality.Hypoglycemia of 1.2 mmol/l was detected during worsening and neurological statusnormalized promptly with treatment. Recurrent hypoinsulinemic hypoglycaemia(glycemia 1.0-2.2 mmol/L; insulin <2mUl/l (5.0-18.0); C-peptide < 0.30 ugL(0.7-3.0)) was documented and required continuous glucose perfusion. Cortisol andIGF-I were normal. IGF-II (740ng/ml) and "big IGF-II” fraction (212 ng/ml; 28.7%of total IGF-II, N<15%) were elevated. An extrapulmonary tumor occupying mostof the right hemithorax was found, consistent with a benign pleural tumor onPET. Thoracotomy revealed a benign pleural fibrous tumor of the fuso-cillary type,identical to the one resected 11 years earlier. After complete resection, no moresymptoms occurred, and IGF-II and its fractions normalized.Conclusion: This elderly lady presented with recurrent ataxic hemiparesis sug-gestive of a lacunar warning syndrome. Demonstration of severe hypoglycaemiaduring these episodes led to the detection of a fuso-cillary pleural tumor. Thisstroke-imitating paraneoplastic syndrome was related to hypersecretion of a fractionof IGF-II, and was completely cured by resection of the tumor.


THREE CASES OF CEREBRAL PEDUNCULAR INFARCTION WITH PUREDYSARTHRIAG.S. Kim, J.H. Lee, S.A. Choi, J.H. ChoNHIC Ilsan Hospital, Goyang Shi, Kyungki Do, South Korea

Background: Pure dysarthria can be seen in patients with stroke involving thecorticobulbar tract, usually at the lenticulocapsular, pontine base, or cortical areas.Infarction of cerebral peduncle is rare. Its manifestation is dysarthria-clumbsyhand or dysarthria-one arm weakness. Pure dysarthria due to cerebral peduncularinfarction is extremely rare. We experienced three cases of cerebral peduncularinfarction with pure dysarthria.Cases: All cases showed inaccurate articulation in the labial sound than palatal orlingual sound. One with left cerebral peduncular infarction showed slight impairof right sided hopping in neurologic examination. The others with right cerebralpeduncular infarction showed dysarthria only. There is no severe stenosis of relevantarterial system in magnetic resonance angiography. In short follow-up period, theyshowed full recovery of symptoms.Discussion: Dysarthria of infratentorial origin has been described in infarctions of


the basis pontis and cerebellum. We also observed pure dysarthria in patients withan infarction of the cerebral peduncle. Focal injury to corticobulbar tract withincerebral peduncle seemed to be a possible cause of pure dysarthria.


CEREBRAL INFARCTION IN ACUTE (MYELOMONOCYTIC) LEUCEMIAAT INITIAL PRESENTATIONS. Koskina, A. Tavernarakis, I. Xydakis, E. Mamouzelos, E. Koutra, N. MatikasEvangelismos Hospita, Athens, Greece

Acute leukemia (A.L) is a rare cause of stroke in young adults. We present the caseof a patient in whom stroke was the first manifestation of the disease.The patient, a 48 years old woman,was admitted to the hospital to investigateunremitting fever. Few days later, she suffered a stroke (left hemiplegia),and then asecond one (right hemiparesis and aphasia), and developped thrombosis of the leftsuperficial femoral vein. The patient didn’t have any known risk factors other thana mild hypertension. Successive computerized tomography scans showed mainly ahypodense area in the right temporo-parietal region.Blood tests were performed, that revealed evidence of disseminated intravascularcoagulation (D.I.C.) and positive lupus anticoagulant,while other ancillary inves-tigations, including lumbar puncture, thoracic and oesophageal cardiac ultrasoundand carotid artery triplex were normal.Finally,the patient was diagnosed from a bone marrow biopsy,as having acutemyelomonocytic leukemia. Treatment was initiated, but the patient died twomonths later from multiorgan failure and sepsis.There are several mechanisms causing thrombotic episodes in acute leukemias.D.I.C.,positive lupus anticoagulant or antiphospholipid antibody in serum, leucosta-sis syndrome in leukemias with leukocytosis, or direct viral damage to endothelialcells in virus induced leukemias are mechanisms encountered in the literature. Theparticularity of the present case is the fact that stroke, probably caused by D.I.C.,was the first manifestation of A.L.Thus, it is of the utmost importance that young adults presenting with stroke, beinvestigated thoroughly to diagnose any underlying hematologic malignancy andinitiate, as soon as possible,the apropriate treatment.


ARTERIAL OCCLUSION AND STROKE AFTER CISPLATINCHEMOTHERAPYK.A. Pasco, P. HartSt George’s Hospital, London, United Kingdom

Background: Malignancy as a risk factor for cerebrovascular disorders and as athromboembolic risk is well described. Cisplatin, used alone or in combination,has become standard treatment for various solid tumours. Side effects of cisplatininclude neurotoxicity and Goldhirsch et al first described the association of acutestroke with cisplatin use in 1983, suggesting this agent increases ones risk of strokeabove that of the tumour alone.Case reports: We describe three cases of acute ischaemic stroke post cisplatinchemotherapy, all supported by clinical and radiological evidence. The cases areaged 17, 36 and 44 years, two female; one male and each with different tumoursubtypes (neuroendocrine, cervical adenocarcinoma and medulloblastoma). Allreceived cisplatin based chemotherapy in conjunction with other treatment. All hadMiddle Cerebral Artery (MCA) territory infarction 2-10 days post cisplatin therapy.All had received more than two treatment cycles. In two cases arterial occlusionwas confirmed, in the third it was found to be highly likely.Discussion: Cisplatin regimes are implicated in acute stroke and the mechanismsmay be multi-factorial. Our cases could be explained by direct cisplatin endothe-lial toxicity and enhanced platelet aggregation thus leading to arterial thrombusformation. Work also suggests that cisplatin predisposes to a hypercoagulablestate through an acquired protein C deficiency or increased von willebrand factor.Other reports suggest cisplatin induced vasospasm through hypomagnesaemia. Ad-ditionally, our cases demonstrate the close temporal relationship between cisplatinadministration and acute stroke onset in the absence of other cardiovascular riskfactors and support the suggestion that stroke tends to occur after several treatmentcycles. A review in 2006 identifying cisplatin use as a stroke risk factor, moreso than other chemotherapeutic agents, also found the MCA territory most oftenaffected, which lends further weight to our own case findings.


PRIMARY ANGITIS OF CENTRAL NERVOUS SYSTEM IN A PATIENT OFACQUIRED DEFICIENCY SYNDROME (AIDS), A CASE REPORTA. Al Memar, N. Akhtar, A. TripAtkinson Morley Wing at St George’s Hospital London, London, UnitedKingdom

Background: Cerebral vasculitis in patients infected with human immunodefi-ciency virus (HIV) is usually secondary to infectious agents rather then HIVitself. It is extremely rare to have cerebral vasculitis where no other cause can befound and role of HIV is postulated in genesis of cerebral vasculitis. This is acase report of 44-year-old, was diagnosed to have HIV in January 2005, when hepresented with features of fever headaches, night sweating and hairy leukoplakia.He presented to us in December 2005 with the features of, personality change washemi paresis and cortical blindness. He had mild leukopenia of 3.7 ANA, ANCAand other antibodies of vasculitic screen were negative. MRI- was suggestive ofgross abnormality involving left occipital lobe. Occipital lobe biopsy was consistentwith clear-cut vasculitis.Methods: We compared the clinical and biopsy results between our case andpreviously published cases.Result: PCR and histological findings looking into the possibilities of HSV 1+2,CMV, Adeno virus, VZV, JC virus and HIV were negative there was no evidenceof cerebral lymphoma. In the view of negative specific viral staining, and absencesof antibodies of vasculitic screen suggest the diagnosis of primary angitis of thecentral nervous system in a patient infected with HIV.Discussion: To our knowledge only eight cases are reported in literature in whichprimary angitis of central nervous system was suspected to be associated with HIV.This case Illustrates a rarity of condition but does raise the strong hypotheticallink between HIV and primary vasculitis. In our case histological studies werecompatible with a diagnosis of primary angitis of the central nervous system, butthe pathogenic role of HIV in the genesis of the vasculitic process cannot beelucidated.


FAMILIAL SNEDDON’S SYNDROMES. Llufriu, A. Cervera, S. Amaro, A. ChamorroStroke Unit, Hospital Clinic, Barcelona, Spain

Background: Sneddon’s syndrome is a non inflammatory arteriopathy character-ized by livedo reticularis and cerebrovascular disease. It is an uncommon causeof stroke in young people and it has been associated to the Antiphospholipidsyndrome. It mainly occurs sporadically, although few familial cases have beenreported. In familial cases the most common pattern of inheritance is autosomaldominant, although the gene responsible is not known.Case report: A 34 year-old woman with right hemiparesis, progressive ataxiasince early childhood and frequent migraine attacks. In the last year, she hadexperienced episodes of dizziness, and dysphagia. The neurological exam dis-closed mild cognitive impairment, right hemiparesis, horizontal nystagmus and leftcerebellar signs. On clinical exam, prominent skin lesions -mainly in the thighs-were consistent with livedo reticularis. Global atrophy, multiple and confluentsubcortical ischemic strokes and abundant microbleeds were found on brain MRI,whereas angio-MRI, carotid ultrasonography and transesophageal echocardiogramwere normal. Prothrombotic states and antiphospholipid antibodies were ruled outas appropriate. A skin biopsy was non specific. Two sisters and one brother hadlivedo reticularis and a history of early-onset stroke (neuroimaging available). Theremainder sister had 2 abortions and livedo reticularis but not neurological deficits.Her father died from a myocardial infarction at the age of 54.Discussion: Sneddon’s syndrome is a devastating cause of stroke in the young withfew cases reported in the literature. We add a new family of this entity which itsmain clinical findings, imaging and immunological traits are reviewed.


INTRACRANIAL PRESSURE AND CEREBRAL BLOOD FLOWCORRELATIONI. Voznjuk, M. Odinak, S. Golokhvastov, N. TsyganMilitary Medical Academy, Russian Federation

Background: Current imaging methods are unable to differentiate neuroinflamma-tory processes such as endothelial activation. We imaged activated endothelium ina mouse model of acute stroke using a novel iron oxide nanoparticle MRI contrastagent - MNP-Psel - targeted to the adhesion molecule P-selectin.



Methods: MNP-Psel was compared to the non-targeted iron oxide MRI contrastagent Feridex. In ten C57 Black 6 mice the left middle cerebral artery wastransiently occluded for 30 minutes while a body temperature of 36.5 ± 0.5 C wasmaintained throughout and after the procedure. 24 hours after reperfusion, six micewere injected with MNP-Psel and four with Feridex at a dose of 2.8 mg Fe/kg.MRI scans (T1, T2, and T2*) were obtained at 9.4 T, and infarct was identifiedfrom T2 maps. Subtraction images (pre-contrast – post-contrast) from T1 maps, T2maps and T2* were obtained at 1, 36, 72 and 108 minutes after contrast injection.Changes in T1 and T2 values in the contralateral hemisphere were subtracted fromthe stroke hemisphere to reveal the infarct tissue-specific contrast accumulationeffect. After imaging, the mice were euthanized and brain sections through thecortex and striatum were taken for iron histochemical staining by the Prussian bluemethod.Results: MNP-Psel and Feridex had similar T1 effects. T2* images demonstrateda peri-infarct prolonged contrast effect with MNP-Psel but not Feridex. T2 sub-traction maps revealed a prolonged MNP-Psel infarct-specific contrast effect notseen with Feridex. Iron staining on vessel walls in the infarct hemisphere indicatedendothelial localization of the MNP-Psel contrast agent.Discussion: Injection of MNP-Psel, but not Feridex, resulted in a prolonged infarct-specific iron oxide contrast effect associated with endothelial iron accumulation.This suggests that MNP-Psel accumulated in infarct and peri-infarct areas viaP-selectin-binding on activated endothelium. This demonstrates the feasibility ofusing MRI to image specific neuroinflammatory processes that contribute to theevolution of stroke injury.


“PERIPHERAL” VERTIGO OF “CARDIAC” ORIGIN. TWO CASES OFCARDIOEMBOLIC PICA INFARCTS, ASSOCIATED WITH PFO,PRESENTING AS A MISLEADING ISOLATED VERTIGOS. Beretta, P. Santoro, C. FerrareseUniversity of Milano-Bicocca, Monza, Monza (MI), Italy

Background: Isolated vertigo as a manifestation of acute cerebellar infarct in theposterior inferior cerebellar artery (PICA) territory was first reported by about 25years ago. Nonetheless, this presentation of cerebellar strokes still represents aclinical challenge, especially in young patients.Results: We report two cases (man, age 48; woman, age 59) presenting withstereotyped-rotatory dizziness, nausea, vomiting, inability to stand unaided andnystagmus, which were evaluated in the ER by a ENT and a neurologist and wereadmitted to the ENT department with the diagnosis of “peripheral vertigo”. Suchinitial manifestation was followed by delayed neurologic signs within 2 days afterthe onset. An acute, large inferior cerebellar ischemic stroke due to occlusion ofthe right PICA was demonstrated by CT scan in both cases. Both patients wereadmitted to the acute stroke unit and received treatment with antiplatelet agents andosmotic diuretics. Both cases demonstrated no evidence of atherothrombosis in thecerebral arteries, normal heart rhythm and morphology, except from a significantpatent foramen ovale (PFO). Both patients recovered completely within 3 monthsand underwent a successful percutaneous closure of PFO.Discussion: Early diagnosis of cerebellar infarction simulating vestibular neuritisis difficult and these patients are likely to be excluded from rtPA therapy. Carefullylooking for subtle neurological signs and the eventual use of diffusion weightedMRI may overcome this problem in selected patients. Finally, PFO needs to beinvestigated as a potential cause of this disorder in young patients.


CARDIAC AMYLOIDOSIS - A RARE CAUSE OF TRANSIENT ISCHAEMICATTACK (TIA)D.M. CollasWatford General Hospital, Watford, Hertfordshire, United Kingdom

Introduction: A 41-year old male with a TIA is presented.Method: After an initial 30 minute episode of motor and sensory loss in the rightarm and leg the patient re-presented within 24 hours with a stroke involving face,arm and leg. Limb weakness resolved within 1 hour and facial weakness within1 day, mild dysarthria persisting. There had been right amaurosis 6 weeks before,and 1 week of mild confusion, 1 month of lethargy, knee pain, and weight lossand urticarial rash over a 5 month period. He had a cholesterol of 6 but no othercommon risk factors for stroke, being a normotensive non-smoker.Results: Computed tomography (CT) revealed a left striato-capsular infarct withmild mass effect confirmed on magnetic resonance imaging (MRI), with signalincrease on diffusion weighting, and normal carotids on duplex scan and MRangiography. Erythrocyte sedimentation rate (ESR) was 40 but C-reactive protein

(CRP) less than 3 and tests for vasculitis and thrombophilia were negative.Electrocardiogram (ECG) showed sinus rhythm with right bundle branch blockand echocardiogram a hyperechoic myocardium with restricted filling. CardiacMRI showed global wall thickening, oedema suggesting an inflammatory process,global subendocardial late enhancement characteristic of amyloid [1], and impaireddiastolic function. There was an IgG lambda paraproteinaemia with partial immuneparesis and 20% plasma cells in bone marrow. Serum amyloid protein scan showedno visceral deposition and rectal biopsy was inconclusive.Discussion: Systemic symptoms preceding amaurosis and a contralateral TIA andan abnormal echocardiogram gave a clue to a rare cause of cardio-embolic stroke,primary amyloidosis. This was confirmed by cardiac MRI, bone marrow and serumelectrophoresis. Anticoagulation is indicated despite sinus rhythm in view of a 33%risk of cerebral embolism [2] due to impaired cardiac functionReferences: [1] Maceira AM et al. Circulation. 2005;111:186-93. [2] Hausfater Pet al. Scand J Rheumatol. 2005;34:315-9.


POSTERIOR INFERIOR CEREBELLAR ARTERY DISSECTION CAUSINGANEURYSM AND TRANSIENT ISCHEMIC ATTACK: ANEURYSMDISAPPEARANCE AND PREVENTION OF RECURRENT BRAIN ISCHEMIAWITH CONSERVATIVE TREATMENT. A CASE REPORTD. Muentener, A. Mironov, A. Valavanis, R.W. Baumgartner, H. HungerbuehlerUniversity Hospital of Zurich, Zurich, Switzerland

Isolated spontaneous dissection of the posterior inferior cerebellar artery (PICA)causing aneurysm formation is rare. Up to 70% present with subarachnoidhemorrhage and remaining cases with ischemic events.We present a 49 year old man with vertebro-basilar transient ischemic attack (TIA).Magnetic resonance imaging showed two cerebellar DWI lesions in the territoryof left PICA with a normal MR angiography (MRA). Transforaminal duplexsonography revealed a stenotic signal, but identification of the affected artery wasnot possible. Digital subtraction angiography (DSA) performed 2 days after MRAshowed a stenosis and a fusiform aneurysm of the proximal left PICA likely due todissection.The patient was treated with oral aspirin. One month later another vertebro-basilarTIA occurred. Aspirin was replaced by oral anticoagulation. No further ischemicevent was observed in the next 12 months. DSA performed 6 months after symp-toms onset showed complete resolution of the aneurysm, which retrospectivelyconfirmed the diagnosis of PICA dissection. Anticoagulation was discontinued.We conclude that dissecting aneurysms of the PICA can resolve spontaneously.Anticoagulation may be an efficient and safe treatment in patients with PICAdissection causing TIA and aneurysm.


ISCHEMIC STROKE AFTER CHEMOTHERAPY WITH CISPLATIN,ETOPOSIDE AND BLEOMYCIN FOR A TESTICULAR NON-SEMINOMACARCINOMA: A CASE REPORTM. Vikelis, M. Xifaras, A. Basta, G. GekasGeneral Hospital of Nikea, Nikea, Greece

Background: Vascular toxicity associated with cisplatin-containing chemotherapyfor testicular cancer is a side effect that its frequency has not been describedprecisely. Nevertheless, major vascular complications such as stroke, myocardialinfarction and pulmonary embolism seem to occur infrequently.Case report: We present the case of a 37-year-old man that was treated witha combination of cisplatin, bleomycin and etoposide (BEP) for a testicular non-seminoma carcinoma. Two days after the first course of BEP he experiencedsudden-onset right hemiplegia with involvement of the lower facial muscles andright homonymous hemianopia. A MRI-scan of the brain revealed an extendedinfarction in the vascular territory of the right middle cerebral artery. Angiographyof the head and neck arteries revealed a completely thrombotic right internal carotidartery. There was no evidence of coagulopathy, vascular, or endocardial disease thatmay have led to a cerebrovascular accident and the patient had no known vascularrisk factor except for smoking. The time sequence between the chemotherapyand the stroke clearly suggest a causal relationship between them in our patient.Particularly, a cisplatin-related cause is probable, since such adverse effects areunknown with bleomycin or etoposide.Conclusion: Whereas a cause and effect relationship is probable for some vas-cular events following chemotherapy, some cases may represent coincidence ormay be disease related. Several factors appear to be responsible for cisplatinvascular toxicity, such as an increased thrombogenicity and vascular spasm due tohypomagnesaemia. Potential late vascular toxicity has also to be taken into account.



CEREBRAL LIPIODOL EMBOLISM WITHOUT PULMONARYINVOLVEMENT DURING TRANSCATHETER ARTERIALCHEMOEMBOLIZATIONB.G. Yoo, J.K. Kim, J.H. Ko, E.G. KimKosin University College of Medicine, Busan, South Korea

Background: Use of lipiodol in transcatheter arterial chemoembolization (TACE)for hepatocellular carcinoma (HCC) treatment has been found to be associated witha number of complications. However, cerebral lipiodol embolism has been rarelyreported. All of the reported cases of cerebral lipiodol embolism have pulmonaryinvolvement.Case report: A 68- year-old woman with advanced HCC underwent a secondcourse of TACE at the hepatic artery using a mixture of 30 mL lipiodol. Duringthe procedure he had dysarthria and deteriorated consciousness, followed by statusepilepticus and semicoma. There was no breathing difficulty, and skin examinationwas normal. Two hours later, MRI showed restricted diffusion in the cortex andcortical-subcortical junction, both cerebral and cerebellar hemispheres. Two dayslater, a follow-up MRI showed multiple cerebellar and cerebral infarcts with hem-orrhagic transformation. A transcranial doppler with a bubble study demonstrateda right-to-left shunt.Conclusion: We report a patient with cerebral lipiodol embolism without pul-monary involvement during TACE of HCC. TACE is not an innocent procedureand clinicians must be alert to complications such as right-to-left shunt. To reducethe risk of lipiodol embolism, a smaller lipiodol dose and survey for detection ofintracardiac shunt before the procedure can be considered.


CROSSED CONDUCTION APHASIA: A CASE REPORTY. Zhang, Y. Liu, X. Zhao, C. Wang, Y. WangBeijing Tiantan Hospital, affiliated with Capital University of Medical Sciences,Beijing, China

Conduction aphasia is known as a disconnection syndrome, and characterized by acomparatively reduced ability to repeat spoken language and well comprehension.According to Geschwind, conduction aphasia results from damage to the arcuatefasciculus, one major pathways connects Broca’s and Wernicke’s areas. BothBroca’s and Wernicke’s areas are left intact. We found that not all conductionaphasia cases met with these standards.A 50 year-old highly educated, right-handed man suffered from aphasia after stroke.And he had fluent paraphasic expression, severe impairment of repetition and poorcomprehension without motor impairment. Western Aphasia Battery showed thathe was conduction aphasia, and the damaged lesion was right hemisphere, thereby,he was crossed conduction aphasia case. We found that he had severe deficitsin repeating no-words and short memory capacity, although he failed to showcognitive limitation in phonological output tasks.By diffusion tensor imaging, we found that the fractional anisotropy (FA) valuesof right arcuate fasciculus were smaller than that of the mirror side, that wereto say the right major pathways connects Broca’s and Wernicke’s areas weredamaged. On the other hand, we also found that the FA values of right Wernicke’sarea were smaller than that of mirror area, that means the Wernicke’s area wasalso damaged, maybe this damaged can explain why the case had fluent outputand poor comprehension and his language disorders liked Wernicke’s aphasia,that were to say he was “Wernicke-like” crossed conduction aphasia. The caseproved that three supposes of conduction aphasia:Wernicke-Geschwind’s theoryof disconnection, the defect pattern of auditory-speech shorten memory and theBidirectional pattern.


STROKE IN THE PUERPERIUM AND PERI-PARTUM PERIODK.M. Tan, A. CarrollNational Rehabilitation Hospital, Rochestown Avenue, Dun Laoghaire, Co.Dublin, County Dublin, Ireland

Stroke in the puerperium and peri-partum period, although rare, causes devastatingconsequences. The following are 3 cases.Case 1: A 28 year old woman with an acute right frontal headache developedleft sided weakness 10 days post caesarean section. CT brain showed right basalganglia, internal capsule and parietal infarcts. CT angiogram showed right carotiddissection. Intravenous heparin was commenced followed by warfarin.Case 2: A 32 year old woman had a severe headache, nausea and vomiting in

the second stage of labour. Three hours post delivery she developed left sidedweakness, left homonymous hemianopia and a GCS of 11/15. CT Brain showed alarge intracranial haemorrhage in the right basal ganglia.Case 3: A 39 year old woman with a severe headache and systolic bloodpressure of 200mmHg eight days post-partum, developed left sided weakness.CT Brain showed infarction in the right middle cerebral artery territory. CTangiogram showed bilateral carotid dissection. Intravenous heparin was started andsubsequently warfarin.Five weeks later she had a new headache in her right temple. MR angiogram showedfurther occlusion of the right vertebral artery. Intravenous methylprednisolone wasstarted followed by oral prednisolone in view of possible recurrent inflammatoryprocess.All 3 women improved. Each patient could walk independently with residualweakness to varying degrees. None had speech or swallow difficulties. One patienthad mild cognitive impairment not evident in everyday function.The challenges in rehabilitation include strategies to cope with infant care, de-pression in not participating fully in nursing and a fear of not being able to bondwith the newborn if hospitalised for a long period. It is vital that patients receiverehabilitation as soon as possible.


IRON DEFICIENCY ANAEMIA AND STROKEJ.C. Wöhrle, M. Silomon, M. Kaspers, R. WernerKatholisches Klinikum Koblenz, Koblenz, Germany

Background: Stroke is a rare complication of iron deficiency anaemia that maybe related to reactive thrombocytosis with thrombophilia or to impaired oxygendelivery.Case report: A 45 year-old woman had a severe right-sided hemiparesis uponawakening resolving within 30 minutes. Subsequently, symptoms recurred andprogressed to hemiplegia and global aphasia on arrival in our stroke unit. She was acigarette smoker and had felt weak for the last 3 months; she had hypermenorrhagia.Immediate cranial computed tomography showed early signs of infarction in theterritory of the left middle cerebral artery (MCA). In the presence of ubiquitoushigh blood flow velocities, transcranial ultrasound revealed a significant reductionin the left MCA (systolic/diastolic velocities 89/34 cm/s left vs. 193/77 cm/sright). There were small hypoechogenic plaques in both proximal internal carotidarteries. Thrombolysis was withheld because of severe microcytic anaemia withhaemoglobin 5.9 mg/dl and thrombocytosis (698.000/μl). We found iron deficiency(iron 22 ug/ml, ferritin 4 ng/ml). Screening for thrombophilic factors and cardiacembolism was negative. The patient received blood transfusions, low dose heparin,and aspirin. Within days, sonography and MR angiography revealed normalizedflow patterns in both MCAs. Hypermenorrhagia remained the only cause ofanaemia. She became ambulatory, but had persistent Broca’s aphasia and a severespastic brachial paresis.Discussion: While extracranial artery thrombosis is recognized as sequelae ofsevere iron deficiency anaemia (e.g. Caplan et al. Neurology 1984), isolatedintracranial artery occlusion is exceedingly rare. Intracranial artery obstruction mayhave been caused by embolism from proximal sites with complete resolution of theoriginal thrombus or by local MCA thrombosis.


ISCHEMIC STROKE OR MULTIPLE SCLEROSIS RELAPSE AFTER 44YEARS OF REMISSION?Y.-M. Huang, O. Nordmark, M. Lee, H. NaverUppsala University Hospital, Uppsala, Sweden

Background: MS and stroke differ in many ways but share a final commonpath with neuronal and axonal loss. It is not known how these diseases influenceeach other, nor their coincidence or best strategy concerning therapy and strokeprevention.Here we report a patient with 60 years history of MS, complete neurologicalremission for >40 years and progressive gait and balance problems with spasticityfor the last 5 years.Case presentation: Woman born 1923, had at age 24, 27 and 29 right-sided opticneuritis, at age 31, 32 and 33 episodes of vertigo and diplopia, and at age 33an episode of slight left leg weakness and spasticity with minor sequelae. CSFat age 31 and 33 showed slight mononuclear pleocytosis, normal protein andglucose. ACTH courses given at age 31 and 33 had beneficial effect. After 1956,she remained healthy till 2001 when gait and balance problems and left-sidedspasticity developed insidiously, making unaided walk difficult but she did notconsult physician.



Fig. 1. Brain CT: slight bilateral signal changes in capsula externa.

In May 2005, she was seen in emergency room because of sudden expressivedysphasia and confusion. Symptoms disappeared after 45 min. BP 170/80. When83 and having UTI, she had head trauma after fall in Febr, 2006, followed byincrease left-sided hemiparesis. Brain CT: unchanged. CSF: >2 oligoclonal IgGbands in CSF absent in serum. MRI showed on T2 multiple pericallosal lesions(Fig. 1, A), on T1 no gadolinium enhancement (B) and no fresh ischemic lesionsat apparent diffusion coefficient image (C) but leucoaraoisis. After UTI treatment,she was mentally intact but had severe gait and balance problems and spasticitypersistent at 10 m follow-up.Questions:1. Did the patient have stroke in 2001? TIA in May 2005?2. Cause of left-sided hemiparesis in February 2006, stroke or MS relapse?3. Best available management both from medical and social aspects?4. How can imaging help in diagnosis?


THROMBOLYSIS IN CAPSULAR WARNING SYNDROME (CWS): FOURCASE-REPORTEDR.M. Vivanco, A. Rodriguez-Campello, A. Ois, M. Gomis, C. Pont,E. Cuadrado, J. RoquerStroke Unit, Hospital del Mar, Barcelona, Spain

Background: CWS, first described by Donnan in 1993, is characterized bystereotyped episodes of motor or sensory deficit (usually more than 3 episodesin 24 hours). It is associated with a high risk of imminent lacunar infarctionwith permanent deficits resembling those of CWS in more than 40% of patients.Pathophysiology of CWS has not been well characterized. Ischemia mechanism isprobably due to small vessel penetrating disease and hemodynamic factors associ-ated as well as molecular mechanisms. Diffusion-weighted imaging (DWI) showsacute lesions in majority of cases. There are no proven therapies for preventingcompleted stroke in this unstable situation. The use of thrombolytic treatment wasnot described previously in these cases. We report four cases of CWS treated withfibrinolysisPatients and methods: Four patients were evaluated between February 2005 andDecember 2006 (0.5% of ischemic stroke). 3 of them were male (75%). Mean agewas 67.5 years. Hypertension was the main vascular risk factor. Stroke symptomswere compatible with lacunar syndrome (motor pure or sensitive-motor), with meanNIHSS 10. Number of episodes varied between 3 and 6 (mean 4). All patients weretreated with rtPA in the first three hours since last episode.Results: 3 patients remained asymptomatic after treatment (mRS 0) and did notpresent any other episode. One patient presented a new episode after the rtPAwith left hemiplegia (mRS 4). Blood pressure was monitored in all patients duringepisodes and no decrease coinciding with the clinical worsening was observed.In all patients laboratory tests, CT scan, non-invasive studies for carotid andintracranial artery disease were normal. Atrial fibrillation was found in one patient.

DWI was performed in three patients. The two asymptomatic patients with DWI didnot show any acute lesion. The other one showed acute lesion in AChA territory.Conclusion: We suggest that fibrinolysis is a therapy to take into account in CWS.DWI was normal in treated patients who recovered.


MIGRAINE WITH AURA AS AN ISOLATED PRESENTING SYMPTOM OFCEREBRAL VENOUS THROMBOSISR.E. Petrea, J.R. Romero, S. Seshadri, J. Viereck, V. Babikian, C.S. KaseBoston University School of Medicine, Boston, MA, USA

Background: Headache is the most common presenting symptom of cerebralvenous thrombosis. In the majority of cases headache is a non-specific symptomaccompanied by other neurological signs.Methods: Case report. A 27 year old right handed man presented to the emer-gency department (ED) with a 30 minute headache with visual aura and sensorysymptoms. He reported a generalized, excruciating throbbing headache followed bybright lights in front of both eyes lasting about 15 minutes. A tingling march fromthe right side of his face down to his right upper and lower limbs progressed overanother 15 minutes. All deficits resolved completely in 30 minutes. He had a totalof four episodes only with visual aura and headache, the first one 2 months prior tothe ED presentation. Two of these episodes were evaluated by head CT (computertomography) and lumbar puncture both of which were normal. His neurologicalexamination was entirely normal.Results: MRI (magnetic resonance imaging) of the brain revealed a hyperintensesignal on T1 and FLAIR sequences in the superior sagittal sinus suggestingthrombosis. Tiny venous infarcts were seen in the parietal cortex bilaterally. MRV(magnetic resonance venography) confirmed lack of flow in the superior sagittal,left transverse and sigmoid sinuses, extending to the jugular bulb. Laboratorywork-up revealed a nephritic syndrome and an abnormal activated protein Cresistance. He was treated with anticoagulation with no recurrence of his headache.Conclusions: New onset recurrent migraines with aura can be an isolated pre-senting symptom of cerebral venous thrombosis. Any headache that progressesin an unusual fashion should also prompt the consideration of cerebral venousthrombosis and the appropriate imaging for diagnosis.


EOSINOPHILIC VASCULITIS: A RARE CAUSE OF DOLICHOECTASIA OFTHE CAROTID AND INTRACRANIAL ARTERIESM. Labuda, S. LanthierUniversité de Montréal, Faculty of Medicine; CHUM-Hôpital Notre-Dame,Montreal, Canada

Background: Present in 12% of strokes, intracranial arterial dolichoectasia (IADE)is associated with atherosclerosis and elastic tissue diseases, and attributed tointernal elastic lamina disruption. Goals: To clarify pathogenesis of IADE and tostress the diagnostic challenge of eosinophilic vasculitis (EV).Methods: Case report.Results: A 46-year-old man presented in 2005 with <2-minute episodes of aphasiaand right limb tremor and weakness without altered vigilance. In 1999, similarepisodes were attributed to carotid and IADE and treated with warfarin for 6 monthsfollowed by aspirin. He denied allergies and other neurological, constitutional orsystemic symptoms, except pruritus since 2003; investigation had shown skininfiltration by lymphocytes and idiopathic blood eosinophilia. Physical exam wasnormal, as well as brain MRI and 24-hour EEG witnessing episodes. Selectivebrain angiography revealed progression of carotid and IADE and no arterialstenosis. Temporal artery biopsy revealed trans-mural non-necrotic infiltration bylymphocytes and eosinophils, multinucleated cells, and histiocytes forming a singlegranuloma. Blood tests showed increased white cell count (14.8 x 109/l; 37%eosinophils) and IgE level, negative HIV, aspergillus and hepatitis serologies, andnormal inflammatory, prothrombotic and vitamin B12 workups. Echocardiography,thoraco-abdominal CT, abdominal CT-angiography, and stool exam were normal.Skin tests indicated pollen and cat hypersensitivity. On bone marrow biopsy, cells(50% eosinophils) had no chromosomal abnormalities. His neurologic symptoms,pruritus and eosinophilia resolved with prednisone (1mg/kg/d for 4 months, taperedover 3 more months). We did not find previous reports of EV with IADE.Discussion: EV can cause IADE by disrupting the internal elastic lamina. In thiscase, idiopathic blood and tissue eosinophilia is consistent with hypereosinophilicsyndrome, but vasculitis as the sole organ infiltrated by eosinophils is unexpected.Differential diagnosis includes atypical Churg-Strauss syndrome.



MYOCARDIAL INFARCTION DURING STROKE THROMBOLYSIS: A CASEREPORTM. Herrera, M.E. Erro, J. Gállego, R. Muñoz, B. Zandio, J.A. VillanuevaHospital de Navarra. Pamplona, Pamplona, Spain

Background: One potential harm of thrombolytic therapy for brain infarction isthrombolytic-induced breakup of thrombi (mostly in the heart or aorta) leading tonew strokes, myocardial damage or limb ischemia.Methods: We report a patient who presented a myocardial infarction during strokethrombolysis.Case report: A 78-year-old man was admitted to the neurology stroke care unit forsudden aphasia and right hemiparesis. The NIH Stroke Scale (NIHSS) score was15. Brain CT scan was normal. A transcranial doppler revealed left proximal ACMoclussion.The ECG showed an atrial fibrilation with normal repolarization pattern.The patient was started on tPA therapy 150 minutes after stroke onset. During druginfusion, he developed hypotension, oxygen desaturation and bradycardia. He alsosuffered neurological deterioration with somnolence, complete right hemiplegiaand left forced gaze deviation. An anaphylactic adverse reaction was suspected andtPA infusion was stopped. A new CT scan ruled out brain hemorrhage. An ECGwas perfomed with signs of anterior acute miocardial damage. A thoracic CT scanruled out an arterial ascending aortic dissection. A coronary angiography confirmedocclusion in distal territories of left anterior descending and circumflex arteriesand mechanical recanalization was unsuccessfully tried. Two days later a trans-esophageal echocardiogram revealed an anterolateral myocardial infarction,denseleft atrial spontaneous echo contrast, non complicated atheromatosis in the aorticarch, and no intracardiac thrombus.Discussion: This patient had an ischemic stroke of a likely embolic origin froman intraauricular thrombi; tPA therapy could have favoured the fragmentation ofthis thrombi and thus facilitated a second-step coronary embolism. A shock duringfibrinolytic therapy should raise the possibility of this rare complication, confirmedby a simple electrocardiographic recording.


LEFT VENTRICLE NONCOMPACTION, MYOPATHY, DYSMORPHICFEATURES AND STROKE IN A YOUNG WOMANA. Mendes, F. Silveira, M. Garcia, E. AzevedoHospital S. João, University of Porto, Porto, Portugal

Background: Left ventricle noncompaction is a rare congenital cardiomyopathycharacterized by numerous prominent trabeculations and intratrabecular recessesin the ventricles. Heart failure is the most common presenting condition. Othermanifestations include arrhythmia and thromboembolic events. We present a caseof stroke associated to noncompaction of left ventricle, unspecific myopathy andsome dysmorphic features.Case report: A 20 years old woman was referred to our department for investi-gation after having a left middle cerebral artery ischemic stroke with aphasia andright hemiparesis. At 5 months of age she was diagnosed with a hypertrophiccardiomyopathy. Since her 15 years she complained with limbs fatigue and musclespasms, and an unspecific myopathy was diagnosed after investigation includingelectromyography and skeletal muscle biopsy. There was no relevant familialhistory. She presented some dysmorphic features such as short stature, webbedneck, low hairline at the nape of the neck and bilateral cubitus valgus. Thepatient had a good mental state and recovered from the aphasia and most ofthe right hemiparesis. There was evidence of a distal muscular atrophy in lowerlimbs. Stroke investigation disclosed a left internal carotid artery (ICA) occlusion.Ecocardiography and cardiac catheterism identified left ventricle noncompaction.Blood investigations were negative, including for serologic, immunologic andprothrombotic changes, as well as for muscle enzymes. Cariotype was 46, XX. Shestarted anticoagulation. Six months later she remained clinically stable, and therewas no recanalization of left ICA.Discussion: Cardioembolism was the probable cause of stroke. Although leftventricle noncompaction may be associated with neuromuscular involvement, likein Barth syndrome, it doesn’t usually affect women. Her dysmorphic featuresresemble Turner or Noonan syndromes, which could not be confirmed in thiscase. To our knowledge, there is no description in the literature of the congenitalassociations found in this patient.


POST-PARTUM CEREBRAL ANGIOPATHY: REPORT OF THREE CASESJ. Willeit, M. Furtner, M. Knoflach, T. Gotwald, S. KiechlDept. of Neurology, Innsbruck, Austria

Background: The post-partum angiopathy (PPA) is a rare disease of unknownetiology characterized by segmental vasoconstriction of medium-sized and largecerebral arteries. No standard management has been established.Methods: We report 3 cases with PPA, in whom serial magnetic resonancetomography (MRT) and MR angiography were performed.Results: All three women delivered a healthy child after an unremarkable preg-nancy. They received dopamin agonists to suppress lactation. After hospitaldischarge the patients complained of severe headache. Two developed seizures,hemianopsia and hemiparesis. Blood pressure monitoring showed intermittent andsevere hypertension. The cerebral MRT demonstrated high signal intensity lesionsin the occipito-parietal/frontal region on T2-weighted and diffusion-weighted imag-ing. Most of the lesions showed high signal intensity on ADC maps compatiblewith vasogenic edema, and MRA showed narrowing of large and medium-sizedcerebral arteries. The results of extensive tests for cerebral vasculitis were negative.In two of the patients, lowering of blood pressure was accompanied by completerecovery of the neurological deficits and normalization of imaging findings. In thelast patient, blood pressure lowering was without any clinical effect. Aggressivemedical therapy including high-dose methylprednisolon and nimodipin resulted ina gradual improvement.Conclusion: These cases provide evidence that apart from hypertension, ergotderivates might act as a trigger of PPA. PPA can be complicated by both va-sogenic edema and ischemic stroke. The clinical course is highly heterogenous,ranging from self-limited and fully reversible to vasculitis-like progressive forms.Apart from blood pressure-lowering therapy some patients may require high dosecorticosteroids.


DISSECTING FUSIFORM VERTEBRAL ANEURYSM IN AN ADOLESCENT:ENDOVASCULAR TREATMENTC. Semedo, M. Manita, J. Reis, P. Raimundo, P. Esperança, J.M. CândidoCentro Hospitalar de Lisboa - Zona Central, Lisboa, Portugal

Introduction: Intracranial aneurysms are rare below 18 years old, correspondingto less than 1% of all the treated aneurysms. Most centres handle no more than 1case a year. Moreover, dissecting fusiform aneurysms of posterior circulation havea difficult approach and have been traditionally treated by parent vessel occlusion.Nowadays aneurysmatic stent placement, with or without coiling, has become avaluable treatment option.Case report: 15 years-old female patient with a right sided cervical and occipitalsevere headache, thunderclapping, with nausea and dizziness, holding on for amonth in spite of oral analgesiae. The patient had no history of previous traumaand had been otherwise healthy. Neurological examination was normal. The CTrevealed an isodense, calcified lesion in the right lateral-cistern and the MRI showedto be probably a displasic aneurysm partially thrombosed with 13mm of diameter,with light mass effect over the adjacent medullary side, continuous with the RVA(right vertebral artery). The cerebral angiogram revealed a displasic/dissectingfusiform aneurismatic formation of the distal RVA (V4), with PICA (posteriorinferior cerebellar artery) originating directly from the aneurismatic sac. A Leo“stent” (4,5mmx20mm) was placed in VA between the two extremities, through allthe aneurism length, and after that partial occlusion of the remaining aneurismaticsac was excluded with coils, with the purpose to induce the progressive thromboticocclusion of the residual sac, in order to protect PICA and allow collateral circula-tion. The control angiogram, 2 days after, confirmed the complete exclusion of theaneurysm, with the main vessel patency. The patient was discharged asymptomatic.Clinical and transcranial doppler revaluation was normal 3 months after.Conclusions: This is an interesting case combining an aneurysm in an earlyage with an fusiform aneurysm difficult approach, turning it into a therapeuticchallenge. In these situations endovascular stent placement and embolization canbe an effective and safe method.




LEFT VENTRICULAR PAPILLARY FIBROELASTOMA, A RARE CAUSE OFPOSTERIOR CIRCULATION INFARCTIONJ. Damásio, R. Almeida, A. Furtado, L. Caiado, A. TunaHospital Geral de Santo António, Oporto, Portugal

Background: Papillary fibroelastoma compromise approximately 7.9% of benigncardiac tumours. They are most commonly detected on the cardiac valves, beingthe intra-ventricular localization very rare. Although mostly asymptomatic, theirclinical manifestations may be sudden death, myocardial infarction or cerebral in-farction of cardiogenic embolic source. The embolic mechanism may be explainedby one of two mechanisms: dislodgment of tumoral fragments or embolization offibrin thrombi that arise on the papillary surface.Case report: A 68-year-old male, hypertensive, presented with sudden dysphagia,dysarthria, gait disequilibrium, vertigo, nausea and vomiting. On neurologic exami-nation he had rotatory nystagmus on lateral gaze; decreased right palatal excursion;broad based stance and gait with right side deviation. Cranial computed tomography(CT) revealed recent infarct on the right cerebellar hemisphere. The diagnosticstudies were normal except the transthoracic echocardiogram that disclosed anectopic, mobile mass with the stalk attached deeply in the left ventricular wall.The patient started anticoagulation. A month later he was submitted to surgicalresection of the tumour. The histology revealed a papillary fibroelastoma. One yearlater he was symptoms free, had a normal neurologic examination. There was noevidence of recurrence of the tumour on control echocardiography.Discussion: We report a case of posterior circulation stroke of cardiogenic embolicorigin as a first manifestation of very rare and treatable intra-ventricular papillaryfibroelastoma.


ISOLATED INFERIOR SAGITTAL SINUS THROMBOSIS (ISST)N. Tran, J. SilvaHamilton General Hospital, McMaster University, Hamilton, ON, Canada

Background and purpose: The inferior sagittal sinus (ISS) is the rarest affectedarea of cerebral vein thrombosis. Only one previous case of isolated inferior sagittalsinus thrombosis (ISST) has been reported. We describe a new case of isolatedISST.Methods: A 70 year-old man presents with a four day history of decreased appetite,global weakness followed by decreased level of consciousness. He had a witnessedright sided focal motor seizure with secondary generalization. He was found withright sided weakness, right facial droop, right sided hyperreflexia, drowsiness,disorientation to time and was febrile (39.9 C). He also had features of frontallobe dysfunction including apathy, decreased insight and volitional activity. Hiscondition did not change despite empiric treatment with acyclovir, cefotaxime,ciprofloxacin and levofloxacin.Results: Initial CT-head showed frontal paramedian hypodensities. CSF demon-strated elevated protein and pleocytosis, primarily lymphocytes. Auto-immune andcoagulation work-up were negative. Peripheral blood cell count showed leuko-cytosis, mostly neutrophils, which improved spontaneously. No obvious systemicmalignancy was detected by imaging. MRI Head demonstrated high signal inthe superior and medial areas of the frontal lobes in the distribution of the ISS.MR-Venogram showed attenuation of the ISS only. After anticoagulation treatmentthe patient improved and managed to go home with some residual frontal lobesdysfunction. Follow-up MRI showed improved ischemic area and incompleterecanalization of the ISS.Conclusion: Primary isolated ISST can present as a febrile non-infectious en-cephalopathy. In our case, we were unable to demonstrate any particular etiology.


STROKE AND EPIGASTRIC PAIN: CONSIDER AORTIC DISSECTIONS. Mavinamane, H.G. M Shetty, M. Robinson, K.R. DavisUniversity Hospital Wales, Cardiff, United Kingdom

Aortic dissection may rarely present with stroke and the diagnosis can be difficultwith atypical symptoms. Thrombolysing such stroke patients can be disastrous. Wereport a patient with Aortic dissection who presented initially with an epigastricpain and stroke.A 73 year old hypertensive man, presented with a left hemiparesis and dysarthriafor 3 hours. He had epigastric pain for 2 weeks, which worsened significantly on theday of admission. Examination revealed epigastric tenderness, dysarthria, and lefthemiparesis. The hemiparesis and dysarthria resolved completely by the following

day. Investigations revealed anaemia (Hb-9.7) and deranged liver function tests(Bilirubin-29, AST-99, Alkaline phosphatase-388). Initial chest radiograph, abdom-inal ultrasound and CT brain were normal. The epigastrc pain persisted and hedeveloped headache, sweating and nausea, whilst his liver function tests graduallydeteriorated. On the tenth day as an inpatient, he suffered a brief episode ofsyncope. A repeat ultrasound showed mild intrahepatic duct dilatation. A magneticresonance cholangiopancreatogram was performed which suggested a dissection ofthe abdominal aorta. Subsequent CT angiogram confirmed type A dissection ofaorta extending from the aortic root to both common carotid arteries and down tothe right common iliac artery. He was commenced on beta blockers and surgicalrepair was undertaken.Aortic dissection can present deceptively and delay in diagnosis can be catastrophicespecially if thrombolytic treatment is given for treatment of the associated stroke.Such an incident has been previously reported. Our case illustrates the need for ahigh degree of suspicion in all stroke patients.

Recovery and rehabilitation

1 Recovery and rehabilitation

OPTIMISING REHABILITATION OUTCOMES FOR APHASIA FOLLOWINGSTROKE THROUGH NEW LEARNINGH. McGraneQueen Margaret University College, Speech and Language Sciences,Edinburgh, United Kingdom

Many people with aphasia retain residual language impairments to varying degreesof severity following rehabilitation. Currently there is no theory of rehabilitationthat explains the therapeutic process involved in the restoration of a damagedlanguage system. Therefore it is not possible to discern what approaches/taskswould be most successful at restoring particular language functions. Does rehabili-tation facilitate the accessing of the damaged language system or could it involvenew learning resulting in the creation of new language representations? The mainobjective of this study was to investigate whether adults with aphasia could learnnew vocabulary. The methodology incorporated procedures based on evidence fromthe literature in order to facilitate and promote optimum learning. The novel stimuli(20 new words) were taught to 12 adults (<65 years) who presented with varyingdegrees of severity of aphasia. The training procedure incorporated learning theoryand a cognitive neuropsychological model of language. The immediate and delayedrecall of this vocabulary was investigated using a range of assessments to facilitatethe capture of new learning which was measured not only in terms of the accurateproduction of the stimuli but also the recognition and knowledge of the wordforms and meanings. Overall findings of this investigation with the presentationof select case studies demonstrate the ability of people to learn new languagerepresentations despite severe language impairment. The findings, which stronglysuggest that language rehabilitation could incorporate the process of new learning,have significant clinical relevance in terms of developing a theory of rehabilitationand to the procedures employed in speech and language therapy.


EFFECT OF RHYTHMIC AUDITORY CUES ON GAIT OF STROKE PATIENTSS.I. LinNational Cheng Kung University, Tainan, Taiwan

Introduction: Sensory regulation is a feature of bipedal locomotion control. Forpatients with chronic sensory loss, it is not clear if the way sensory inputs areused for locomotion control would be altered. This study examined if the use ofrhythmic auditory cue (RAC) for locomotion control was affected by the residualproprioceptive function in patients with chronic stroke.Methods: Fourteen chronic stroke patients went through a joint repositioning testof the knee and ankle, and were classified into intact and impaired joint positionsense (JPS) groups. EMG activity and peak joint angular acceleration of theaffected leg during two walking conditions, normal and with RAC, were recordedand compared. In RC, subjects were asked to match their foot-floor-contact with abeeping tone delivered by a metronome.Results: Walking speeds did not differ significantly between the two conditionsor the two patient groups. Compared to normal condition, RAC had shorter EMGactivation time in patients with impaired JPS, but not in patients with intact JPS.Stride characteristics did not differ significantly between walking conditions or thetwo patient groups.Discussion: Chronic stroke patients were found to use RAC to regulate locomotion.


However, the effect of RAC was not equal between patients with intact andimpaired lower extremity JPS: walking with cues lead to shorter EMG activationtime in patients with impaired JPS, but not in patients with intact JPS, compared towalking without cues. These findings imply that after chronic sensory loss, changesin sensorimotor processing for locomotion might have occurred and affected theuse of augmented sensory cues. Thus clinically, instead of using augmented sensoryinputs routinely, their use should be planned and effects carefully monitored.


THE EFFECTS OF TYPE AND INTENSITY OF PHYSIOTHERAPY ONLOWER LIMB STRENGTH AND FUNCTION AFTER STROKEE.V. Cooke, R.C. Tallis, S. Miller, V.M. PomeroySt. George’s University of London, London, United Kingdom

Background: Stroke survivors often have permanent residual motor impairment.This may be due to a sub-optimal dose of conventional physical therapy (CPT) andthe discouragement of strength training. However, experimental evidence suggeststhat strength training might be beneficial. Hypothesis: adding functional strengthtraining (FST) to CPT improves muscle function, gait and functional mobility morethan either CPT alone or CPT plus “neuro-facilitation” (NF).Methods: Multi-centred randomised controlled observer-blind trial. Subjects werewithin 3 months of stroke with the ability to voluntarily move their paretic lowerlimb. A power calculation estimated sample size as 102. Subjects underwentbaseline measurements before being allocated randomly to; CPT; or CPT + NF;or CPT+FST for 6-weeks. All additional therapy was provided up to1-hour/day,4 times/week. Outcome measures were made at 6 weeks after baseline. Measure-ment battery included: muscle strength; walking speed; and functional mobility(Rivermead). Analysis followed the intention to treat principle. Data, outcomeminus baseline, was tested for differences between groups using the Kruskal-Wallistest. Results: 109 subjects were recruited. Mean age was 68.3 (SD12.03) years.The attrition rate was 8.3%. Only Rivermead data is reported here. Median (IQR)change in Rivermead score following intervention was 5.0 (9.7) for control, 6.5(14.3) for NF and 7.0 (13.7) for FST. The Kruskal-Wallis statistic was 1.06(p = 0.59). Discussion: Immediately after intervention no statistically significantdifferences were found between groups for functional mobility, however there wasa trend towards CPT + FST. Muscle strength (for which the trial was powered) andgait data are currently undergoing analysis.


MOTOR NETWORK CHANGES AND FUNCTIONAL RECOVERY IN STROKEPATIENTS TREATED WITH VERY EARLY MOBILISATION IN AN ACUTESTROKE UNIT. A LONGITUDINAL FUNCTIONAL MRI STUDYT. Askim, B. Indredavik, S. Mørkved, O. Haraldseth, A. HåbergNorwegian University of Science and Technology, Trondheim UniversityHospital, Trondheim, Norway

Background: Functional MRI (fMRI) might elucidate mechanisms of brain plas-ticity. The aim of this study was to investigate the relationship between functionalrecovery and brain activation patterns after an acute stroke.Methods: 14 patients (62-75 years) with first ever ischemic stroke and unilateralhand paresis, but intact language were included. 16 age and gender matchedcontrols were also investigated. All patients were treated in an acute stroke unitwith very early mobilisation and early supported discharge. They underwent MRI,fMRI and functional tests 4-8 days from onset and after three months. fMRIparadigms were 1 Hz and self-paced (SP) index finger tapping.Results: No patients had infarction involving primary motor cortex (M1). Therewas significantly improved hand function as measured by all functional tests. 1 Hztask: Patients in the acute phase activated more prefrontal regions than the controls.There was increased activation in contralateral thalamus, anterior cingulate cortexand ipsilateral prefrontal cortex for patients in the chronic compared to the acutephase. SP task: Controls had significantly larger activation in contralateral M1and ipsilateral cerebellum than patients in the acute phase. In the chronic phasepatients had increased bilateral M1 activity compared to the controls. There wasincreased activity in contralateral M1 for patients in the chronic compared to theacute phase. Discussion: For the 1 Hz task the difference in activation between thechronic and the acute phases did not involve increased activation in motor areas,but encompassed other cortical regions. This was at great variance to the resultsfrom the SP task. These findings indicate that the injured brain adapts to differentmotor task demands using different networks.


FACTORS PREDICTING EARLY HOSPITAL DISCHARGE FOLLOWINGADMISSION FOR ACUTE STROKEJ. White, L. Dacey, R. Navaratnasingam, M. WaniMorriston Hospital, Swansea, Cardiff, United Kingdom

Background: Reducing hospital length of stay following admission for acute strokehas economic benefits for the healthcare provider. However, identifying patientssuitable for a reduced length of stay with early supported discharge is difficult atthe time of admission. This study explored the relationship between clinical andsocial characteristics at the time of admission, with length of hospital stay for acutestroke inpatients to predict a safe early discharge.Methods: A retrospective case control study of all patients admitted with acutestroke to a UK hospital, over a 3 year period. Patients were categorised into twogroups, those requiring a 10 day or less hospital admission and those requiringlonger. 23 separate clinical and social characteristics were assessed to evaluate theirimpact on length of hospital admission.Results: 359 patients, mean age 76.2 years, mean length of stay 54 days (median24 days), mean Barthel Score on admission 10.8.Ten factors, on admission, were independently associated with a reduced lengthof stay: CT brain scan showing no sign of haemorrhage, (odds ratio (OR) 11.07(95% confidence interval (CI) 1.87 to 65.11)); no receptive dysphasia (OR 7.59 (CI1.98 to 29.03)); no sensory deficit on admission (OR infinite (CI 5.49 to infinite));no hemi neglect (OR infinite (CI 5.72 to infinite)); sinus rhythm (OR 2.72 (CI1.35 to 5.49)); living with support prior to admission (OR 2.52 (CI 1.49 to 4.26));Functional Ambulation Categories (FAC) score ≥4 (OR 10.68 (CI 6.23 to 18.31));Barthel Score ≥19 (OR 12.32 (CI 6.63 to 22.88)); and urinary continence (OR11.98 (CI 5.66 to 25.29)).Discussion: This study has identified 10 clinical and social factors present atadmission that may help identify patients who could be potentially discharged earlyfrom hospital, allowing prompt referral to early supported discharge teams andother intermediate care services.


DIFFERENCES IN VISUAL ATTENTION BETWEEN HEMIPLEGIC SIDES INPOSTSTROKE PATIENTSS. Shimizu, M. Maeda, Y. Ikeda, H. NagasawaFaculty of Rehabilitation, School of Allied Health Sciences, KitasatoUniversity, Kanagawa, Japan

Background and purpose: We previously reported that visual attention in thecircumferential field in healthy persons was greater in the left lower field than inthe right upper field. The present study investigated differences in visual attentionbetween left and right hemiplegic sides based on simple reaction times (RTs) tovisual stimuli.Methods: Participants were 10 stroke patients with right hemiplegia (RH group),10 stroke patients with left hemiplegia (LH group), and 20 normal control subjects.RTs were recorded using RT estimation software on a personal computer. Fixationpoint and reaction stimuli were presented on a screen. Stimuli were presented atone of 16 sites located on circles with radii of 1 cm or 11 cm; visual angles (VA)were 2 degrees (VA2) or 20 degrees (VA20). Stimuli were randomly presented fivetimes at each site for a total of 80 trials, and the delay from presentation to thesubject pressing a key was recorded as RT. Stroke patients used their unaffectedhand to press the key, while half of the control subjects used their right hand (CRgroup) and the other half used their left hand (CL group).Results and discussion: In the comparison of the LH and CR groups, RTs for theLH group were slower; however, no significant differences were observed betweenstimuli positions. In contrast, the comparison of the RH and CL groups showed nosignificant difference in RTs. For the RH and CL groups, RTs to left lower fieldstimuli were significantly shorter than RTs to upper field stimuli. These resultsindicate that patients with left hemiplegia may have decreased attention in all visualfields compared to patients with right hemiplegia.


DEPRESSION IN CAREGIVERS OF LONG-TERM STROKE SURVIVORSD. Varga, E. Boros, J. Kenez, Z. NagyNational Institute for Neurology and Psychiatry, Budapest, Hungary

Objective: Caregivers of stroke patients’ may experience high levels of burden,that can result in deterioration of the caregivers’ mental and/or physical health. Ouraim was to examine the prevalance of depression among caregivers.

Poster SessionRecovery and rehabilitation


Methods: A sample of 87 stroke survivors and their informal caregivers wasstudied. Caregiver burden was evaluated with Beck Depression Inventory andCaregiver Strain Index. Patients’ functional, cognitive and behavioural status wasalso assessed with a questionnaire yielding information pertinent to these items.Results: 69% of caregivers of long-term stroke survivors suffered some kindof depression. In 50% the patients’ serious residual functional status, in 77%mental-behaviourial symptoms were mentioned as cause of depression. Presenceof both caused depression in 87% of caregivers. Patients generate high caregiverburden are significantly older and had significantly higher volume of cerebralinfarction.Conclusion: High percentage of caregivers suffer from depression. The level ofself-percepted burden was stronger associated to the patients’ mental-behaviourialsymptoms, than to the degree of their disability.


A PILOT STUDY INVESTIGATING THE COMPARISON BETWEENFAMILIAR AND UNFAMILIAR ENVIRONMENT ON PATIENT’S ABILITIESTO COMPLETE AND PROCESS ACTIVITIES OF DAILY LIVING, POSTBRAIN INJURYJ. ScottHammersmith Hospitals NHS Trust, London, United Kingdom

Background: Cognitive impairment is common post-brain injury and is associatedwith poor long-term outcomes. Environmental context is thought to impact oncognitive processing, and it is possible that the venue where cognitive abilities areassessed will influence the patients’ performance.The Assessment of Motor and Process Skills (AMPS) is an observational tool usedto measure the quality of a person’s ability to perform activities of daily living,based on normative data for age and injury.Hypothesis: Patients’ will perform better in terms of cognitive ability, at homecompared to in hospital.Objective: To compare the process abilities of patients with brain injury in hospitaland at home using the AMPS, with a same subject design.Methods: Patients admitted with brain injury, who had cognitive impairment butwere orientated, were assessed in the hospital and at home within 48 hours usingAMPS. The scores for process skill (the skills needed to organize and adapt actionsto complete tasks) were compared at the end of the study.Results: 7 out of the 8 patients assessed performed better within the hospitalcontext.

Median AMPS process score Interquartile range P

Home 0.16 0.2-0.54 0.05Hospital 0.94 0.47-1.29

Discussion/Conclusion: The structure and lack of external stimuli may accountfor the improved performance within the hospital context. Patients with cognitiveprocessing deficits performed better within the hospital setting. This implies thatthe level of challenge therapists use to assess patients’ cognition through functionmay need to be higher in hospital to ensure sufficient process ability at home.


THE EFFECTS OF CURRENT PHYSICAL THERAPY AND FUNCTIONALSTRENGTH TRAINING ON UPPER LIMB FUNCTION ANDNEUROMUSCULAR WEAKNESS AFTER STROKE: A PILOT STUDYC. Donaldson, R. Tallis, V. PomeroySt George’s, University of London, London, United Kingdom

Background: Scientific evidence suggests that functional strength training (FST)might have a better effect than current physical therapy (CPT) on upper limb motorrecovery after stroke. Aim: to assess the feasibility and sample size for a clinicaltrial to compare CPT and FST.Methods: A randomized, single-blinded, clinical trial. Subjects were within 3months of infarction (anterior circulation) with upper limb weakness. Subjects wererandomised into 3 groups. All received routine CPT. Control Group 1 received noextra physical therapy (CPT), Group 2 received increased intensity (CPT+CPT),Group 3 received CPT and FST. Intervention lasted 6 weeks. A treatment recordingform, developed by our team and validated in clinical practice, was used to stan-dardise and record CPT. Measurements included: muscle strength, Action ResearchArm Test (ARAT) and 9-Hole Peg Test. Measurements were taken before treatmentbegan, after six weeks of intervention, and 12 weeks thereafter. Analysis: Attrition

rate was calculated and differences between groups were analysed with descriptivestatistics. ARAT data was used to calculate the sample size for a Phase III trial.Results: 30 subjects were recruited with an attrition rate=0%. No statisticallysignificant differences were found between groups, however there was a trendfor CPT+FST to produce greater improvement at a clinically relevant level: meanARAT change of +13 (CPT) and +22 (CPT+FST). The power calculation estimateda sample size of 246 for a Phase III trial.Discussion: Results of this pilot study suggest that increased intensity of CPT maynot further enhance motor recovery after stroke but adding FST to CPT might.A Phase III trial is feasible using the methods of this study although a largemulticentre trial will be required.


MODIFIED OBSERVATIONAL PERSPECTIVE AND STROKEREHABILITATIONL.M. Ewan, T. Haire, K. Kinmond, H. Chatterton, N. Smith, P. HolmesManchester Metropolitan University, Alsager, United Kingdom

Background and purpose: Motor imagery has been questioned as a neuropsy-chological rehabilitation technique for stroke patients with structured observationbeing proposed as a more valid approach (Holmes, 2006). Similar to imageryhowever, observation conditions provide two spatial visual perspectives: first andthird person. Since there is evidence that the different perspectives are linked todifferent brain activity, the use of each perspective may be moderated post-stroke.The practical considerations of this change have not been examined experimentally.This research explored observational visual perspective in individuals who hadexperienced stroke to aid the development of observation-based rehabilitationprogrammes.Method: 21 individuals who had experienced stroke were matched against 19individuals who had not had a stroke. Following ethical approval and full writteninformed consent, a stroke and observation specific questionnaire was employed toexplore viewing experiences; specifically kinesthisis. Participants viewed DVDs ofactivities of daily living from both visual perspectives.Results: Non-parametric analysis indicated that individuals who had not had astroke showed a statistically greater preference for a first person visual perspectiveand reported stronger kinesthisis with this perspective. In contrast, individualswho had experienced stroke showed no preference for either perspective and weregenerally unable to report kinesthisis with either the first or third person perspective.Discussion: Stroke may lead to a change in preferred visual perspective and experi-ence of observational kinesthisis as a consequence the lesion damage and functionalmotor inactivity. Observation-based rehabilitation interventions may support neuralchange. However, further research is required to investigate these changes linkedto individual differences in stroke aetiology. This study provides evidence thatvisual perspective should be considered in all imagery and observation-based strokerehabilitation interventions.


FUNCTIONAL RECOVERY IN ISCHAEMIC STROKE PATIENTS YOUNGEROR OLDER THAN 65 YEARSJ. Jansa, Z. Sicherl, K. Angleitner, A. GradUniversity Medical centre Ljubljana,Ljubljana, Slovenia

Background: To compare duration of occupational therapy (OT), functional out-come and quality of life in group of ischaemic stroke (IS) patients younger or olderthan 65 years.Methods: IS patients who were consecutively included into OT within early hospi-tal stay and contacted three months after stroke; 24R-hemiplegia, 26L-hemiplegia;31 males, 19 females. Assessment tools: Extended Barthel index (EBI), CanadianOccupational Performance Measure (COPM) and EuroQol (EQ). They were doneduring OT and at follow up.Results: There were 25 patients younger than 65 years (mean 55 yrs, range 39-65)and 25 were older than 65 yrs (mean 73 yrs, range 67-81). Mean duration of OTin younger group was 15 days (range 4-54); mean duration of OT in older groupwas 8 days (range 2-28). This difference was statistically important (p<0,007). TheEBI and EQ were assessed in the whole sample; we were able to follow client’spriorities with COPM in 32 patients; 15 in younger group and 17 in older group.Difference in EBI was statistically important from the first assessment to follow-upin both groups (p<0,0001;p<0,0001);there were no statistically important differ-ences among groups in initial or follow-up assessment (p=0.28; p=0.67). Differencein EQ was statistically important in both groups (p<0,01;0,002);there were nostatistically important difference among groups (p=0,63;p=0,45). Regarding the useof COPM; 32 patients (64%) of our sample expressed in total 87 activities. There


were 15 patients from younger group and 17 patients from older group. 14 patientsfrom each group named 58 activities (68%) of basic activities of daily living; 6younger, 4 older expressed in total 14 (18%) issues of productivity and 5 from eachgroup named in total 14 (14%) leisure activities.Discussion: Results showed that both groups need less help as measured by EBI.Equally, the quality of life, as measured with EQ has improved. We were ableto follow client functional priorities in both groups, although the younger groupreceived more OT. Further work is needed to clarify this issue.


A SERIES OF INVESTIGATIONS INTO SENSORY REHABILITATION POSTSTROKES.L HillierUniversity of South Australia, Adelaide, Australia

Background: Sensory loss post-stroke is well documented, however the rolesensation plays in the recovery of motor function is less understood. We haveconducted a series of clinical trials investigating different applications of sensoryinput as part of rehabilitation post-stroke.Methods: Firstly we have applied combinations of peripheral and central stimula-tion to either the foot or hand in chronic and acute stroke rehabilitation participants.The rationale is that such stimulation will excite the cortex and facilitate positiveneuroplastic responses to subsequent task specific training. The second series oftrials involved training awareness and appreciation of sensory input to the lowerlimb post stroke. The rationale for these experiments was based more on a learningor attention-training paradigm.Results: Results from these trials offer some functional evidence that afferent inputis effective in driving recovery, however data from TMS investigations have beenless convincing. We found sensory appreciation could be improved via specifictraining in chronic stroke participants but not as convincingly in acute. Therewas also some preliminary indications that the increased sensory awareness led toimproved postural control in some participants.Discussion: Our findings warrant further investigation and current studies areexploring the relationship between sensory input in the lower limbs and functionalactivity.


IS THE BIO-PSYCHOSOCIAL MODEL SUITABLE TO EXPLAIN THEDEVELOPMENT OF DEPRESSION AFTER STROKE?T.A. Barskova, G. WilzTechnical University Berlin, Department of Clinical and Health Psychology,Berlin, Germany

Background: Depression after stroke is common and increases morbidity andmortality in the first years after stroke onset. Nevertheless little is known about therole of psychosocial factors on the etiology of depression in early as well as in thelate poststroke stages. With reference to the bio-psychosocial model, the aim ofour study was to investigate the influence of stroke survivors’ mental impairmentas well as the quality of their social relationships on the development of poststrokedepression.Method: The study used a longitudinal design. Eighty-one German stroke patientswere investigated twice, directly after discharge (on the average three month afterstroke onset) and one year later. Hierarchical regression analyses and cross-laggedpartial correlation analyses tested direct and indirect mediating effects of potentialpredictors on poststroke depressive symptoms.Results: Time 1 patients’ perceived cognitive and emotional functioning predictedpsychological depressive symptoms at time two. Quality of patients’ social rela-tionships mediated the effect of the stroke-related emotional deficits on depression.Discussion: In contrast to the previous research the study provided more evidencefor causal influence of different risk factors on PSD. Results support the biopsy-chosocial model of poststroke depression. Early and late poststroke depressionseem are based on partially different etiological mechanisms.

Management and economics

1 Management and economics

COMPARISON OF BLOOD PRESSURE MANAGEMENT AFTER STROKEAND CORONARY EVENT. THE REDUCTION OF ATHEROTHROMBOSISFOR CONTINUED HEALTH (REACH) REGISTRYE. Touzé, J. Röther, D. Batt, F. Aichner, M. Alberts, M. Ohman, P. Durieux,J. Coste, S. Goto, G. StegHôpital Sainte-Anne, Paris, France

Background: Management of blood pressure (BP) is not optimal in patients withatherothrombotic diseases. We looked for differences in BP control and the use ofantihypertensive drugs in patients with cerebrovascular disease (CVD, includingstroke or TIA) and coronary artery disease (CAD).Methods: 68,236 patients were enrolled in the REACH Registry, an international(44 countries worldwide) prospective, observational study of patients with or ≥3risk factors for atherothrombotic disease. Of these patients 12,153 had isolatedCVD and 33,611 had isolated CAD. At inclusion BP was measured and treatmentdata were collected.Results: There were no major differences in age, previous hypertension and otherrisk factors between the CVD and CAD patient groups. However, CAD patientsexperienced lower mean BP values, were more likely to have a BP<140/90 mmHg,and to receive 3 or more antihypertensive drugs (34.2% vs. 22.7%, p<0.0001).The use of 3 or more antihypertensive drugs was also more common in CADpatients with elevated BP (≥140/90 mmHg) (39.4% vs. 26.2%, p<0.0001). Afteradjustment for age, sex, other risk factors, and world regions, the CAD groupmaintained significantly better control of BP (OR=1.4; 95%CI: 1.3-1.5, p<0.0001)and the use of ≥3 drugs (OR=1.4; 95%CI: 1.3-1.5, p<0.0001). Similar trend wasobserved across world regions.

CVD only (12,153) CAD only (33,611)

Mean age (SD) 68.9 (10.2) 67.8 (10.2)Male, % 56.4 70.8Previous hypertension, % 80.7 78.7Mean systolic BP (SD) 140.2 (19.6) 134.5 (18.7)

All P values < 0.0001.

Conclusion: Blood pressure management is better after a coronary event thanafter a cerebrovascular event. This result could be explained by the under use ofcombination therapies in stroke patients compared to CAD patients.


US SURVEY OF STROKE NEUROLOGISTS ANDNEUROINTERVENTIONALISTS ON TREATMENT CHOICES FORINTRACRANIAL STENOSIST. Turan, M. Lynn, M. ChimowitzEmory University School of Medicine, Atlanta, GA, USA

Background: We sought to determine the effect of an NIH-sponsored clinical trialon treatment choices of physicians managing patients with intracranial stenosis.Methods: Surveys of treatment choices were sent pre- and 1 year post-publicationof the Warfarin vs. Aspirin for Symptomatic Intracranial Disease (WASID) Trialresults. The pre-WASID survey was sent to neurologists and the post-WASID sur-vey was sent to neurologists and neurointerventionalists. The post-WASID surveyincluded questions about the minimum benefit that physicians would require tomake stenting their treatment of choice. Data was analyzed using the chi-square test.Results: There was a significant difference in the choice of antithrombotic agentsfor the treatment of both anterior circulation (p<0.001) and posterior circulation(p<0.001) stenoses after publication of WASID (see Table).There was no significant difference in risk reductions required by neurologistsvs. neurointerventionalists for stenting to become their treatment of choice. For

Table 1. Antithrombotic choices before and after WASID

MCA or IC siphon % of MD % of MD Basilar or vertebral % of MD % of MDpre-WASID post-WASID pre-WASID post-WASID

n=181 n=199 n=181 n=199

Warfarin 41 7 Warfarin 49 15Antiplatelet 44 85 Antiplatelet 36 74Combination 11 4 Combination 10 7Other 4 4 Other 4 4

Poster SessionManagement and economics


a primary endpoint rate of 20%/2 yrs in the medical arm (the rate in WASIDfor high-risk patients with 70-99% stenosis), 33% of physicians required a 25%reduction from stenting, 15% required a 33% reduction, 14% required a 40%reduction, 21% required a 50% reduction, 13% required a 60% reduction, and 4%would continue to use medical therapy regardless of the stenting rate.Discussion: The results of WASID had a significant impact on physician treatmentpractices. For high-risk patients with intracranial stenosis, 40% is the minimumrelative risk reduction required from intracranial stenting to make it the treatmentof choice for a clear majority (at least 60%) of physicians.


HOSPITAL DISCHARGE CODING UNDERESTIMATES BURDEN OF STROKEM. Taylor, C. McAlpine, M. WaltersStobhill Hospital, Glasgow and Western Infirmary, Glasgow, United Kingdom

Background: Hospital discharge coding is an important process which informshealth resource planning. During an audit project we noted a discrepancy betweenthe information services division (ISD) of the Scottish Executive numbers forstroke disease workload and the numbers recorded by the local stroke consultants.Methods: 280 stroke service discharges were reviewed from 2 different stroke unitsduring different time periods within the health board division. One doctor codedthem according to ICD10. The actual coding was then reviewed and compared tothis.Results: 219 out of the 280 (78%) stroke service discharges had a final diagnosisof stroke when coded by the doctor. Of those 219 only 166 (76%) of patients wererecorded as stroke by hospital coding. In particular, of those diagnosed as a lacunarstroke (43 patients) only 23 (53%) were coded as stroke on discharge. Of the 61diagnosed as “not stroke” by the doctor 19 had a “false positive” coding of strokeby coders on discharge.Discussion: There are several potential sources of error in the coding process. Thishas major implications for both national stroke statistics and local stroke serviceplanning and resources. Urgent further work is required to identify the extent towhich “false positive” stroke codes are generated and whether this pattern of “falsenegative” stroke coding is reproduced elsewhere.


FIRST IRISH NATIONAL AUDIT OF STROKE CAREF. Horgan, A. Hickey, S. Murphy, M. Wiley, R. Conroy, H. McGee, D. O’Neill,on behalf of the Irish National Audit of Stroke CareTrinity College Dublin, Dublin, Ireland

Improving services for people with stroke represents a global challenge, espe-cially in the light of the proven efficacy of many treatment modalities. The UKNational Sentinel Stroke Audit was a pioneer in developing a national profile ofhospital services for stroke. We report on the design of a project which buildson this methodology, but which also profiles nation-wide preventive, communityrehabilitation and long-term care services for people with stroke.The Irish Heart Foundation, in association with the Irish Department of Healthand Children, commissioned a national audit of stroke services in March 2006.The project involves hospital audits, and community-based surveys of generalpractitioners (GP), allied healthcare practitioners (AHPs), patients and carers, andnursing homes.All 37 public hospitals (100%) providing acute services to stroke patients areparticipating in the.Organisational and Clinical Audit Proformas of the UK Na-tional Sentinel Stroke Audit 2004. A random sample of 484 GPs were surveyedby post and 46% responded. The AHP survey, involved interviews with regional,disciplinary and nursing managers, and frontline staff, with 85% response. 200people with stroke and a family member will be interviewed one year afterdischarge using a questionnaire on health status and service needs and utilization,as well as 200 people with stroke who have been discharged to nursing homes, anda family member. Final results are due in September 2007.This audit is of interest for two main reasons: it provides evidence of the feasibilityof using the UK National Sentinel Audit in another jurisdiction, but also providesa methodology which allows for the measurement of availability of the full rangeof services for people with stroke across the modalities of primary and secondaryprevention, acute treatment, rehabilitation and long term care. This global overviewis vital to the delivery of services across the full spectrum of stroke prevention andcare.


PRIMARY CARE MEDICINE AND STROKE: THE IRISH NATIONAL AUDITOF STROKE CARED. O’Neill, D. Whitford, F. Horgan, M. Wiley, R. Conroy, S. Murphy,H. McGee, D. O’Neill, on behalf of the Irish National Audit of Stroke CareTrinity College Dublin, Dublin, Ireland

Primary care services have a key role in the prevention and management of stroke.As part of the Irish National Audit of Stroke Care, a survey of general practitioners(GPs) was performed to document the availability of evidence-based structures forsupporting stroke care and prevention in general practice and to profile the views,experiences, and needs of Irish GPs in this context. In a cross-sectional study ofrandomly selected GPs practising in the Republic of Ireland was surveyed by postalsurvey.Of the target sample of 484 GPs, 36 were ineligible and 204 responded (responserate = 46%). Regarding the acute management of stroke, nearly a fifth of GPs (17%)reported initially managing at least a substantial minority (20%) of their patientsat home. The majority of GPs viewed existing rehabilitation services for theirstroke population as inadequate. Overall, general practice showed little structuredorganisation for long-term follow-up of stroke patients.There was little or no organised system of care for the primary prevention of strokewithin primary care in Ireland. Three quarters of GPs believed there were barriersto implementing secondary prevention strategies in their practiceJust over 60%reported time as a barrier, 57% of reported staffing issues and almost a third (33%)funding as barriers. Other barriers listed included lack of protocols/guidelines(17%) and lack of space (almost 10%). The main barriers listed for secondaryprevention were very similar to those recorded for primary prevention.There was little or no organised system of care for the prevention and managementof stroke within primary care in Ireland. However, there were encouraging signs ofdevelopment. GPs in practices involved in a national cardiac prevention programmeand those with good or excellent access to practice nurses were more likely toengage in evidence-based activities to manage stroke.


EDUCATIONAL MULTIMEDIA CAMPAIGNS HAVE DIFFERENTIALEFFECTS ON PUBLIC STROKE KNOWLEDGE AND AWARENESS OFINDIVIDUAL STROKE RISKJ.J, Marx, M. Nedelmann, B. Haertle, M. Dieterich, B.M. EickeJohannes Gutenberg-University Mainz, Mainz, Germany

Background: Aim of the study was to evaluate the educational effects of differentmedia in a multimodal educational program on public knowledge of risk factorsand warning signs of stroke.Methods: Computer-assisted telephone surveys were conducted among an averagesample of 500 members of the general public, before and immediately after anintense three months educational campaign in a German area of 400.000 inhabi-tants. The multimodal educational program comprised of 400 poster advertisementson billboards, busses, local emergency transport cars etc. Print media includedflyers dispensed in pharmacies and at the doctors’ office and mail circular to allhouseholds. Slogans, stroke interest stories and interviews appeared regularly inlocal newspapers, on television and radio, and several public events focussed on thesubject.Results: Before the educational intervention stroke knowledge was generally low,especially in men and elderly individuals. General knowledge of the nature ofstroke (65.7% correct answers before versus 84.9% after the campaign, p<0.01)and the awareness of being at risk of stroke (32.7% vs. 41.9%, p<0.01) significantlyincreased due to the campaign, especially in respondents of lower educationalBackground: In contrast, there was hardly any effect on detailed knowledge ofspecific stroke warning signs or different risk factors. Mass media were most fre-quently reported as the main information source (66.5%). Information flyers werealso remembered by a high proportion of respondents (59.0%), while widespreadposter advertisements received far less attention (26.7%).Discussion: Our data indicate that educational programs are effective in increasinggeneral knowledge of stroke in the public. They improve awareness of individualstroke risk and this may influence behavior in acute stroke. Especially in indi-viduals of lower educational background repeated information using short-tailoredmessages presented in mass media proved to be effective. It is difficult, however, totransfer detailed information by means of a large educational campaign.



THROMBOLYTIC THERAPY IN ACUTE ISCHEMIC STROKE IN TAIWANM. Tseng, K. Chang, J. LiuNational Sun Yat-Sen University, Kaohsiung, Kaohsiung, Taiwan

Background and purpose: Starting Jan 1, 2004, the use of intravenous recom-binant tissue plasminogen activator (rtPA) in patients with acute stroke becomesavailable for reimbursement from the National Health Insurance (NHI) Programin Taiwan. The purpose of this study was to study the frequency of intravenousrtPA for stroke treatment in Taiwan the first year after the reimbursement, and toexamine the characteristics regarding the usages in the health care system.Methods: We studied the administrative claims data of NHI beneficiaries of 2004.The compulsory and universal NHI covers more than 96% of the total population ofTaiwan since the implementation in March 1995. We identified patients treated withrtPA by searching the database of Details of Inpatient Orders with the rtPA-specificorder code. The associated data of Inpatient Expenditures by Admissions wereexamined.Results: Among 90,550 admissions with cerebrovascular diseases in 2004, therewere 93 patients treated with rtPA. The mean age was 64.0 ± 11.0 years (range 35.5to 82.4). Fifty-nine or 63% of patients were men, 69% had a Charlson comorbidityindex of 0, 27% 1, and 4% 2. Neurologists were admitting physicians for 66.7% ofpatients. About one-forth of patients were treated in medical centers, 57% regionalhospitals, and 17% district hospitals. Median length-of-stay (LOS) was 10 days(range 1 -136), and in-hospital death were found in one case (1.1%). Patients caredfor by neurologists as compared to non-neurologists had significantly lower medianLOS (8.5 versus 19 days, P = 0.016). Patients admitted into medical centers hadlonger median LOS (16 days), as compared to regional hospitals (12) or districthospitals (6) (Kruskal-Wallis test, P = 0.002).Discussion: Intravenous rtPA was not widely applied within the first year followingreimbursement from NHI in Taiwan. Because the analysis was based on NHIclaims data, some important patient-level data, particularly initial stroke severityand functional outcomes, were not available.


COST OF STROKE FOR SOCIETY IN YOUNGER PERSONS RECEIVINGREHABILITATIONA.K.S. Sunnerhagen, A.A. BjörkdahlGöteborg University, Göteborg, Sweden

Background and purpose: In recent years a number of costs of stroke studies havebeen conducted based on incidence or prevalence estimating costs a given time.As there still is a need for a deeper understanding of factors influencing the costthe aim of this study was to calculate the direct and indirect costs for society ina “younger” (<65) sample of stroke patients and to explore factors affecting thecosts.Methods: 58 patients, included in a study home rehabilitation, were followed 1year and interviewed about use of health care services, assistance, medicine andassistive devices. Costs were calculated. A linear regression of cost and variablesof functioning, ability, community integration and health-related quality of life wasundertaken.Result: Inpatient care contributed substantially to the direct cost with a mean lengthof stay of 92 days. Rehabilitation during the first year constituted of in average 28days in day clinic, 38 physiotherapy sessions and 20 occupational therapy sessions.Total direct mean cost was 33 604 Euro and indirect cost 32 129 Euro. The directcosts were influenced by the process skill (the ability to plan and perform a giventask and to adapt when needed) and presence of aphasia. Indirect cost for informalcaregiving increased for patients less health-related quality of life scoring low onhome integration.Conclusion: Costs are high in this group compared to other studies partly due tolength of stay as well as loss of productivity.


DOES ORGANISATION OF STROKE CARE CORRELATE WITH RESEARCHACTIVITY IN STROKE?D.F. Jenkinson, G.A. Ford, A.R. Rudd, A. Hoffman, G.D. LoweUK Stroke Research Network, Leazes Building, Royal Victoria Infirmary,Newcastle upon Tyne, London, United Kingdom

Background: It is unclear whether participation of stroke services in research leadsto improved patient care, or conversely whether better organized stroke servicesfacilitate greater participation in research. We determined the association between

quality of organization of stroke care and research activity in stroke services inEngland.Methods: The 2006 National Sentinel Audit of Stroke assessed by questionnairethe quality of stroke service infrastructure and included two questions on researchactivity: (A) “How many stroke research studies are you involved in?”, and (B)“How many Whole Time Equivalent staff are employed in stroke research?”. Datawere collected by local staff from 235 sites in England in April/May 2006.Data analysis was performed using SPSS. As research activity formed part of thetotal score for Organisation of Care, we did not compare with the total score. Weexamined the association between responses to the 2 questions on research activitywith each of the other 9 domains of Organisation of Care.Results: Correlation was shown at the 0.01 level (2-tailed) between responses tothe 2 questions research and the following 6 domains: acute care organisation,organisation of care, interdisciplinary services (overall service), TIA/neurovascularservices, team working – agreed assessment measures and communication withpatients and carers. The strongest correlation was with acute care organisation(Spearman’s rho 0.381 for research question (A) and 0.387 for question (B)). Nosignificant correlation was found with the domains interdisciplinary services (strokeunit), team working (records) and team working – team meetings.Discussion:The significant correlation between quality of organization of stroke care andresearch activity suggests that well organized stroke care facilitates stroke patientparticipation in research and/or that participation of stroke services in researchfacilitates improved services.


COST EFFECTIVENESS OF STROKE UNIT (SU) CARE FOLLOWED BYEARLY SUPPORTED DISCHARGE (ESD)Ö. Saka, V. Serra, Y. Samyshkin, S. Merkur, A.J. Mcguire, C. WolfeKings College, London, Division of Health and Social Care Research, London,United Kingdom

Introduction: Stroke is the second leading cause of death in England and Walesand the leading cause of adult disability. Annual cost of stroke care in the UKgovernment is over £7bn including. SUs provide improved outcomes for strokepatients with respect to non-stroke specialised hospital units. In addition to thatanother trend has been ESD of some stroke patients. This allows inpatient beds tobe available for the care of stroke patients faster, decreases the necessary numberof expensive hospital beds to be maintained whilst providing further rehabilitationcare to stroke patients at home. We assessed the cost effectiveness (CE) of SU carefollowed up by ESD (SUESD).Methods: Data from the South London Stroke Register, and local ESD wereutilised for clinical and resource use data. The cost effectiveness of SUESD wascompared with with SU without ESD (SUNESD) and general medical ward carewithout ESD (GWNESD). We used a Markov model to simulate the care processfor 10 years. Societal perspective was used for costing and included direct carecosts as well as informal care costs and productivity losses due to mortality andmorbidity.Results: SUESD option leads to better outcomes, although it increases totalcare costs. We found that it costs £9,200 per additional quality adjusted lifeyears (QALY) when SUESD was compared with GWNESD and £8,600 whencompared with SUNESD. The incremental cost effectiveness ratio (ICER stayedwithin accepted limits of £30,000 per QALY gained. The multi-way (+10%) andprobabilistic sensitivity analyses did not have a significant impact on the results.Discussion: This is the only study to date looking at the CE of SU followed by ESD,comparing it with other treatment options. The results of the study suggests thattreatment in stroke unit followed by early discharge of patients with an enhancedoutpatient care policy (SUESD) offers the best results in terms of effectiveness,with an additional cost within accepted reasonable CE levels. GWNESD, althoughcheaper than the other two, appeared the least effective strategy.


STROKE OCCURRENCE AND DISEASE CLASSIFICATION IN GERMANY- ANATIONWIDE ANALYSIS BASED ON THE GERMAN DRG REPORT 2004P.L. Kolominsky-Rabas, B. Griewing, J. RüthemannDep. of Health Economics, Institute for Quality and Efficiency in Health Care(IQWiG), Cologne, Germany

Background: Data are limited regarding the number of stroke patients and theirfraction in each ICD-10 (International classification of diseases) and DRG (Di-agnosis Related Groups) category in German hospitals since so far data are onlyavailable from cohort studies and stroke registers with a limited number of patients.

Poster SessionManagement and economics


Methods: Based on the analysis of more than 1.780 German hospitals transferringtheir DRG (B70 A- D) and ICD-10 data to InEK (Institut für Entgeltsystem imKrankenhaus) we compiled a detailed overview of stroke occurrence and diseaseclassification of all German inpatient strokes in 2004.Results: 235.097 stroke patients were reported for 2004. 5.658 (2%) thereof diedwithin 3 days after admission (B70 C), 189.403 (81%) had a non-hemorrhagicstroke (DRG B70 B), 20.929 (9%) had a hemorrhagic stroke (B70 A), and 19.107(8%) stayed in hospital for one day only (B70 D). As categorised with the ICD-10coding system there were 11.500 (4%) patients with subarachnoid haemorrhage(SAH, I60), 33.075 (12%) with intracerebral haemorrhage (ICH, I61), 6.482 (2%)with other non-traumatic intracranial bleeding (I62), 180.863 (66%) with ischaemicstroke (I63), and 40.882 (15%) with stroke neither defined as haemorrhagic norischaemic (I64). This implied an overall ratio of 19% haemorrhagic (I60,61,62) and81% non-haemorrhagic (I63,64) strokes for Germany in 2004.Conclusions: This nationwide analysis based on the German DRG report 2004gives a detailed overview of stroke occurrence and disease classification in Ger-many. Numbers revealed are higher than previous assumptions made by assessingdata available from cohort studies and stroke registers containing a limited sampleof patients.


A NEW METHOD OF ESTIMATING THE TIME USED BY HEALTH CAREPROFESSIONALS (HCP) FOR ACUTE STROKE CARE IN EUROPE,EUROPEAN REGISTERS OF STROKE (EROS) COLLABORATIONÖ. Saka, A.J. Mcguire, P. Heuschmann, A. Rudd, C. WolfeKings College London, Division of Health and Social Care Research, London,United Kingdom

Intro: The EROS project is assessing the provision of stroke care in 8 Europeancentres. As stroke is a labour intensive disease, time spent by the HCP make up amajor part in the resources used. The purpose of our study was to analyse the timespent by HCP.Methods: A questionnaire was developed, piloted and used in 7 of the participatingcentres to the EROS project. The forms included the description of 5 case scenarioswith different severities (case 1 the least, case 5 the most severe according to NIHscores). In the forms HCP were asked to define the activities they would carry outdaily, weekly, on admission and on discharge for each case scenario. The answerswere recorded in minutes (mins). We analysed the mean and the median time spentby HCP groups (classified as nurses, doctors and therapists which included speechand language therapists, physiotherapists, occupational therapists) for different casescenarios. We used Kruskall Wallis test to test the correlation between the timesspent by HCP in different countries.Results: 145 interviews with HCP were included. The nurses spent more timewith patients than the other HCP (mean total daily time; for nurses 25-305 mins,for doctors 5-55 mins, for therapists 0-66 mins). For all of the specialists groupsthere was a tendency to spend more time with mid range severity cases (cases 2 &3), total time spent increasing steadily from mild to moderate and declining frommoderate to severe. There was statistically significant correlation between the timespent by the nurses and the doctors (p values between 0.0001 and 0.05) but not forthe therapists.Conclusion: The HCP time use questionnaire helped gather data when the formswere filled through an interview with the HCP. Although this tool mainly will beused in costing the stroke service provision we found strong correlation betweenthe time spent by doctors and nurses in different countries. The lack of such acorrelation in the case of therapists can be explained by the differences in the waythe function of therapists are defined in EROS centres.


IMPORTANCE OF THERAPEUTIC INERTIA IN SECONDARY STROKEPREVENTION: IMPLEMENTATION OF PREVENTION AFTER ACEREBROVASCULAR EVENT (IMPACT) STUDYE. Touzé, M. Voicu, J. Kansao, R. Masmoudi, B. Doumenc, A. Ferreira,P. Durieux, J. Coste, J.-L. MasHôpital Sainte-Anne, Paris, France

Background: Many patients do not receive prevention according to recommen-dations after stroke, but the relative importance of patient- and physicians-relatedfactors is uncertain.Methods: We prospectively assessed individual factors associated with bloodpressure (BP)<140/90 mmHg and LDL-cholesterol<1g/L in a cohort of 240consecutive stroke/TIA patients (Rankin<4; ≤80 years; no major comorbidity)from a stroke unit and 3 emergency departments. A standardized assessment of risk

factors was performed 6 and 12 months after the initial event by an investigator notinvolved in the usual follow-up of patients.Results: At 6 months, 41% of patients with diagnosed hypertension at inclusionhad BP<140/90 mmHg and 55% of those with diagnosed hypercholesterolemia hadLDL<1g/L. Compliance to treatment was excellent in 81% of patients. In univariateand multivariate analyses, initiation or reinforcement of appropriate treatments wasthe main factor associated with BP<140/90 mmHg (OR=2.2; 95%CI: 1.0-4.5)and LDL<1g/L (OR=3.3; 1.3-8.7) or with decrease in BP (p<0.0001) and LDL(p<0.0001). Patients’ characteristics including sociodemographic characteristics,education, income, and knowledge of disease and risk factors were not associatedwith control of BP or LDL. Among patients with BP≥140/90 mmHg, about 40%received either no treatment or only one drug, and treatment was reinforced in only20% of them. Results were similar at 12 months with no improvement in the rateof control of risk factors.Conclusion: In-hospital initiation of secondary stroke prevention could influencethe long term quality of secondary prevention. Therapeutic inertia is an importantimpediment to achieve the BP and LDL control goals after stroke, even in relativelymotivated/compliant patients.


LONGER TERM STROKE CARE: HOW SHOULD SERVICES BEDEVELOPED IN THE POST-ACUTE PHASE?A.M. Cox, L. Kalra, A.G. Rudd, C.D.A. Wolfe, C. McKevittKing’s College London, Division of Health and Social Care Research, London,United Kingdom

Background: Organised stroke care, such as inpatient stroke units and earlysupported discharge, improves outcomes for patients. There is less evidence forthe optimal organisation and delivery of stroke care in the longer term. Followingthe MRC Framework for developing complex interventions, we undertook Phase1 work to identify the potential components of an intervention to improve longerterm stroke care.Methods: In-depth interviews with a purposive sample of health professionals(n=25) working in stroke. Participants were asked to describe existing services forpatients in the post-acute phase, identify successes and shortcomings and proposeways to improve service provision. Interviews were recorded, transcribed andanalysed using framework analysis.Results: Participants highlighted gaps in the available evidence base regarding theoptimal setting, timing, intensity and focus of therapy delivery.There are theoretical questions that need to be resolved to improve post-acuteservices. These include whether specialist care is necessary and what specialismmeans in this context, the nature of multi-disciplinary working, and how toovercome the structural and professional divisions that currently influence serviceprovision.Problems relating to the delivery of services include transfer of care between ser-vices, lack of psychology support, capacity to provide intensive therapy, inadequateservice provision for the cognitively and perceptually impaired.Conclusion: This interview study identified gaps in the evidence base, theoreticalquestions that underpin the organisation of services as well as practical problems incare delivery. These will need to be addressed in formulating an improved modelof post-acute stroke care.


IMPACT OF CHANGES TO THE DRG CLASSIFICATION ON ACUTEISCHAEMIC STROKE CARE IN GERMANYP.L. Kolominsky-Rabas, V. Ziegler, J. RüthemannDep. of Health Economics, Institute for Quality and Efficiency in Health Care(IQWIG), Cologne, Germany

Background: In 2005, the German Diagnosis Related Groups (DRG) classificationwas modified to allow, for the first time, differentiation between different forms ofacute ischaemic stroke (AIS) care.Prior to this, the German DRG system consisted of only one classification forAIS: B70B. The new classification distinguishes between complex (B70B) andnon-complex (B70E) neurological treatments for AIS. The advanced category ofcomplex neurological treatment (OPS 8-981) comprises specific diagnostic andtreatment features including potential thrombolysis commonly provided in strokeunits (SU).Methods: To assess the influence of the changes made to the DRG system in2005 on patient disease management, we analyzed the data, from more than 250of the 1.750 German hospitals, sent to the InEK (Institut für Entgeltsystem imKrankenhaus) for the years 2004-2006.


Results: Prior to the changes in the DRG classification (2004), 6.480 (34%)patients received either plain or contrast MRI (Magnetic Resonance Imaging) aspart of their stroke care. After changes made to the DRG system, 1.112 (59%) AISpatients in the B70B group and 5.971 (34%) patients in the B70E group underwentMRI in 2005. The corresponding figures for 2006 were 1.180 (71%) patients inthe B70B group and 4.613 (27%) patients in the B70E group. Concurrently, thepercentage of thrombolytic treatment given to patients increased in the complextreatment group from 15% in 2005 to 33% in 2006.Conclusions: Distinguishing between complex and non-complex AIS care is rele-vant and more accurately reflects patient disease management. Thus, the changesto the DRG system improve the frequency of appropriate MRI imaging andthrombolytic treatment in hospitals providing complex neurological treatment.


THE APPLICATION OF TELEMEDICINE FOR STROKE IN THE BALEARICISLANDSS. Tur, I. Legarda, M.J. Torres, C. JimenezSon Dureta Hospital, Palma de Mallorca, Spain

Background and purpose: Systemic thrombolysis for acute ischemic stroke isadministered only in Son Dureta University Hospital (SDUH) in the island ofMallorca. It is impossible to move a patient from Ibiza or Menorca to Mallorcain time to treat. Our main aim is to extend the use of this treatment to the otherBalearic Islands (Menorca, Ibiza) through telemedic support.Methods: The Department of Neurology of SDUH has Stroke Unit, neurologiston duty 24 hours per day and experience in thrombolytic treatment. Our hospitalsare connected by a video conference system (red ATM). All hospitals have specificequipment (Tandberg MXP) to allow us to explore a patient at a distance and a CTimage transferer.After a first phone contact, emergency physicians consult stroke neurologistsvia a two-way video conference system. Medical history, neurologic examinationaccording to National Institute of Health Stroke Scale (NIHSS) and head CT scanare reviewed to select a candidate patient for t PA treatment.Can Misses Hospital in the island of Ibiza has organized a stroke team. There isa specific stroke bed in the Intensive Medical Care Unit and we share the samestroke protocol. Both doctors and nurses have been trained. This activity began inthe island of Ibiza in July 2006. Verge del Toro Hospital in the island of Menorcais developing its own assistance process.Results: We are registering clinical data, number of contacts, number of thrombol-yses, onset to contact and to treatment time, complications, mortality, neurologicaldeficits and disability at admission, discharge and after 3 months (NIHSS andmodified Rankin Scale) and no treatment reasons.Conclusions: Telemedicine allows us to extend specialized assistance and throm-bolytic treatment to underserved areas.


COSTS OF STROKE BORNE BY INDIVIDUALS AND FAMILIES: USER-LEDDEVELOPMENT OF A PATIENT BASED COST MEASUREC. McKevitt, N. Fudge, A. Sriskantharajah, C. Coshall, C. Wolfe, KCL StrokeResearch Patients & Family GroupKing’s College London, London, United Kingdom

Background: The high costs to the state/health service associated with stroke careare documented in several economic analyses. These provide little informationabout the costs borne by families/individuals. Our Stroke Research Patients andFamily Group (PFG) identified personal costs resulting from stroke as a researchpriority but methods to assess these costs are not well developed. We report auser-led study to adapt an existing generic, but untested, cost questionnaire forcompletion by stroke patients.Method: PFG discussions and 5 individual qualitative interviews were held toidentify preferred research methods and specific cost items. These were used todevelop a topic guide for a novel qualitative method, “guided conversations”,between 10 stroke survivors/carers. These were recorded, transcribed and analysedto finalise items for the cost measure. The existing generic questionnaire wasadapted to incorporate stroke specific items, and reviewed by the PFG to ensure alltopics were covered and approve wording.Results: User-led qualitative methods led to identification of items to include ina questionnaire. These include: payment for adaptations, medications, alternativetherapies, diabetic/organic food, nutritional supplements, clothing suitable for dis-ability, transport; direct and indirect loss of family income. The cost measure wasincorporated into a structured interview questionnaire with content validity, for usein a pilot study.

Conclusion: The PFG identified an under-researched area that they regard asimportant to understanding the consequences of stroke. Their participation led tothe development of a cost measure relevant to the population under study. Themeasure is currently being piloted with people recruited to the South London StrokeRegister.


MARKET SHARE OF INPATIENT CARE STROKE UNITS FROM THE TOTALHOSPITAL CARE MEASURED BY THE DIAGNOSIS RELATED GROUPS(DRG) SYSTEMI. Boncz, A. Sebestyén, J. Betlehem, L. GulácsiDepartment of Health Economics, Policy and Management, University of Pécs,Pécs, Hungary

Aim: In the early 1990s, a Diagnosis Related Groups-like financing system wasintroduced in Hungary including all the Hungarian acute care hospitals. The aimof the study is to analyse the market share of acute stroke units according to DRGsystem.Data and methods: Data were derived from the financial database of the NationalHealth Insurance Fund Administration, the only health care financing agency inHungary (1996-2005). All the acute care stroke units were involved into the study.The following indicators were used for the analysis: number of cases, the numberof DRG cost-weights, hospital days. Regression lines and Pearson coefficients (R2)were calculated.Results: Although it was no significant changes in the number of stroke cases(18.000-20.000 patients/year), the market share of stroke care within in-patientscare – measured by the number of cases – decreased continuously from 1,07%(1996) to 0,77% (2005). The market share of stroke care within in-patients caremeasured by the number of DRG cost-weights has been also decreased from 1,06%(1996) to 0,84% (2005). The market share of stroke care from the total in-patientscare hospital days has been also decreased from 1,41% (1996) to 1,01% (2005).The market share of acute stroke care in 2005 compared to 1996 felt to 72,3% inthe number of cases, 79,3% in DRG cost-weights and 71,6% in hospitals days. ThePearson coefficients (R2) for number of cases, DRG cost-weights and hospitalsdays are 0,74, 0,72 and 0,55 respectively.Conclusion: The market share and health insurance financial conditions of acutestroke care units varied significantly between 1996-2005. The overall financialeffect of DRG system on the Hungarian stroke care seems to be relatively good orneutral, but not disadvantageous.

Experimental studies

1 Experimental studies

AN EXTENDED WINDOW OF OPPORTUNITY FORGRANULOCYTE-COLONY STIMULATING FACTOR TREATMENT IN RATFOCAL CEREBRAL ISCHEMIAJ. Minnerup, R. Wysocki, R. Laage, A. Schneider, W.R. SchäbitzUniversity of Münster, Münster, United Kingdom

Background: Granulocyte-Colony Stimulating Factor (G-CSF) is known as aregulator of white blood cell proliferation and differentiation. We and others haveshown that G-CSF is effective in treating cerebral ischemia in rodents, both relatingto infarct size as well as functional recovery. We assessed the hypothesis thatG-CSF has acute neuroprotective effects and long-term recovery effects in rat focalcerebral ischemia after delayed treatment onset.Methods: Wistar rats (n=24/group) underwent middle cerebral artery occlusion(MCAO) for 90 min. Four hours after onset of occlusion animals received 60μg/kg G-CSF i.v. over 20 min or vehicle. Infarct volumes were determined by TTCstaining. For evaluation of long-term functional outcome photothrombotic ischemiawas induced in the parietal cortex. For treatment, animals (n=10/group) were givenvehicle or 10 μg G-CSF/kg i.v. starting 24 or 72 hours after induction of ischemiaand daily repeated for 10 days. Rotarod testing was performed at 1, 2, 3, 4, 5and 6 weeks after ischemia. Brain sections were immunostained for anti-BrdU andNeuN.Results: After MCAO we observed a infarct volume reduction by 33% in the totalinfarct volume as compared to vehicle-treated rats (334.0 ± 31.5 mm3 vs. 223.3 ±27.3 mm3, p < 0.05). G-CSF treated rats in both the 24 h and 72 h time-windowperformed significantly better in the rotarod test than vehicle-treated animals. Thenumber of newly generated neurons (BrdU+/NeuN+) in the dentate gyrus wasincreased (p < 0.01) by G-CSF treatment.

Poster SessionExperimental studies


Discussion: In Conclusion we demonstrate an infarct reducing effect of a 4 htreatment delay with G-CSF in a severe hemispheric stroke model (MCAO). Inthe photothrombotic model we have shown that the time window for initiation ofpoststroke functional recovery by G-CSF is at least 72 h. This effect correlated witha profound induction of neurogenesis. G-CSF is therefore thought to activate dualmechanisms within the brain such as the previously shown neuroprotective effectplus a substantial recovery/neurogenesis inducing function.


EFFECTS OF THE TP RECEPTOR ANTAGONIST S 18886 IN A RAT MODELOF MIDDLE CEREBRAL ARTERY PHOTOTHROMBOTIC OCCLUSIONT. Hirvonen, V. Blanc-Guillemaud, L. Lerond, J. JolkkonenCNServices, Kuopio, Finland

Background: Platelet activation represents the major source of increase in TXA2biosynthesis after a cerebro-vascular event and plays an essential role in thepathogenesis of thrombotic stroke.S 18886, a TXA2 and prostaglandin endoperoxide (TP) receptor antagonist is beingdeveloped as an anti-atherothrombotic agent and is expected to be effective incerebral artery thrombosis. The aim of this study was to investigate the effects ofS 18886 in comparison to aspirin and clopidogrel, in a rat middle cerebral artery(MCA) model of focal thrombosis induced by photochemical reaction.Methods: S 18886 (10 mg/kg), aspirin (100 mg/kg) and clopidogrel (30 mg/kg)were orally administered to male Wistar rats (n=10-16/group) 2 hours beforethrombosis induction. A group receiving the vehicle served as control. Blood flowin the distal MCA was monitored by a laser Doppler flowmeter for a 70-min period.Time to occlusion (TO) and total patency time (TPT), expressed as a percentage ofobservation period, were measured.Results: There were no statistical differences in time to occlusion between treat-ment groups. However S 18886 and clopidogrel prevented MCA occlusion in 2/10and 2/11 animals, respectively. S 18886 significantly increased the incidence of re-opening (87.5% versus 25% in control group; p<0.05) and the TPT measured whenMCA blood flow was above 70% of baseline (67.9 ± 8.3% versus 20.9 ± 4.2%in control group, p<0.001). The increases in the incidence of reopening (55.5%)and the TPT (41.1 ± 12.4%), observed with Clopidogrel were not statisticallysignificant. Aspirin had no effect on these parameters.Conclusion: The present study demonstrates that single oral administration ofS 18886 increases incidence of MCA reperfusion and improves MCA bloodflow during reperfusion. The findings suggest that anti-platelet therapy with theTP receptor antagonist S 18886 can be potentially beneficial in cerebro-vasculardiseases


HYPERBARIC OXYGEN TREATMENT COMBINED WITH THROMBOLYTICTHERAPY REDUCES INFARCTION SIZE IN EXPERIMENTAL ISCHEMICSTROKEL. Küppers-Tiedt, A. Manaenko, A. Günther, D. Michalski, A. Wagner,D. SchneiderKlinik und Poliklinik für Neurologie; Universität Leipzig, Leipzig, Germany

Background: In acute ischemic stroke effective treatment is still limited. The onlyapproved therapy is thrombolysis with rtPA within the first three hours, but thistherapy is only an option for a small number of patients due to the time windowand the risk of hemorrhage. This study investigated the effects of thrombolysis incombination with hyperbaric oxygen therapy (HBOT) in acute ischemic stroke inrats.Methods: In 22 male Wistar-Rats an ischemic stroke was induced by embolicocclusion of the middle cerebral artery using clots of 20 mm length. After strokeinduction the animals were randomized to one of three groups: 1) Control (room airand placebo), 2) Thrombolysis (room air and rtPA), 3) HBOT (hyperbaric oxygena-tion (2.5 ATA) and rtPA). rtPA was given intravenously two hours after the embolicstroke. The animals were exposed to room air/HBOT before or during thrombolysisfor one hour. 6 hours after the stroke the animals were sacrificed and brain slicesstained with Triphenyltetrazoliumchloride to calculate the infarct volume.Results: The ischemic infarctions calculated 6 hours after the embolic strokeextended to 50% of the hemisphere. Thrombolysis alone did not reduce theinfarction size (about 43% of the hemisphere), but the combination of throm-bolysis and HBOT lead to a significant decrease of infarction size to about20% of the hemisphere (HBOT before rtPA p=0.01; HBOT during rtPA p=0.02,Student-Newman-Keul-Test).Discussion: In this study we could demonstrate a significant reduction of infarctionsize in an embolic stroke model in rats by combining thrombolytic and hyperbaric

oxygen therapy. These effects were observed 6 hours after the ischemic stroke.Hence further studies are needed to investigate the long term effects of thecombination therapy.


BLOOD-BRAIN BARRIER (BBB) OPENING AFTER TRANSIENT FOCALCEREBRAL ISCHEMIA IN RATSI. Marinkovic, A. Durukan, U.A. Ramadan, D. Strbian, M. Pitkonen,E. Pedrono, L. Soinne, T. TatlisumakBiomedicum Helsinki, Helsinki, Finland

Background: It is widely believed that BBB breakdown after transient focalcerebral ischemia occurs in a biphasic pattern. The aim of this study was to evaluatequantitatively the pattern of BBB damage after transient focal cerebral ischemia andits correlation with the size of the ischemic lesion by the use of contrast-enhancedmagnetic resonance imaging (MRI).Methods:Adult male Wistar rats (n=10) were underwent 90 minutes of transient focalcerebral ischemia with the suture occlusion method and imaged with MRI at 4.7Tesla at 2, 24, 48, 72 hr and 1 week after reperfusion. Diffusion weighted imaging,FLAIR (fluid attenuated inversion recovery), and T1-weighted (with and withoutcontrast agent gadolinium) sequences were acquired. After calculating lesion areaand contrast-enhanced areas, their ratio was obtained. The gadolinium permeability(by calculating Ki values via the Patlak plot approach) was estimated for allimaging time points.Results: In all post contrast images, gadolinium enhancement occurred in a similarspatial pattern with ischemic lesion and there was no statistically significant differ-ence between ratios (mean ratio was 0.94 for T1-weighted sequence, p=0.06 and0.91 for FLAIR, p=0.6). The Ki values of ischemia regions (cortex and subcortex)for all groups were statistically significant (p<0.01) compared with the identicalregions in the contralateral brain hemisphere. The difference in Ki between differenttime points was not statistically significant (p=0.38).Discussion: Our results showed that BBB leakage to gadolinium (molecular weight590 Da) occurs in a monophasic pattern during the period of 2 hr to 1 week aftertransient focal cerebral ischemia in rats and BBB damaged brain area is highlysimilar to the ischemic area depicted on DWI.


EFFECT OF HIGH-DOSE OESTROGEN THERAPY ON CEREBRALPLASTICITYAFTER TRANSIENT FOREBRAIN ISCHAEMIA IN GERBILE.A. Wappler, A. Gal, G. Szilagyi, J. Vajda, J. Skopal, K. Felszeghy, C. Nyakas,Z. NagyNational Institute of Psychiatry and Neurology, National Stroke Center,Budapest, Hungary

Background: After ischaemic injury repair mechanisms in the brain tissue reducethe functional deficit. Neuroprotective effect of oestrogen is well documented,however its effect on repair mechanisms are still not elucidated. In our workwe focused on the expression of plasticity genes and functional recovery afteroestrogen treatment in transient ischaemic model.Methods: 3 month-old ovariectomized femail gerbils (n=40) were subjected to 10min transient forebrain ischaemia or sham procedure. Half of the ischemic animalswere pre-treated i.p. with 4 mg/kg body weight oestrogen 20 min previous tosurgery. From one series of animals brain samples were collected on postoperativeday 4 for histological and molecular biological examinations. Paraffin-embeddedbrain slices were stained with TUNNEL-caspase double labelling fluorescent anti-sera. Marker mRNA levels were determined with real-time PCR. Gene expressionlevels were assessed by ddCT method using TaqMan gene expression assays. Onother series of animals attention and learning behaviour were tested in spontaneousalternation, novel object recognition and spatial learning paradigms beginning frompostoperative day 7.Results: Oestrogen significantly decreased the number of apoptotic and necroticcells in CA1 region. Oestrogen treatment resulted in a significant increase inBcl-XL, nestin and GAP-43 mRNA expression. In ischaemic insult inpaired atten-tion and working memory in all behavioural tests, while oestrogen pre-treatmentimproved attention and prevent or decreased memory deficit.Discussion: Our novel finding is that oestrogen is not just neuroprotective in ourmodel, but augmented the expression of plasticity genes and these correlate well tobetter outcome in behaviour tests.This work was supported by OTKA T037887, GVOP-3.1.1.-2004-05-0389/3.0.



S 18886, A THROMBOXANE A2 RECEPTOR ANTAGONIST, PREVENTSOCCURRENCE OF SPONTANEOUS BRAIN DAMAGE IN STROKE-PRONERATS VIA ANTI-INFLAMMATORY ACTIVITIESP. Gelosa, E. Nobili, A. Gianella, V. Blanc-Guillemaud, L. Lerond, U. Guerrini,E. Tremoli, L. SironiDept. Pharmacol. Science, Univ. Milan; Italy; Inst. Recherches InternationalesServier, Courbevoie, Milan, Italy

Background and aim: Spontaneously Hypertensive Stroke-prone rats (SHR-SP)are an established model of human cerebrovascular disease. In this rat strain, thedevelopment of hypertension and inflammation precedes the appearance of brainabnormalities. The aim of the present investigation was to assess the efficacy of S18886, an orally active antagonist of TP-receptors (the receptors for thromboxaneA2), in protecting the brain of SHRSP and whether this effect was related to itsanti-inflammatory properties.Methods and results: Male SHR-SP (n=10 per group), fed with a high-salt diet,received by gavage vehicle or S 18886 (3 or 30 mg/kg/day). In vehicle-treatedanimals, brain lesions, as detected by magnetic resonance imaging, developed spon-taneously after 40 ± 2 days (mean ± SEM). Treatment with S 18886 had no effecton arterial blood pressure, significantly delayed the appearance of brain damage,at the dose of 30 mg/kg/d (p<0.001), and increased survival, in a dose dependentmanner (p<0.001 and p<0.0001 at the dose of 3 and 30 mg/kg/d respectively).In comparison with vehicle-treated SHRSP, treatment with S 18886 (30 mg/kg/d;n=5), preserved brain tissue by preventing macrophage infiltration (ED1 positivecells) (p<0.05), and reduced the accumulation of perivascular macrophages (ED2positive cells) and lymphocytes T helper (CD4+ positive cells) as assessed byimmunohistochemistry. Furthermore, S 18886 attenuated the transcription of thepro-inflammatory cytokines IL-1beta, TNF-alpha, IL-6, and MCP-1, as assessed byRT-PCR.Conclusion: These data indicate that S 18886 prevents the occurrence of sponta-neous brain damage in SHRSP by reducing inflammation, suggesting that S 18886may exert a beneficial anti-inflammatory effect in cerebrovascular disease.


CDP-CHOLINE INCREASES EAAT2 ASSOCIATION TO LIPID RAFTS ANDAFFORDS NEUROPROTECTION IN EXPERIMENTAL STROKEI. Lizasoain, O. Hurtado, J.M. Pradillo, D. Fernández-López, T. Sobrino,T. Gasull, M. Castellanos, F. Nombela, J. Castillo, M.A. MoroFacultad de Medicina, Universidad Complutense Madrid, Madrid, Spain

Background: EAAT2 is responsible for up to 90% of all glutamate transportand has been reported to be associated to lipid rafts. In this context, we haverecently shown that CDP-choline induces membrane translocation of EAAT2. SinceCDP-choline preserves membrane stability by recovering sphingomyelin levels aglycosphingolipid present in lipid rafts, we have decided to investigate whetherCDP-choline increases association of EAAT2 transporter to lipid rafts.Methods and results: For lipid rafts isolation, brain homogenates from each groupwere subjected to a discontinous sucrose gradient in the presence of Brij-58 and 8fractions were collected. Flotillin-1 was used as a marker of lipid rafts due to itsknown association to these microdomains. We have found that flotillin-1 was foundmainly in fractions 2 and 3 and their levels were similar in all the groups studied.EAAT2 protein was predominantly found colocalised with flotillin-1 in the fraction2, and CDP-choline increased EAAT2 levels in fraction 2 at both times examined(3 and 6 hours after 1g/Kg CDP-choline administration). Furthermore, exposure tomiddle cerebral artery occlusion also increased EAAT2 levels, an effect which wasfurther enhanced in those animals receiving 2 g/Kg CDP-choline 4 hours after theocclusion. Infarct volume measured at 48 h after ischemia showed a reduction inthe group treated with CDP-choline 4 h after the ischemic occlusion.Conclusions: We have demonstrated that CDP-choline induces a re-localisationof EAAT2 into lipid raft microdomains in rat brain. This effect is also foundafter experimental stroke, when CDP-choline is administered 4h after the ischemicocclusion. We have also shown that this delayed post-ischaemic administration ofCDP-choline induces a potent neuroprotection.


ULTRASOULD-THROMBOLYSIS WITH 488KHZ - SAFETY-STUDIES WITHAN MRI-BASED RAT STROKE MODELM. Walberer, M. Nedelmann, D. Schiel, K. Volk, P. Reuter, M. Kaps,T. Saguchi, G. Bachmann, H. Furuhata, T. GerrietsUniversity Giessen, Germany; Kerckhoff-Clinic Bad Nauheim; Jikei UniversityTokyo, Giessen, Germany

Objective: Ultrasound can enhance the effect of i.v.-thrombolysis in acute stroke.First clinical trials with 2MHz-ultrasound revealed an improved recanalisation ratebut yet no convincing clinical benefit. Lower ultrasound frequencies might bemore effective. However, clinical trials as well as animal experiments have shownsevere side-effects. Safety of new therapeutic ultrasound devices thus needs to bedetermined.Methods: Male Wistar rats were subjected to middle cerebral artery-occlusionfor 90 minutes followed by reperfusion (suture technique). Rt-PA (Actilyse®) wasinjected intravenously thereafter. Then transcranial ultrasound treatment (488kHz;sweep: 10%; 0.7W/cm2; continuous wave) was started and continued for 60minutes. Sham treated animals were used as controls. Intracerebral temperaturewas recorded during ultrasound application in a sub-study.MRI (Bruker PharmaScan, 7.0 Tesla) was performed after 24h. Ischemic lesionvolume (T2-WI and DWI) and vasogenic brain edema (T2-relaxation time) werequantified. T2*-WI was used to determine hemorrhagic complications.Results: 488kHz-ultrasound treatment did not noteworthy affect brain temperature.Ultrasound did not increase lesion volume or edema formation. No hemorrhagiccomplications could be detected on T2*-weighted imaging.Conclusion: The 488kHz-device did not exert any side-effects in our MRI-basedrat stroke model. Further safety- and efficiency-studies are required prior firstclinical applications.


GW3965, AN AGONIST OF THE LXR NUCLEAR RECEPTORS, ISNEUROPROTECTIVE IN EXPERIMENTAL STROKEM.A Moro, J.R Morales, M.P. Pereira, J.R. Caso, O. Moldes, J. Vivancos,C. Gubern, J. Serena, A. Davalos, I. LizasoainFacultad de Medicina, Universidad Complutense Madrid, Madrid, Spain

Background and purpose: Liver X receptors alpha (LXR-alpha) and beta (LXR-beta), also known as NR1H3 and NR1H2, respectively, are ligand-activatedtranscription factors that belong to the superfamily of nuclear receptors. Apart fromtheir role in the regulation of cholesterol homeostasis and fatty acid metabolism,LXR receptors have been described to inhibit the expression of inflammatory medi-ators such as inducible nitric oxide synthase (iNOS), cyclooxygenase-2 (COX-2) ormatrix metaloproteinase-9 (MMP9). Since these anti-inflammatory actions mightbe useful in stroke, we have investigated the effects of the LXR agonist GW3965on stroke outcome in a rodent model of cerebral ischaemia by permanent occlusionof the middle cerebral artery (MCAO).Methods: Male Fischer rats were used. Infarct size: 48 after MCAO, animalswere sacrificed with an overdose of pentobarbitone and a serial of 2 mm thickcoronal slices were made and stained with 2,3,5-triphenyltetrazolium chloride 1%in 0.2 phosphate buffer. Infarct size was determined using a computer imageanalysis system. Experimental groups were control, permanent middle cerebralartery occlusion (MCAO), and MCAO+GW3965 (20mg/kg). GW3965 or vehicle(DMSO) were administered i.p. 10 min after MCAO. Stroke outcome was assessedby measurement of infarct size. Protein expression of iNOS, COX-2 and MMP9in cerebral cortex were studied by Western blot and data were expressed as % ofdensitometry of bands in the MCAO group.Results: The administration of the LXR agonist GW3965 reduced infarct volume(180.4 ± 7.7 mm3 vs. 150.2 ± 5.5 mm3 in MCAO and MCAO+GW3965,respectively, n=6-10, p<0.05). Furthermore, GW3965 reduced MCAO-inducedexpression of iNOS (41 ± 1% of MCAO, n=4, p<0.05), COX-2 (43 ± 3% ofMCAO, n=4, p<0.05) and MMP-9 (43 ±1% of MCAO, n=4, p<0.05) was reducedin animals treated with GW3965.Conclusions: Activation of LXR receptors induce neuroprotection in experimentalstroke, very likely due to anti-inflammatory mechanisms.




ENHANCEMENT OF SENSORIMOTOR RECOVERY UNDERCEREBROLYSIN TREATMENT IN A RAT MODEL OF FOCAL CEREBRALINFARCTIONM. Hitzl, J.M. Ren, D. Sietsma, S.M. Qiu, H. Moessler, S.P. FinklesteinEBEWE Pharma GmbH Nfg.KG, R&D Neuro Products, Unterach, Austria

Background: Many efforts have been made to find drug products having the po-tential to reduce infarct volume and/or promote neurological recovery after stroke.Cerebrolysin a unique drug product composed of neurotrophic and neuroprotectivepeptides from biological origin seems to fit perfectly into this specification. It hastherefore been tested in an rodent model of stroke.Methods: Focal cerebral infarction was produced in mature male Sprague-Dawleyrats by occlusion of the proximal right middle cerebral artery. Animals havesubsequently been treated with Cerebrolysin at a daily dose of 1.0, 2.5, or 5.0 ml/kgbody weight starting 24, 48 or 72 hours after stroke onset for 21 days. Behavioralassays were performed before and during Cerebrolysin treatment (forelimb andhindlimb placing and body swing tests). At the end of Cerebrolysin treatment theinfarct volume has been determined using a computer-interfaced imaging system.Results: Cerebrolysin treatment of rats after focal cerebral infarction resultedin enhanced recovery of sensorimotor function compared to vehicle-treated ani-mals. Enhancement of sensorimotor recovery has been found when Cerebrolysintreatment at a dose of 2.5 ml/kg was started 24 or 48 hours after stroke onset.Discussion: These findings clearly demonstrate that administration of Cerebrolysinafter stroke can enhance neurological recovery.


RHEOENCEPHALOGRAPHY: A NON-INVASIVE METHOD TO ASSESS THEELECTRICAL IMPEDANCE CHANGES RELATED TO THE PULSATILITY OFTHE CEREBRAL BLOOD FLOWJ.M. Pons, J.J. Pérez, P. Ortiz, E. Guijarro, A. Navarré, J. Sancho1Consorcio Hospital General Universitario Valencia; 2Centro de investigación einnovación en bioingeniería Universidad Politécnica de Valencia, Valencia,Spain

Objective: Rheoencephalography (REG) measures the electrical impedancechanges of the head caused by the pulsatility of the cerebral blood flow (CBF).However, the use of REG in the clinical practice is limited because signal is buriedby the extracranial component. Our research group has formulated a mathematicalalgorithm that allows the extraction of the intracranial component from the REGsignal. The main goal of this work is to validate the separation method. For thispurpose, an experimental model that arrests mechanically the extracranial bloodflow is used.Material and methods: REG signal was acquired in healthy volunteers in twoconditions: (i) normal and (ii) during the arrest of the scalp blood flow by means thescalp compression with a pneumatic cuff. Subsequently, the intracranial componentextracted with our algorithm was statistically compared with the REG tracesrecorded in scalp compression condition.Results: Intracranial component extracted by our algorithm matches well withthe REG trace recorded in scalp compression condition. Additionally, the mor-phology of the extracted intracranial component agrees with the intraparenchymalimpedance traces previously described in the literature.Conclusions: Our results suggest that the intracranial REG component can bereliably extracted from the raw REG signal by cancelling the scalp blood artifact.This method could provide a new non-invasive technique to assess the cerebralblood flow. Nevertheless, additional works would be necessary to check and toassess the diagnostic capability of our REG technique.This work was supported by grant PI04/0303 from the Instituto de Salud CarlosIII (Fondo de Investigación Sanitaria) in the framework of the “Plan Nacional deInvestigación Científica, Desarrollo e Innovación Tecnológica (I+D+I)”.


FIRST EXAMINATIONS WITH AN AUTOMATICALLY OPTIMIZEDCOMPUTER MODEL FOR INDIVIDUAL SIMULATIONS OF CEREBRALHEMODYNAMICSF.C. Roessler, V. Metzler, R. Grebe, G. SiegelClinic for Neurology, UK-SH, Campus Lübeck, Lübeck, Germany

Background: During the medical treatment of some cerebrovascular diseases itis necessary to occlude brain supplying arteries. Sometimes these interventionscause a cerebral ischemia and the patient will suffer from strokes. Until now this

risk can only be estimated by invasive diagnostics, which themselves bare the riskof cerebrovascular accidents. Therefore we searched for a reliable non-invasivediagnostic tool for preoperative risk estimation.Methods: A computer model of the brain supplying arteries was designed, whoseparameters can be determined by non-invasive measurements and picture-givingprocedures to fit the model to the individual physiological state of the patient.Comparing time series generated by the model with those measured at the patientthe model can be evaluated and in the case of insufficient results its parameters canbe changed by an optimization process based on evolutionary algorithms.Results: The patient adapted models behaved physiologically and showed goodagreement between the modelled data and those recorded from the subjects. Thereaction of individual cerebrovascular systems in critical situations similar to theocclusion of the internal carotid artery was investigated by special scenarios. Eventhough in this first step of examination the optimization process was only relatedto a few parameters, it became obvious that evolutionary algorithms are suitableprovided that some physiological laws are considered.Discussion: Although some difficulties remain concerning the parameter estimationand optimization we hope, that this flexible, time saving, cheap and non-invasivemethod makes a valuable contribution to avoid complications of induced vesselocclusion during medical treatment by an improved operation planning.


NATURAL REGULATORY CD4+CD25+FOXP3+ T-LYMPHOCYTES (TREG)PREVENT DELAYED INFARCT GROWTH BY AN INTERLEUKIN-10DEPENDENT MECHANISMR. Veltkamp, E. Suri-Payer, C. Sommer, C. Veltkamp, H. Doerr, T. Giese,A. LieszUniversity Heidelberg, Heidelberg, Germany

Background and aims: Inflammatory cascades contribute to secondary ischemicbrain damage. Tregs are important anti-inflammatory modulators in various inflam-matory diseases. We studied the role of Tregs in ischemic stroke.Methods: Focal ischemia was induced by transtemporal MCAO. Tregs wereeliminated either by preischemic depletion with mAb (clone PC61) in C57Bl/6mice or by adoptive transfer of CD4+CD25- into rag2-/- mice. Infarct volume andcerebral cytokine expression (RT-PCR) were measured at various time points afterMCAO. Effect of IL-10 was tested by intraventricular injection or by transfer ofTregs from IL -10 -/- mice.Results: Treg depletion had no effect 24h after MCAO, but Treg-depletedmice had significantly larger infarct volumes 7d after MCAO (control: 7.4mm3;antibody-treated: 12.1mm3). Correspondingly, transfer of CD4+25- T cells intolymphocyte-deficient rag2-/- mice resulted in larger infarcts (12.9mm3) than trans-fer of CD4+ cells (6.8mm3; p<0.05). Intraventricular IL-10 reversed this effect.Treg derived IL-10 was particularly important as adoptive transfer of CD4+CD25+cells derived from IL-10 -/- failed to prevent infarct growth. In mice lackingTregs, RNA levels of proinflammatory cytokines were significantly more elevatedin ischemic hemisphere compared to control (6h after MCAO: TNFa 4x; IL-1b 2x;72h after MCAO: IFNg: 5x). Cerebral invasion of Tregs became first detectable72h after MCAO by FACS and immunohistochemistry.Conclusions: Tregs are master anti-inflammatory modulators in ischemic strokewhich reduce secondary infarct progression by downregulating proinflammatorycytokine –induced cell death. Based on our findings, this effect is probably mediatedby early humoral IL-10 signalling and by delayed Treg invasion.


NONLINEAR ANALYSIS OF BRAIN SPIROGRAPHY SIGNALS - THE WAY TOA NEW NON-INVASIVE DIAGNOSTIC TOOL (A PILOT STUDY)M. Swierkocka-Miastkowska, G. OsinskiDepartment of Neurology for Adults, Medical University of Gdansk, Gdansk,Poland

Background: Ischaemic stroke is associated with disturbances of respirationrhythm. The purpose of this study was to analyze breathing patterns of acute phasestroke patients in comparison to healthy subjects.Methods: Brain spirography (BSG) as a new method of experimental clinicalbreath research was deviced and tested in Medical University of Gdansk. It has adetecting system coupled with pressure sensors. Signals from the sensors throughthe analog-digital converter are transferred to the computer for making visualrepresentation of respiration activity on a screen. Data from 55 patients withfirst-ever supratentorial ischaemic stroke and a control study of 25 volunteers wereobtained. In patients the respiration curves were measured 3 times during first5 days of hospitalization. All patients were assessed with the National Health


Institues Stroke Scale and (if needed)Glasgow Coma Scale. The curves wereanalyzed with 3 nonlinear methods: Return Map Plot (RMP), calculation of FractalDimension (FD) with Higutchi algorithm and graphical representatiom of VisualMap Plot (VMP).Results: We analyzed 25 physiological and 165 pathological data sets. As a resultwe obtained average values of FD for both groups: in the group of volunteersFD=1.599 ±0.072, in stroke patients FD=1.873±0.076. The values analyzed withANOVA test are significantly different (p<0.003).Discussion: As a number of stroke patients is instantly high, it is very importantto develop quick, easy-to-use and non-invasive methods to monitor acute phase ofstroke. A set of graphical interpretation of RMP and VRP together with value ofFD and estimation of trends in groups of signals could give us clinically importantinformation by visual representation of the analysis of dynamic status of respiratorypattern.


MRI AND BEHAVIOR EFFECTS OF EARLY INTRAVENOUS DELIVERY OFMESENCHYMAL STEM CELLS AT EXPERIMENTAL CEREBRAL INFARCTIN RATSL. Gubskiy, K. Yarygin, O. Povarova, Yu. Pirogov, R. Tairova, A. Dubina,I. Cheblakov, D. Kupriyanov, V. SkvortsovaFundamental and Clinical Neurology Department, Russian State MedicalUniversity, Moscow, Russian Federation

Background: To measure the effect of early intravenous delivery of mesenchymalstem cells (MSCs) on neurological and neurobehavioral functional deficits andMRT volume of experimental cerebral infarct in rats.Methods: 3-month-old 19 male Wistar rats (weight 180 to 250 g) were subjected tofocal ischemia in the region of MCA by electrocoagulation before bifurcation intofrontal and parietal branches under intraperitoneal anesthesia by chloral hydrate(300 mg/kg). The rats were randomized into three groups: sham (3), control (8)and experimental (8 animals with intravenous delivery of 6 million of MSCs on the1 - 2nd days after operation). MRT was performed on 1-2 and 7 days at BioSpec70/30, neurological and behavior functional tests (elevated cross - maze, open-fieldtests) also where performed.Results: Neurological severity scores in experimental and control groups on 1, 7and 14 days after operation were equal but lower than those in the sham-operatedgroup. At open-field test the horizontal activity of control rats was higher thanexperimental animals. There were no significant differences between control andMSC groups on the cross-maze on 10 but not 20 day. Before delivery of MSCsvolumes of cerebral infarct (at T2-weighted imaging) were without significantdifference between control and experimental groups. There was significant decreaseof the volumes in both groups on 7 day.Discussion: Early intravenous delivery of MSCs did not change the rate of decreaseof the volume of cerebral infarct at the first 7 days of experimental cerebral infarctin rats. There were no difference between groups at neurological and behaviorfunctional tests except for more high activity of control rats at open field test.


OXYGEN-GLUCOSE DEPRIVATION-INDUCED CELLULAR CHANGES INORGANOTYPIC SLICE CULTURES OF THE HIPPOCAMPUS: PROTECTIVEEFFECT OF (-)DEPRENYLB. Bali, Z. Nagy, K.J. KovácsSemmelweis University, Budapest, Hungary

Background: (-)Deprenyl is an irreversible inhibitor of type B monoamine oxidase(MAO-B), which is now used as a neruoprotective compound for treatment ofParkinson’s or Alzheimer’s diseases. Evidence suggests that the therapeutic efficacyof deprenyl may not be related exclusively to the inhibition of the enzyme MAO-B,however the cellular mechanisms underlying its neuroprotective effect remainedunknown.Methods: To test the impact of deprenyl on ischemia-induced changes in vitro, wefollowed the time course of propidium iodide (PI) uptake as an indicator of neuronalcell death in organotypic hippocampal slice cultures exposed to oxygen-glucosedeprivation (OGD) for 45 min. The expression of apoptotic factors (Bcl-2, Bcl-xland Bax) as well as of the proinflammatory citokine interleukin-1 (IL-1) wasmeasured by polymerase chain reaction (PCR)Results: The first signs of neuronal death were detected 2 hours after OGD andwere extended to all subfields of the hippocampus by 24 hours post-injury. Presenceof deprenyl (10-9 M) significantly delayed the cell death induced by the insult.Exposure of control cultures to deprenyl significantly increased the abundance ofBcl-2 and Bcl-xl mRNAs as revealed by RT-PCR. OGD resulted in an elevation of

anti-apoptotic factors and IL-1, while the expression pro-apoptotic bax remainedunchanged.Discussion: These data suggest that deprenyl is neuroprotective in an in vitromodel of ischemia. Although deprenyl upregulates the expression of Bcl-2 underbasal conditions, its effect on anti-apoptotic factors is not significantly manifestedduring OGD.


BCL-2 AND BCL-XL GENES THERAPY INCREASES PLASTICITY ANDCELL CYCLE GENES EXPRESSION AFTER HYPOXIA IN PC12 CELLSA. Gal, G. Szilagyi, E. Wappler, Z. Bori, J. Vajda, J. Skopal, Z. NagyNational Institute of Psychiatry and Neurology, National Stroke Center,Budapest, Hungary

Introduction: Hypoxia induces cell necrosis and/or apoptosis. Antiapoptotic genetherapy could be an option to prevent the cell death and activate the repair mecha-nisms. In this study we measured the expression of plasticity and pro/antiapoptosisgenes (Bcl-2, Bcl-XL, Bax, synapsin-1, nestin and c-fos) in PC12 cell culturesystem after adenovirus containing Bcl-2 or Bcl-XL gene delivery. We found pre-viously that the gene transfer has a cytoprotective effect, protects the mitochondrialfunction and augmented repair protein GAP-43 expression after hypoxic insults.Materials and methods: The cells were treated by Argon gas (1 hr) for inductionhypoxic cell injury followed by 24 hrs of restored oxygen. The cells were infectedwith adenovirus constructs contaning Bcl-2 or Bcl-XL gene utilized before orafter hypoxia. We examined the selected mRNA levels with real-time PCR. Thegene expression levels were determinated by ddCT method using TaqMan geneexpression assays.Results: Hypoxia and reoxygenization increased the pro-apoptotic Bax gene ex-pression while the c-fos mRNA level was decreased. Gene transfers of Bcl-2 orBcl-XL resulted in a significant increase of Bcl-2, Bcl-XL, synapsin-1, nestin andc-fos mRNA expression levels after hypoxic insults.Conclusions: In our in vitro model, Bcl-2 or Bcl-XL anti-apoptotic gene deliverywas not only cytoprotective but it augments repair genes expressions after hypoxicinsults. The double actions of these genes appear to be beneficial in preventinghypoxic cell injury. However, the link between the augmented anti-apoptotic andrepair mRNA expression is not clear at the moment.Keywords: PC12, hypoxia, Bcl-2, Bcl-XL gene transfer, plasticity genes


(-) DEPRENYL-N-OXID IS NEUROPROTECTIVE AFTER TRANSIENTHIPPOCAMPUS ISCHEMIA IN GERBILSG. Szilágyi, K. Magyar, Z. NagyNational Stroke Centre, Budapest, Hungary

Introduction: As we reported previously, (-)deprenyl-N-oxid (DNO) has signifi-cant cytoprotective effects in PC12 cell culture. Decreased cell death and reactivespecies production and increased mitochondria trans-membrane potential weredemonstrated after hypoxic injury. In this paper we tested the neuroprotective effectof DNO on hippocampus in a model of transient, bilateral common carotid arteryocclusion in Gerbils.Methods: We performed transient, bilateral carotid artery occlusions for 10 min-utes. The treated group was injected 0.4 mg/kg/day of N-oxid-deprenyl intraperitoneal for 4 days. The control rats were injected with the vehicle only. Thegerbils were killed after 4 days. On the formaldehyde fixed sections TUNEL andcaspase-3 immuno-histochemistry were performed and NIKON/BIORAD confocalmicroscope was used for analysis. Consecutive images were taken from the hip-pocampus. The TUNEL and caspase-3 positive cells were counted automaticallywith Image J 1.37 software (NIH, USA).Results: All the caspase-3 positive cells were TUNEL positive too. In the CA2regions we found significantly less caspase positive neurons az there were found inthe controls. Similarly in the same region significant decrease in number of TUNELpositive cells were counted.Discussion: In our previous study we demonstrated the neuroprotective effect ofdeprenyl. This drug decreased significantly the TUNEL labelled and TUNEL-caspase-3 double labelled neurons in the penumbra region after permanent MCAocclusion in rat. In the present experiment one of the deprenyl metabolite DNOdecreased similarly the number of TUNEL positive cells in the CA2 hippocampusregion. DNO cytoprotective effect was previously documented in PC12 cell culture.The neuroprotective effect of the highly polar molecule in the in vivo model couldbe related the transient opening of BBB in this model.Grant: OTKA T037887.



Vascular surgery and neurosurgery/Interventinalneuroradiology

1 Vascular surgery and neurosurgery/Interventinalneuroradiology

INTRA-ARTERIAL AND INTRAVENOUS THROMBOLYSIS IN ACUTEISCHEMIC STROKE FROM CAROTID “T” OCCLUSIONP. Nencini, S. Mangiafico, M. Nesi, I. Romani, G. Cagliarelli, M. Pratesi,V. Palumbo, M. Cellerini, A. Rosselli, D. InzitariCareggi Hospital, Florence, Italy

Background: Outcome in acute ischemic stroke from internal carotid artery oc-clusion is poor with high mortality or severe long-term disability. We evaluated ifintra-arterial (IAT) or intravenous thrombolysis (IVT) may influence outcome.Methods: From February 2004 to August 2006, twenty-nine patients with acuteinternal carotid artery “T” occlusion were admitted to two Hospitals in Florence,Italy. All patients underwent screening for IVT (SITS-MOST protocol), colourduplex sonography or cerebral angiograph, and had a 3-month follow-up withmRankin scale (mRS).Results: Eleven patients (male 64%, mean age 66 years, mean NIHSS 20) weretreated with IATwithin 6 hours from symptom onset; 11 patients (male 27%, meanage 72 years, mean NIHSS 18) with IVT within 3 hours from symptom onset; and7 patients (male 86%, mean age 64 years, mean NIHSS 18) had standard treatment.Recanalization was achieved in 8/29 (27.6%) patients (6 TIMI 3 and 2 TIMI 2), alltreated with IAT. Four out of 11 (27.3%) IAT patients had a 3-month good (mRS0-2) outcome compared with no patients in both the IVT and standard treatmentgroup. The 3-month mortality rate was 36% in IAT, 27% in IVT, and 43% inthe standard group, respectively. Symptomatic haemorrhage occurred only in IATgroup (27% of patients).Conclusions: The prognosis of ischemic stroke due to internal carotid artery“T” occlusion remains severe. Our data may suggest a favourable effect on theintra-arterial approach. More data are needed to confirm this hypothesis.


RETROSPECTIVE VALIDATION OF THE ABCD SCORE IN PATIENTSPRESENTING WITH TRANSIENT ISCHAEMIC ATTACKS UNDERGOINGCAROTID ENDARTERECTOMYS. Shaikh, J. Brittenden, E. MacAulay, M.J. MacleodUniversity of Aberdeen, Aberdeen, United Kingdom

Introduction: Patients who have experienced a transient ischaemic attack are atrisk of developing a stroke especially within the first two weeks after a TIA. TheABCD score aims to predict an individual patient’s risk, thus facilitating the abilityto fast-track investigation and treatment of the high-risk group who have a score of5 or 6. We aimed to assess the ABCD score of patients who had undergone carotidendarterectomy.Method: 194 of patients who underwent CEA between January 2001 and December2005 were identified from a prospectively collected database.Results: 90 (46.4%) patients undergoing CEA presented with a TIA. The remainingoperations were performed for cerebrovascular accident (n=59), amaurosis fugax(n=36) and asymptomatic carotid disease (n=9). All patients had an ipsilateralhigh-grade internal carotid artery stenosis (>70%). Of the 89 patients with TIAs(case-notes of 1 patient were destroyed), the median age was 71years (range45-83) with a male to female ratio of 1.6:1. Post-operative complications included2/89 (2.22%) TIAs, 2/89 (2.24%) lingual nerve paraesthesia, and 4/89 (4.5%)haematomas none of which required drainage. The ABCD scores were as follows:1, n=4 (4.49%); 2, n=15 (16.85%); 3, n=19 (21.34%); 4 n=22 (24.71%); 5 n=17(19.10%); 6 n =12 (13.48%).Conclusions: All patients undergoing CEA for TIAs were treated as per therecommendations of the European carotid trial, yet according to the ABCD scoretwo thirds of these patients would be considered to be at low risk of a subsequentneurological event. These patients would not have been fast-tracked for treatmentand thus further validation of this score is urgently required.


CAROTID ENDARTERECTOMY: COMPLICATIONS AND CLINICALOUTCOMER. Martinez, Y. Silva, J.A. Amado, O. Andres, J. Puig, S. Pedraza,M. Castellanos, J. SerenaHospital Universitari Dr. Josep Trueta, Girona, Spain

Background: To investigate the rate of perioperative complications and the clinicaloutcome of endarterectomy in patients with symptomatic carotid-artery stenosis ofat least 70% in daily clinical practice. High risk patients were included.Method: Ninety-five patients consecutively diagnosed as having symptomaticcarotid stenosis >70% were included over a 4 year period up to 2006. Thetherapeutic decision was taken collectively by a multidisciplinary team from theneurology, neurosurgery and neuroradiology departments. Vascular risk factors,neurological examination, neuroimaging findings, carotid and transcranial colourduplex study, intra- and post-operative complications (local and systemic), strokerecurrence and mortality at 3 months were recorded. Patients were admitted tothe neurology department both on the occurrence of stroke and for immediatepostoperative attention.Results: Seventy-six percent of patients were men, mean age 70±8.5 years. Hyper-tension had a frequency of 72.6%; hypercholesterolemia, 51.6%; diabetes, 31.6%;smoking, 29.5%; and ischaemic cardiac heart disease, 23.2%. Sixty percent had suf-fered a stroke and 40% a TIA. Forty percent of patients had a symptomatic stenosis>90% and 24% had a contralateral asymptomatic stenosis >70% or occlusion.The occurrence of stroke or death within 3 months of carotid endarterectomy was4.2%. Two patients died perioperatively (2.1%), one as a result of a hyperperfusionsyndrome and the other due to a brain infarction. Two patients (2.1%) sufferedstroke recurrence.Minor haemodynamic alterations were detected in 42.1% of patients during thefirst days after endarterectomy (hyper or hypotension) and 43.2% had localcomplications (XII or VII minor pareses).Conclusions: Among patients with severe carotid-artery stenosis and coexistingconditions, carotid endarterectomy in clinical practice, including high risk patients,has a low perioperative risk although minor complications are frequent, which maybe attributable to the close neurological control.


CAROTID ENDARTERECTOMY AUDIT OF GREAT BRITAIN & IRELANDA. Rudd, T. Lees, A. Halliday, P. Rothwell, A. Hoffman, D. KamugashaRoyal College of Physicians, London, United Kingdom

Background: A prospective two-year audit involving all hospitals that offerCarotid Endarterectomy (CEA), aiming to capture data on all CEA cases performedbetween Dec 05 & Dec 07, to assess quality of process of care & outcomes againstthe available evidence base. Main reporting spring 2008. A firm evidence basesupports the role of CEA & its urgency in the prevention of stroke. In the UK110,000 patients per annum suffer first stroke & 30,000 suffer TIAs. 10%-15% ofthe stroke patients should have CEA. The Healthcare Commission commissionedthe audit following a pilot funded by the Stroke Association.Methods: All surgeons who undertake CEA are eligible to participate and un-dertake to contribute to: Organisational Survey (2 rounds) describing individualsurgeon routine practice: investigations, case selection criteria, surgical technique& post-operative assessment and Clinical Audit to collect patient level data(indications, investigation, surgical technique & 30-day morbidity/mortality).Results: Organisational Survey (Round 1): Median number of CEAs performedper annum overall is 17 (IQR 10-25) - 90% of these done for symptomatic diseaseOver 70% of surgeons say they are able to see patients referred by letter within 2weeks99% of surgeons would not operate on symptomatic carotid stenoses <50% and53% would not operate <70%.62% of surgeons would be prepared to undertake CEA within 2 weeks followinga non-disabling stroke if the CT scan showed no infarct and 39% if the CT scanshowed a small infarctVascular surgeons perform nearly all CEA but about one quarter of surgeonsperform fewer than 10 cases per year. There are significant variations across thecountry in access to CEA services.Discussion: Recruitment is still open. 86% of eligible surgeons have registered andso far contributed 1300 cases for the Clinical Audit for which data collection willcontinue until end of December 2007. The first round of the Organisational Surveycompleted May 2006 and the second is currently Round 2 is underway will reportApril 2006


Venous diseases

1 Venous diseases

DEVELOPMENT OF A RISK SCORE TO PREDICT THE PROGNOSIS OFCEREBRAL VEIN AND DURAL SINUS THROMBOSIS (CVT)J.M. Ferro, T. Rodrigues, L. Bacelar-Nicolau, H. Bacelar-Nicolau, P. Canhão,I. Crassard, A. Dutra, A. Massaro, M.A. Mackowiak-Cordiolani, D. Leys, J.FontesDepartment of Neurosciences, Hospital de Santa Maria, Lisboa, Portugal

Background: Although cerebral vein and dural sinus thrombosis (CVT) has anoverall favourable prognosis, a variable proportion of patients die or becamedependent after CVT. It is relevant to identify such high-risk patients.Method: We used the ISCVT sample (624 patients) with a median follow-up timeof 478 days to develop a Cox proportional hazards regression model. Because ofnon proportional hazards, the used model was stratified by the median age of 37years. A treatment of influential observations (dfbeta analysis) led us to not include9 outlier subjects. The model was tested in the whole ISCVT sample and in twovalidation samples 1) the VENOPORT (91 patients), 2) of 169 consecutive CVTpatients admitted to 5 ISCVT participating centres after the end of the ISCVTrecruitment period.Results: In the ISCVT sample the model (HR - CNS infection 5.11; malignancy3.96; deep system 3.32; coma 3.17; mental 2.25; haemorrhage 1.57; male 1.76)accurately predicted 89% of good and 47% of bad outcomes (accuracy: 85%) at 6months, for a cut-off of 83% in the estimated survival probability. Area under theROC curve was 0.79 (p=0.000). In the VENOPORT validation sample the modelaccurately predicted 91% of good and 25% of bad outcomes (accuracy: 84%). Areaunder the ROC curve was 0.69 (p=0.077). In the 5 ISCVT centres validation samplethe model accurately predicted 93% of good and 24% of bad outcomes (accuracy:86%). Area under the ROC curve was 0.79 (p=0.000).Conclusion: The prognostic model presents a good external validity. The modelaccurately predicts the majority of favourable outcomes and 1/4 to 1/2 of un-favourable outcomes. It can be used to avoid dangerous interventions in low-riskpatients and to select patients for intensive monitoring and aggressive interventions.From hazard ratios a prognostic score and estimated survivor probability at 6months can be computed.

2 Venous diseases

DURAL ARTERIOVENOUS FISTULAS AND PREMATUREANTICOAGULATION CESSATION AFTER CEREBRAL VENOUSTHROMBOSISP. Cardona, H. Quesada, P. Sanchez, M.A. Fong, A. EscrigBellvitge Hospital, Hospitalet de Llobregat, Barcelona, Spain

Dural arteriovenous fistulas (DAVF) rarely are associated with cerebral venousthrombosis (CVT). We report five cases of symptomatic intracranial dural arteri-ovenous fistulas during follow-up of CVT.Methods: We retrospectively review forty patients with intracranial venous throm-bosis between 1996-2006. In five cases DAVFs were developed during follow-upperiod 1 year after anticoagulation stopping (after 6 -9 months of period treatment);leptomeningeal drainage were present in all the cases.Results: Symptoms as pulsatile tinnitus or headache appeared 3-12 months intervalafter anticoagulation cessation. All five patients where the initial angiogram studieshad showed abnormalities of the venous transverse or sigmoid sinuses, persistentabnormalities were seen on the later angio-MR previous to stopping oral anticoag-ulant (6-12 months period) as occluded or filiforme sinus. Two of five patients hadfactor V Leiden previously unknowned. Embolization of DAVF was performed inthree cases with good outcome.Discusion: DAVF appeared over previous ocluded or filiforme transverse sinusdemonstrated in angio-MR. All fistulaes were on the transverse or sigmoid sinuses.Itwas hypothesized that factor V Leiden and other inhereted deficiencies of coagula-tion factors, might be involved in the pathogenesis of DAVFS secondary to venousthrombosis predisposition over damaged venous wall. Also the anticoagulationcessation may predispose to DAVF formation.Conclusion: The longterm anticoagulant therapy in occluded or partial thrombosedsinus might be important for prevention of thrombosis and DAVF formationalthough patients were asymptomatic. Due to a potential risk of intracranial hemor-rages, embolisation previous to prompt anticoagulation may be developed in thesecases.

3 Venous diseases

INCIDENCE OF INHERITED THROMBOPHILIA IN GREEK PATIENTSWITH CEREBRAL VENOUS THROMBOSISK. Lysitsas, I. Gravas, G. Papaioannou, P. Kyriakidis, E. Dermitzakis, J. RudolfPapageorgiou General Hospital, Thessaloniki, Greece

Background: Hereditary thrombophilia has been reported to be present in approx-imately 30% of all patients with cerebral venous thrombosis (CVT). However, dataon the incidence of inherited thrombophilia in Greek CVT patients are scarce.Methods: We report the results of the diagnostic work-up including a full throm-bophilia screening in a consecutive case series of 27 patients (8 males, 19 females,age range 17 – 59 years) with CVT from a Greek tertiary healthcare facility.Results: Cephalalgia was the leading symptom in 85% of the patients (n=23),focal neurological signs were present in 48% (n=13), and epileptic seizures in 22%(n=6). Multiple thrombosis of cerebral sinus was a common finding in MRI andMRV: Thrombosis of the superior sagittal sinus was found in 78% (n=21), of thetransverse sinus in 41% (n=11), the sigmoid sinus in 7% (n=2), of the sinus rectusin 18% (n=5) and of the cavernous sinus in one patient only. Elevated D-dimerswere found in 48% (n=13, hyperhomocysteinaemia in 30% (n=8), heterozygousmutation of the MTHFR gene in 44% (n=12) and homozygous MTHFR mutationin 18% (n=5). Other hereditary thrombophilias (e.g. FV-Leiden mutation, n=1, orthe prothrombine G20210A mutation, n=2) were found in single cases only.Conclusion: In this consecutive open case series of Greek patients with CVT, theincidence of inherited thrombophilia was considerably higher than reported fromother comparable study populations.

Heart & brain

1 Heart & brain

DIAGNOSING PATENT FORAMEN OVALE(PFO) IN CRYPTOGENICSTROKE:TRANSCRANIAL DOPPLER VS TRANS OESOPHAGEAL ECHOS. Kumar, M.S. Randall, L.O.’ Toole, J.N. West, G.S. VenablesSheffield Teaching Hospitals NHS Trust, Sheffield, United Kingdom

Background:PFO is associated with cryptogenic stroke in young patients (<55 years). TransOesophageal Echo (TOE) has been the standard for diagnosing PFO. Contrast en-hanced Trans Cranial Doppler (TCD) is a simpler, easier & less invasive techniquethat detects a right to left shunt (RLS).The aim of this study was to compare theutility of the two techniques in the management of young patients with stroke.Methods:TCD & TOE were performed in 100 consecutive patients with ischaemic strokeor TIA (< 55 years; mean age 40years). Statistical analysis was performed usingSPSS software.Results:51% of patients had RLS on TCD and only 41% on TOE. After a positive resulton ce TCD 2 patients who had a negative TOE on the first occasion were shownto have a shunt on repeat TOE. Other structural abnormalities detected by TOEwere inter atrial septal aneurysm (19), aortic atheroma (7), atrial thrombus (2), leftventricular hypertrophy (2) & mitral valve abnormalities (3). Using TOE as thestandard, TCD sensitivity was 90% and specificity was 76%.With the combinationof the two tests as the standard, the sensitivity of TCD & TOE was 93% & 75%respectively. The negative predictive value of TCD was 92%,while that of TOEwas only 76%.McNemar’s test showed a significant difference between TCD &TOE(P=0.03).Discussion:This study reveals the added value of TCD in combination with TOE. The highersensitivity and negative predictive value of ceTCD may be due to an extracardiacshunt or inadequate valsalva during TOE. The size of the shunt on TCD mayassist in the risk assessment for stroke recurrence in young people with stroke.TOE is useful to exclude other sources of cardiac emboli. TCD has been shownto be reliable, more sensitive, less invasive and easy to use in a clinical settingmaking it the ideal screening tool. All young cryptogenic stroke patients shouldhave both TCD & TOE; undergo risk stratification based on degree of shunt onTCD, the presence of intra cardiac abnormalities and other concomitant risk factorsto facilitate appropriate management.

Poster SessionHeart & brain


2 Heart & brain

MODERATE HYPOTHERMIA FOR ANOXIC ENCEPHALOPATHY AFTERIN-HOSPITAL AND OUT-OF-HOSPITAL CARDIAC ARREST: RESULTS INROUTINE CLINICAL PRACTICER.A. Bernstein, K. Dewan, A.M. Naidech, M.J. Alberts, D. Fintel, D. Bergman,R. OakleyNorthwestern University, Chicago, IL, USA

Background and purpose: Randomized trials have shown that induced moderatehypothermia (MH) improves neurological outcomes after out-of-hospital cardiacarrests due to ventricular fibrillation/tachycardia (VF/VT). However, the effective-ness and safety of this treatment for other arrhythmias or for in-hospital arrests isless clear. In addition, the prognostic significance of early brainstem dysfunctionafter hypothermia is unknown. We devised a standard protocol for MH in patientswith anoxic encephalopathy after cardiac arrest of any type, including both inhospital and out-of hospital arrests.Methods: For this retrospective case series, we collected demographic, clinical,and outcome data on our first 21 consecutive patients treated with MH for anoxicencephalopathy after cardiac arrest. Target temperature was 33o C for 24 hoursfrom start of cooling, with controlled re-warming over 8 hours. Neurological out-comes were dichotomized based on discharge disposition as “good” (discharge tohome or rehabilitation) or “poor” (discharge to nursing home or death). Brainstemdysfunction was defined as any of the following: pupillary non-reactivity, absentcaloric or oculocephalic reflexes, absent corneal reflexes, or absent gag reflex.Results: Of the 21 patients who underwent MH after cardiac arrest, 62% weremale, and the mean age was 60 [range 35-88]. In hospital arrests (n=16, 76%) out-numbered out-of-hospital arrests (n=5, 24%). Arrest arrhythmias included pulselesselectrical activity (44%), asystole (24%), VF/VT (19%), primary respiratory arrestfollowed by cardiac arrest (10%) and unknown (5%). The mean time to returnof spontaneous circulation was 16 minutes [95% CI 9.7-22.7 min]. Dischargedispositions were to home (10%), rehabilitation (19%), nursing home (14%) anddead (57%). Good outcome (discharge to home or rehabilitation) occurred in 3 of5 (60%) of out-of-hospital arrests, and 3 out of 16 (18%) of in hospital arrests.Among patients surviving 3 or more days, all of those with brainstem dysfunctionhad poor outcomes (n=8); of those without brainstem dysfunction on day 3, 6 outof 7 (86%) had good outcomes.Conclusions: Moderate hypothermia after cardiac arrest is feasible in routine clin-ical practice. Good neurological outcome may be more common in out-of-hospitalarrests; only a randomized trial can determine if this therapy is effective forin-hospital arrests. Consistent with experience in the pre-hypothermia era, patientswith brainstem dysfunction on or beyond day 3 have poor outcomes.

3 Heart & brain

IS HEART DISEASE A PROGNOSTIC FACTOR FOR ACUTE STROKEOUTCOME? A PROSPECTIVE STUDYI. Ybot, M.J. Abenza, B. Fuentes, B. San José, M.A. Ortega-Casarrubios,P. Martínez, E. Díez-TejedorUniversity Hospital La Paz, UAM, Madrid, Spain

Background: Heart disease in ischemic stroke (IS) may be the cause of stroke,a coexistent illness, or even a consequence of stroke, but its presence means ahigher risk for vascular death. Objective: To analyse the presence of cardiopathy inpatients with acute IS and its impact on stroke outcome.Methods: Prospective study with inclusion of consecutive IS patients in a 4-monthrecruitment period. Previous or current cardiopathy, vascular risk factors, strokeseverity on admission, in-hospital complications and modified Rankin Scale (mRS)at discharge were analysed.Results: 91 patients included, 33% with known heart disease. Most frequententities were arrhythmia, including atrial fibrillation (AF) (53.3%) and ischemiccardiopathy (36.7%). They were older (72 vs 63 years old; p<0.05), had greaterfrequency of hypertension (80% vs 42%; p<0.05), hypercholesterolemia (60% vs19%; p<0.05) and peripheral artery disease (20% vs 4,9%; p<0.05), had moresevere strokes on admission (p<0.05) and worse outcome at discharge (mRS>2:48.1% vs 18.2%;p<0.05) than patients without previous history of heart disease.It was diagnosed cardiopathy in 11 among 61 patients without known heartdisease (18%), being AF the most frecuently diagnosis (6 patients). In the logisticregression analysis, the only independent factor of poor outcome was the strokeseverity on admission, without significant influence of heart disease.Discussion: Although previous cardiopathy seems to be associated to higher strokeseverity on admission and worse recovery at discharge, when adjusting for otherprognostic factors it was not independently associated to poor outcome.

4 Heart & brain

THE ROLE OF THE MORPHOLOGICAL CHARACTERISTICS OF PATENTFORAMEN OVALE IN CRYPTOGENIC STROKE: AN MRI STUDYC. Bonvin, K.O. Lovblad, H. Müller, R. SztajzelUniversity Hospitals of Geneva, Genève, Switzerland

Background and purpose: Patent foramen ovale (PFO) is an established cause ofcryptogenic stroke in young patients. The aim of our study was to evaluate, in pa-tients admitted for a cryptogenic stroke or transient ischemic attack (TIA), whetherthe number and distribution of ischemic lesions on MRI differed according to themorphological characteristics of the PFO including size and degree of interatrialright-to-left shunting (RLS) and presence of atrial septal aneurysm (ASA).Patients and methods: We included 220 consecutive patients less than 60 years oldadmitted from 2000 to 2006 for a cryptogenic stroke or TIA (absence of any otherdetermined stroke etiology following TOAST criteria after complete diagnosticworkup). Hypercoagulable state was not an exclusion criterion, since it may playa role in the paradoxical embolism. Demographic data have been analyzed fromthe patients’ personal records. MRI scans and echocardiographies were evaluatedby independent experienced investigators, blinded to the patients’ history. Twodifferent methods were systematically assessed to diagnose PFO and ASA: contrasttranscranial Doppler (c TCD) and transesophageal echocardiography (TEE) as wellas transthoracic echocardiography (TTE) in most patients.Results: Recruitment of patients is completed and neurologists, cardiologists andneuroradiologists are currently working intensively on the data. We will especiallydetermine (i) the prevalence of PFO and ASA in cryptogenic strokes, (ii) comparec-TCD, TEE and TTE methods in their ability to detect and quantify the PFO,(iii) correlate the number and size of MRI lesions with size of PFO, ASA anddegree of RLS in univariate and multivariate analysis (logistic-regression modeland ANCOVA). To our knowledge, this study is the first to compare simultaneouslyc-TCD, TEE and TTE findings with MRI lesions in patients with cryptogenicstroke. This study may help to better assess the risk of stroke in these patients andthus have critical impact on treatment options.

Large clinical trials (RCTs)

1 Large clinical trials (RCTs)

DELAYS IN TREATMENT FOR SYMPTOMATIC CAROTID STENOSIS ATRESEARCH ACTIVE CENTRESR.L Featherstone, J. Ederle, M.M. BrownUCL Institute of Neurology, London, United Kingdom

Background: Treatment of symptomatic carotid artery stenosis is an effectivesecondary prevention measure for stroke. The earlier endarterectomy is performedafter symptoms, the better the long-term outcome. We have used baseline data fromthe International Carotid Stenting Study (ICSS), an ongoing multicentre study ofsymptomatic patients randomized between carotid endarterectomy and stenting, toassess delays in treatment.Methods: The interval between the most recent TIA or non-disabling stroke,recorded at randomization, and the date of procedure (carotid endarterectomy orstenting) was calculated for all ICSS patients where data on the procedure wasreturned by December 2006. Data came from 36 centres in the UK, Europe, NorthAmerica and Australia.Results: The median delay between event and treatment was 55 days (n=854). Mostof the delay occurred before randomization, median delay between randomizationand treatment was 14 days. The three centres with the shortest average delaybetween event and treatment were compared with the three that had the longest.The median delay at the fastest centres was 14 days (N=42 patients) comparedto 123 days in the three slowest centres (N=72 patients), a significant difference(p<0.001).Discussion: Significant differences exist in treatment delays between centres. Eventhe most efficient research active centres are failing to treat many patients withsymptomatic carotid stenosis within 2 weeks of the presenting symptoms, whenthe benefit is greatest. Such treatment delays result in a substantial proportion ofpatients being left at high risk of a recurrent event while awaiting investigationand treatment. The results emphasise the need to reorganize stroke services toinvestigate and treat carotid stenosis urgently.



THE SAFETY AND EFFICACY OF CLOPIDOGREL VERSUS TICLOPIDINEIN JAPANESE STROKE PATIENTS — COMBINED RESULTS OF TWO PHASEIII MULTICENTRE RANDOMISED CLINICAL TRIALSS. Uchiyama, T. Yamaguchi, Y. FukuuchiTokyo Women’s University, Tokyo, Japan

Background: Compared with aspirin, both ticlopidine and clopidogrel have demon-strated superior efficacy in preventing recurrent vascular events, but adverse drugreactions (ADRs) are more common with ticlopidine. For the first time, the safetyand efficacy of these agents have been compared directly among Japanese strokepatients in a combined analysis of 2 phase III randomised, controlled studies.Methods: Patients aged 20–80 years with a history of noncardioembolic strokewere randomised to receive clopidogrel 75 mg once daily (n=941) or ticlopidine200 mg once daily (n=928) for 26 or 52 weeks. The primary endpoint wassafety (incidence of ADRs; n=1869). The major secondary endpoint was efficacy(combined incidence of cerebral infarction, myocardial infarction, and vasculardeath; n=1862).Results: Baseline characteristics of each group were similar (mean age, 64.7years; 71.3% males). The clopidogrel group had significantly fewer ADRs than theticlopidine group (35.0% vs 48.7%; p<0.0001) and a significantly lower incidenceof hepatic dysfunction (13.4% vs 25.6%; p<0.001). The frequency of hemorrhagicADRs was similar in the two treatment groups (clopidogrel group, 11.9%; ticlo-pidine group 10.1%; p=0.612). Both drugs were equally effective in preventingvascular events (clopidogrel group, 3.6%; ticlopidine group, 3.8%; p=0.510).Conclusions: In Japanese patients with a history of noncardioembolic stroke, bothclopidogrel and ticlopidine are equally effective in the prevention of recurrentvascular events.However, clopidogrel is safer than ticlopidine. Now that clopidogrel is availablein Japan, it should be considered as the preferred treatment choice, as occurs forWestern patients.


ETHNICITY DOES NOT AFFECT THE HOMOCYSTEINE-LOWERINGEFFECT OF VITAMIN THERAPY IN SINGAPOREAN STROKE PATIENTSK. Kasiman, J.W. Eikelboom, G.J. Hankey, H.M. Chang, M.C. Wong, C.P. ChenNational Neuroscience Insititute, Singapore General Hospital Campus,Singapore

Background: Increased plasma total homocysteine (tHcy) levels are a risk factorfor stroke and can be reduced with vitamin therapy. However, data on the tHcy-lowering effects of vitamins are limited largely to white populations. Thus, weaimed to determine in Singaporean patients with recent stroke: (1) the efficacyof vitamin therapy (folic acid, vitamin B12, and B6) on lowering tHcy, and (2)whether efficacy is modified by ethnicity (Chinese, Malay & Indian).Methods: A total of 506 eligible patients (420 Chinese, 41 Malays and 45 Indians)were recruited after presenting with ischemic stroke within the past 7 months.Patients were randomized to receive either placebo or vitamins as part of alarge multi-centre double-blinded clinical trial. Fasting blood samples collected atbaseline and at 1 year were assayed for levels of plasma tHcy.Results: Mean baseline tHcy was similar in the 2 groups, At 1 year, mean tHcywas significantly higher in the placebo group compared with the vitamin group.Ethnicity was not an independent determinant of tHcylevels at baseline. The magnitude of the reduction in tHcy levels at 1 year withvitamin therapy was similar, irrespective of ethnicity: mean change in tHcy Chinese(-3.2 vs 0.6 micromol/L); Malay (-3.5 vs 1.5 micromol/L) and Indians (-3.0 vs 0.2micromol/L).Discussion: Vitamin therapy reduces mean tHcy levels in the Singaporean strokepopulation studied. Ethnicity did not impact on the tHcy-lowering effect of vitaminsused in this study, despite possible differences in dietary intake and genetic makeup.This suggests the generalisability of vitamin therapy efficacy in lowering tHcyacross Asian populations.

Behavior and mood

1 Behavior and mood

SYMPTOMS AND DIAGNOSIS OF DEPRESSION IN APHASIC STROKEPATIENTSA.C. Laska, B. Mårtensson, T. Kahan, M. von Arbin, V. MurrayKarolinska Institutet, Department of Clinical Sciences, Stockholm, Sweden

Background: We investigated the feasibility of assessing depression, and symptomoccurrence, in patients with aphasia.Methods: 89 acute stroke patients with aphasia of all types were followed for sixmonths. The diagnosis of depression was made in accordance with DSM-IV criteriaat baseline, 1, 3, and 6 months. A standard aphasia test was performed. A batteryof “yes and no” capability questions from the comprehension part of the aphasiatest were selected for depression diagnostic purposes.Results: In 60 patients (67%) at baseline, and successively increasing to 100%at six months, comprehension allowed reliable DSM-IV diagnosis. The possibilityto undertake a DSM-IV interview was related to the degree of aphasia (p<0.01),and was least in patients with global and mixed non-fluent types of aphasia. Acomparison at one month between patients reliably fulfilling the criteria for adepression (D) and those who did not (Non-D) revealed: none of the two cardinalsymptoms occurred among the Non-D. Of the other symptoms, weight loss (36% inD, 16% in Non-D); insomnia (50% in D, 33% in Non-D); loss of energy (25% in D,20% in Non-D); and impaired concentration (27% in D, 19% in Non-D) occurred(all n.s.). At six months weight loss, insomnia, and loss of energy still occurred inabove 10% in Non-D. In all, criteria for depression were fulfilled in 24%.Discussion: It is possible to verify the presence or absence of a depression accord-ing to DSM-IV criteria in two thirds of aphasic stroke patients in the acute setting.Some depression symptoms occur irrespective of depression diagnosis. Hence, ifa cardinal symptom is fulfilled depression may possibly be over-diagnosed in theindividual stroke patient with aphasia.

2 Behavior and mood

SYMPTOM PROFILES IN MAJOR AND MINOR POSTSTROKE DEPRESSIONV. Murray, P. Gustavsson, B. MårtenssonKarolinska Institutet Danderyd Hospital, Dept. Clin. Sciences, Stockholm,Sweden

Background: Findings on clinical profiles in poststroke depression are conflict-ing. Meta-analysis is difficult due to different methodologies. Hence, new dataare needed. Symptom profiles in major and minor poststroke depression werecompared. For clarity, or basis for hypotheses, the symptom profiles of the majordepressed stroke patients were compared with those of major depressed psychiatricpatients.Methods: Stroke patients fulfilling DSM-IV diagnostic criteria for major or minordepression (n=127), and “psychiatric” in-patients with a major depression (n=40)were assessed by the Montgomery-Åsberg Depression Rating Scale (MADRS).Results: The MADRS profiles of major and minor depressed stroke patients weresimilar, with lower scores in minor depression but for “inner tension”; “pessimisticthoughts”; and “suicidal thoughts” where scoring was equally high. No basicclinical or neuroradiology differences were identified. Stroke patients with lesions

Poster SessionBehavior and mood


involving the cortex rated higher for “sadness”, while in those with a central lesiononly, “inner tension” (anxiety) was more pronounced. In the comparison of profilesbetween major depressed stroke patients and “psychiatric” patients, the items on“sadness” and “reduced sleep” were similarly pronounced.Discussion: Given the similarity in “sadness” i.e. depressed mood, equivalent tocardinal criterion 1 in the DSM classification of a depression, it is striking that“inability to feel”, the equivalent of the DSM cardinal criterion 2, was much lesspronounced in stroke than among the “psychiatric” patients.Conclusions: The difference between major and minor poststroke depression maybe more quantitative than qualitative, while the difference between poststroke and“psychiatric” major depression could be more qualitative.

3 Behavior and mood

ENDOREACTIVE POST-STROKE DEPRESSIONS ARE ASSOCIATED WITHPOORER MOTOR RECOVERYV. Kontzevoj, V. Skvortsova, M. Savina, E. PetrovaRussian Medical State University, Russian Federation

Background: The majority of recent studies showed that poststroke depression(PSD) influence the recovery of neurological deficit and daily activities. However,some studies didn’t take into account the clinical heterogeneity of PSD. Some ofPSD are known to have endogenic structure.Objective: We hypothesized that PSD with different psychopathological structurewould influence differently on motor recovery.Methods: 115 subjects with first stroke (57 males, 58 females, the mean age 65years) were observed in fixed terms. Depression was diagnosed using criteria ofICD-10. The elaborate psychopathological analysis of their clinical features wasmade. The degree of neurological impairment was assessed by the Orgogoso Scale.The recovery was assessed by criterion of Wilxoson.Results: During first year after stoke depressions were observed in 38 patients(33%). 6 cases with manifestation of depression before stroke were not includedin further analysis. In 21 patient were diagnosed reactive PSD. In 12 patientswere diagnosed endoreactive PSD that had both reactive and endogenic features(vitalized affects, circadian rhythmus with worsening of depressive symptoms atthe morning etc.). In patients without PSD (n = 77) Orgogozo scale total scorechanged significantly from 1–3 days to 2. week (p = 0,005), from 2. to 4. week (p= 0,000), from 4. week to 3 month (p = 0,000) and from 3. month to 6. month (p= 0,017); changes from 6 to 12 month after stroke were insignificant. In patientswith reactive PSD changes of Orgogozo scale total score were significant from 2.to 4. week (p = 0,001) and from 4. week to 3. month (p = 0,002). In patients withendoreactive PSD Orgogozo scale scores changed insignificantly in all defined timeintervals.Conclusions: Endoreactive PSD compared with reactive ones are associated withpoorer motor recovery.

4 Behavior and mood

NEUROPSYCHIATRIC PROFILE OF ELDERLY ACUTE STROKE PATIENTSC.O. Santos, L. Caeiro, J.M. Ferro, M.L. FigueiraServiço de Neurologia e Serviço de Psiquiatria, Department of Neurosciences,Hospital de Santa Maria, Lisboa, Portugal

Background: Neuropsychiatric disturbances after acute stroke are relatively fre-quent. Elderly stroke patients have a high proportion of concomitant diseases, aworse recovery and an aging brain. We aim to describe the neuropsychiatric profileof a sample of elderly acute stroke patients.Methods: Consecutive acute stroke patients (≤4 days after stroke onset) hospi-talised in a Stroke Unit were assessed with a standardized protocol including:MMSE, Delirium Rating Scale, Montgomery Asberg Depression Rating Scale, De-nial of Illness Scale, Catastrophic Reaction Scale, Mania Rating Scale, Apathy Scaleand Apathy Evaluation Scale. Neuropsychiatric profile of patients aged </≥65years old (younger vs elderly) was compared. Bivariate analysis was performed tofind associations between neuropsychiatric disturbances and demographic, clinicaland imaging data in the elderly patients.Results: We studied 55 elderly patients (mean age of 72.5 years old), 13 (24%)of them presenting an acute cognitive impairment, 7 (13%) delirium, 27 (49%)acute depression, 27 (49%) denial, 8 (15%) catastrophic reaction, 1 (2%) mania,9 (39%) were identified as clinically apathic and 8 (35%) considered themselvesas apathic. Elderly patients presented a higher frequency and severity of acutecognitive impairment (p=.01), a higher severity of delirium (p=.04) and catastrophicreaction (p=.02) and they considered themselves as more apathic (p=.02).Discussion: Although the frequency of neuropsychiatric disturbances was similarto that presented by younger patients, we found a higher severity of cognitive

impairment, delirium, catastrophic reaction and self-referred apathy in elderly. Theidentification of those neuropsychiatric disturbances is clinically relevant for theirfunctional state recovery.

Stroke and movements disorders

1 Stroke and movements disorders

MYOCLONUS AFTER ACUTE STROKEL. Idrovo Freire, F. Vivancos Matellanos, M. Lara Lara, E. Diez-TejedorHospital Universitario La Paz, Madrid-Spain

Introduction: Hyperkinetic abnormal movements during acute stroke are uncom-mon, with an estimated prevalence of 1%. Myoclonus is a clinical manifestationdefined as a sudden, brief, involuntary and shock-like movements caused bymuscular contraction (positive) or inhibitions (negative).Methods: We report 3 patients that on examination during the acute phase ofstroke showed hemi-asterixis (two of them) and a positive myoclonus in the other(videos).Results: The patients with asterixis had similar clinical features and both sufferedcardioembolic ischaemic infarcts of the posterior cerebral artery territory. Thepatient showing a positive myoclonus had a thalamo-mesencephalic haemorrhage.On all cases, neuroimaging studies revealed that the postero-lateral thalamus wasinvolved. The patients who showed asterixis, in addition to the thalamic compro-mise, also had temporo-occipital lesions (one of them the cerebellar hemisphere wasalso affected). On the other hand, in the patient with the haemorrhagic stroke therostral mesencephalus was also affected. In all cases, these abnormal movementshad a good outcome.Discussion: Hyperkinetic abnormal movements during stroke are unusual neurolog-ical manifestations and acute-onset hemi-asterixis is even less frequently reported.Asterixis is usually associated with thalamic lesions (ventral and posterolateral)though any lesion of the cerebellar-rubrothalamic-cortical pathway can enhancemyoclonic activity.

Intracranial aneurysms and vasospasm

1 Intracranial aneurysms and vasospasm

LOCAL LEVELS OF ENDOTHELIN-1 AND NITRIC OXIDE METABOLITESIN BASILAR ARTERY AND CEREBROSPINAL FLUID AFTEREXPERIMENTAL SUBARACHNOID HEMORRHAGE IN RABBITSV. Neuschmelting, S. Marbacher, A.R. Fathi, R.W. Seiler, S. Jakob, J. FandinoUniversity Hospital Berne, Berne, Switzerland

Objective: The genesis of Endothelin-1 (ET-1) and Nitric Oxide (NO) as twoimportant mediators in the development of cerebral vasospasm (CVS) after sub-arachnoid hemorrhage (SAH) is controversially discussed. The objective of thisstudy was to determine whether local levels of ET-1 and NO in cerebral arterialplasma and/or in cerebrospinal fluid (CSF) are associated with the occurrence ofCVS after SAH.Methods: CVS was induced using the one-hemorrhage-rabbit-model and confirmedby digital subtraction angiography of the rabbits’ basilar artery (BA) on day 5. Priorto sacrifice local samples of CSF and basilar arterial plasma (BAP) samples wereassessed by transclival approach to the BA in addition to systemic arterial plasma(SAP). ET-1 levels were determined by an immunometric technique (in pg/ml ±SEM) and total nitrate/nitrite level spectrophotometricly (in μmol/L ± SEM).Results: Angiographically detectable CVS could be documented in animals withinduced SAH (n=12, p<0.05). The ET-1 level in CSF was significantly elevated by27.3% to 0.84 ± 0.08 pg/ml in SAH animals (n=7) in comparison to control (0.66± 0.04 pg/ml, n=7, p<0.05). There was no significant difference of ET-1 levels inSAP and BAP samples of SAH animals compared to controls. Highly significantlack of local NO metabolites could be documented in BAP of SAH animals (36.8± 3.1 μmol/L, n=6) compared to controls (61.8 ± 6.2 μmol/L, n=6, p<0.01).Decreasing tendency of local NO level remained insignificant in CSF and SAP(n=6, p>0.05).Conclusions: This study demonstrates elevated ET-1 level in CSF and local lack ofNO in BAP samples to be associated with CVS after experimental SAH. Possiblegenesis of local changes of ET-1 and NO level after SAH are discussed in respectto controversial data reported to date.



SHARED EXPERIENCE IN THE ACUTE TREATMENT OF PATIENTS WITHCEREBRAL ANEURYSMM. Carvi y Nievas, J. Hattingen, H. Höllerhage, M. Müller-SchimpfleNeurosurgery, Frankfurt am Main/Höchst, Germany

Objective: evaluation of our shared endovascular-microsurgical experience treatingcerebral aneurysms.Method: An interventional neuroradiologist and a neurosurgeon administeredthe emergent treatment of 97 consecutive SAH-patients harboring 108 cerebralaneurysms. Decisions were based on the patient-clinical-condition, aneurysm mor-phology and location, degree of ICP and on CBF distribution patterns. The benefitsof joined case-assessment were retrospectively analyzed evaluating the rate ofaneurysm occlusion, the employment of complementary treatments, the patientclinical evolution and number of observed complications.Results: Initially, 39 aneurysms in 34 patients were endovascular (EV) and 65aneurysms in 63 patients were surgical (S) treated. Four non-ruptured aneurysmsremain under control. Three patients in the surgical group died without angio-control. Complete radiological aneurysm occlusion was achieved in 35 (89.5%)EV- and 60 (95.2%) S treated aneurysms. Complementary treatments included24 CSF-drainages and 7 decompressive surgeries in the EV-group as well as 5EV-procedures to treat severe vasospasm in the S-group. Four aneurysms-remnants(2 EV and 2 S-treated) remain unchanged. One EV-treated-aneurysm grew andwas surgically occluded. Favorable evolution was observed in 27 (79.4%) EV and52 (82.5%) S treated patients. CT-documented rebleedings (7 - 1), angiographicvasospasm (3 - 5), occlusion of main vessels (1 - 2) were the method relatedcomplications in EV and S-groups respectively.Conclusion: After comparing literature data, shared decisions in the emergentaneurysm-treatment increases the rate of aneurysm-occlusion, improves patient’sevolution and allows appropriated complementary treatments reducing the numberof complications.


NOREPINEPHRINE INDUCES DILATION IN THE RABBIT BASILARARTERY DUE TO ALPHA ADRENOCEPTOR DEPENDENT MECHANISMAFTER EXPERIMENTAL SUBARACHNOID HEMORRHAGE IN VIVOV. Neuschmelting, A.R. Fathi, S. Marbacher, R.W. Seiler, S. Jakob, J. FandinoUniversity Hospital Berne, Berne, Switzerland

Objective: Norepinephrine (NE) is routinely administrated for prevention andtreatment of cerebral vasospasm (CVS) after subarachnoid hemorrhage (SAH).The aim of this study was to determine mechanisms responsible for angiographicdilation and hypertension observed during continuous NE infusion in the rabbitbasilar artery (BA) after SAH.Methods: CVS was induced using the one-hemorrhage-model. On day 5 theanimals underwent control angiography prior to continuous intravenous adminis-tration of NE. Alpha-1 adrenoceptor antagonist (prazosine) and alpha-2 antagonist(rauwolscine) were added for partial inhibition. Changes in diameter of the BAwere digitally calculated in μm and expressed in percentages ± SEM. Prior tosacrifice, local samples of cerebrospinal fluid (CSF) and BA blood were obtainedby transclival approach. Endothelin-1 (ET-1) and nitric oxide (NO) levels weredetermined in random samples of both groups.Results: SAH induced CVS in the BA (-13.9% ± 2.0, n=36, p<0.0001). NEcaused hypertension from 83.2 ± 0.8 mmHg to 170.3 ± 0.9 mmHg (p<0.001).A dilation of 12.4% ± 2.6 (p<0.0001) of the BA during NE administration couldbe documented. Alpha-2 adrenoceptor inhibition partially reversed NE-dependentblood pressure plateau and significantly narrowed BA diameter by 11.3% ± 1.7(n=12, p<0.05). Additional alpha-1 inhibition instead showed similar antihyper-tensive effect while its narrowing effect on the dilated BA was less (-4.8% ±0.9, n=12) and remained insignificant (p>0.05). ET-1 and NO levels in CSF, BAand systemic plasma remained unchanged after NE administration and were notaffected by additional alpha antagonism (n=7 each, p>0.05).Conclusion: This study demonstrates the novel finding that NE causes dilation ofthe BA in the SAH rabbit model due to alpha adrenergic dependent mechanism,independently, however, from ET-1 and NO system.


INTEREST OF PERFUSION AND DIFFUSION MR IMAGING TO FOLLOWPATIENTS WITH CEREBRAL VASOSPASM AFTER ANEURYSMALSUBARACHNOID HEMORRHAGEE. Le Bars, H. Brunel, M. Moynier, G. Boubotte, A. BonaféCHU Hôpital Gui de Chauliac, Montpellier, France

Objective: study the potentiality of Diffusion and Perfusion MRI to improve thevasospasm diagnosis sensitivity in case of aneurysmal subarachnoid bleeding.Methods: Thirty cases of aneurysmal SAH were evaluated with TCD, DWI andPWI within the first three days and the following sixth and tenth day after thebleeding. The fourth MRI examination is done at 6 months to evaluate braindamages. For each patient, the apparent diffusion coefficient, the cerebral bloodvolume, the cerebral blood flow, the tissue mean transit time, the Time to Peak(TTP), the time inflow of contrast agent were evaluated for each exam. Twomethods for the evaluation of DWI and PWI analysis were carried out: a qualitativeanalysis for the thirty cases; a longitudinal quantitative analysis of PWI based ontwo groups of patients. The control group showed no modification of PWI duringthe study. In the other group variations of PWI time data outside the ischemic areawere found.Results: We found in two patients a complete reversibility in DWI anomalies.Three patients showed PWI anomalies without DWI modification. The amplitudeof relative perfusion time data at the acute stage of vasospasm is statisticallysignificant between the two groups. The evolution of relative perfusion time datafor the group with altered perfusion is statistically significant compared to thecontrol group. The relative TTP evolution is correlated with the clinical symptomsduring the acute stage of vasospasm, MRI lesion and with the neurological deficitsat 6 months. The longitudinal analysis of the rTTP value was the most sensitiveparameters witch was correlated with the deficit and with a risk of a lesion at sixmonth.Conclusion: The DWI and PWI appear to be sensitive imaging techniques forcerebral vasospasm evaluation. According to these preliminary results, perfusionappears to be an important tool for the evaluation of symptomatic or asymptomaticvasospasm and for the follow up of those patients.


PILOT STUDY OF NON-INVASIVE MEASURES OF ENDOTHELIALDYSFUNCTION IN ACUTE ISCHAEMIC STROKES.L. Soiza, I. Ford, H. Clark, M. Bruce, K.K. Kalal, D.J.P. WilliamsUniversity of Aberdeen, Aberdeen, United Kingdom

Introduction: Endothelial dysfunction (ED) is believed to be important in thepathogenesis of ischaemic stroke. Studies show serum markers of endothelialactivation are acutely raised after stroke. A direct, non-invasive measure of globalED has recently been developed.1 The method relies on pulse wave analysis(PWA) before and after administration of endothelium-dependent (salbutamol) andindependent (glyceryl trinitrate (GTN)) vasodilators. We believe this is the firststudy employing this technique in stroke patients.Methods: 29 patients with recent ischaemic stroke, 21 controls matched for riskfactors, and 9 healthy controls underwent PWA at the right radial artery usingSphygmoCor. ED was assessed by the ratio of the change in augmentation indexafter 400mcg inhaled salbutamol via spacer over the change after 400mcg ofsublingual GTN.1 Serum markers of ED (vonWillebrand factor, E-selectin andsVCAM-1) were obtained simultaneously. One-way ANOVA was used to look forsignificant differences between the groups.Results: See Table 1. Correlation between the various measures of ED was poor.

Table 1. Mean values of measures of endothelial function

Healthy Controls Matched Controls Acute Stroke P

Augmentation Index (AIx), % 9.7 32.7 32.6 <0.001AIx drop after Salb/GTN 0.57 0.23 0.30 0.02vonWillebrand Factor, U/ml 0.91 1.01 1.30 0.05E-selectin, ng/ml 36.2 50.9 49.9 0.43sVCAM-1, ng/ml 274.2 361.7 353.4 0.40

Conclusions: Patients who have suffered an acute ischaemic stroke have evidenceof endothelial dysfunction, but this was not significantly different from that foundin a population matched for risk factors for stroke.Reference: [1] Hayward et al. J Am Coll Cardiol 2002;40:521-528.

Poster SessionIntracranial aneurysms and vasospasm


Vascular biology

1 Vascular biology

POTENTIAL OF ERYTHROPOIETIN IN TREATMENT OF SPINAL CORDINJURY AND VASCULAR FAILUREI. Voznjuk, M. Odinak, N. TsyganMilitary Medical Academy, Russian Federation

Nowadays the use of cytokines that promote neuron survival and growth is apromising trend in the treatment of spinal cord injury and vascular failure. One ofthe few nerve growth factors which are used in clinical practice is erythropoietin.The purpose of the study was to evaluate the potential of erythropoietin in thetreatment of spinal cord injury and vascular failure.The experiment involved 48 male adult rats. The rats were subject to spinecompression at 20 N for 30 sec at level L3-L4. The animals were divided intothree groups: intact rats (8 animals); the rats that received sodium chloride (0.9%,0.5 ml) intraperitoneally 10 min after the trauma (20 animals); the rats thatreceived erythropoietin (5,000 units/kg) intraperitoneally 10 min after the trauma(20 animals). All the rats underwent daily neurological examination (the assessmentof hind limb and tail muscle strength, the pain sensitivity of the hind part of thebody, and pain reflexes of hind limb flexion and tail withdrawal). The results of theexamination were assigned numerical scores. Histological spinal cord examinationwas carried out on the 1st and 21st day after the trauma.The lifetime of the rats that received erythropoietin (4.63±1.69 days) was reliably(p<0.05) higher than in rats that received sodium chloride (2.75±1.28 days). On the11-16th day after the injury, the muscle strength in rats that received erythropoietinwas reliably (p<0.05) higher than in rats that received sodium chloride. The speedof the recovery of the muscle strength was reliably (p<0.025) higher in rats thatreceived erythropoietin. Histological examination showed a smaller number ofdamaged neurons and a smaller area of spinal hemorrhage in rats that receivederythropoietin.The data proves that erythropoietin increases the lifetime, extent and rate ofneurologic recovery after spinal cord injury in rats. This may be due to theneurotrophic and vascular-protective effect of erythropoietin. Thus, erythropoietinmay have high potential in the treatment of spinal cord injury and vascular failure.

2 Vascular biology

REDUCED ADAMTS-13 (VON WILLEBRAND FACTOR-CLEAVINGPROTEASE) ACTIVITY IN THE EARLY PHASE AFTER TIA OR ISCHAEMICSTROKED.J.H. McCabe, R. Starke, P. Harrison, P.S. Sidhu, M.M. Brown, S.J. Machin,I.J. MackieThe Adelaide and Meath Hospital, Trinity College Dublin, Dublin, Ireland

Background: Reduced ADAMTS-13 (von Willebrand factor-cleaving protease)enzyme activity is well described in patients with thrombotic thrombocytopenicpurpura (TTP), and may lead to the accumulation of very large von Willebrandfactor (VWF) multimers. Large VWF multimers may promote platelet activationand thrombus formation in vivo and could exacerbate ischaemia or infarction inpatients with TIA or ischaemic stroke who do not have TTP.Methods: Using a collagen binding assay, we performed a pilot study to measureADAMTS-13 activity in platelet poor plasma in 56 patients in the early phase (≤4weeks) and 46 patients in the late phase (≥3 months) after a TIA or ischaemicstroke while they were on treatment with aspirin (75-300 mg daily). We comparedthese data with those obtained from 22 controls subjects who were not on aspirin.The results were expressed in percentages relative to pooled normal plasma.Results: Mean ADAMTS-13 activity was significantly lower in the early phase(70.3%, P = 0.002) but not in the late phase (80.1%, P = 0.07) after TIA or strokecompared with controls (94.5%).Discussion: We have shown that ADAMTS-13 activity is significantly reduced inthe early phase after TIA or ischaemic stroke. Studies in larger cohorts of patientsare required to assess the importance of this finding, and further work is ongoingto assess the impact of reduced ADAMTS-13 activity on platelet function ex vivounder high shear stress conditions.

3 Vascular biology

IMPAIRED FLOW MEDIATED DILATATION IS ASSOCIATED WITH POOROUTCOME IN ISCHEMIC STROKED. Santos, M. Blanco, N. Perez de la Ossa, S. Arias, J. Serena,X. Rodriguez-Osorio, F. Nombela, M. Rodriguez-Yañez, R. Leira, A. DavalosHospital Clinico, University of Santiago de Compostela, Santiago deCompostela, Spain

Background: Brachial arterial flow-mediated dilatation (FMD) reflects endothelium-dependent vasodilator function. FMD is diminished in patients with atherosclerosis,is a marker of low nitric oxide bioavailability, and is associated with an increasedrisk of vascular or cardiac events. Our aim was to investigate the relationshipbetween FMD and outcome in patients with acute ischemic stroke.Methods: In 120 consecutive patients (58.3% male, median age 73 years) withacute ischemic stroke within the first 24 hours of evolution we measured FMD byhigh-resolution ultrasonography. FMD was calculated as the relationship betweenbasal diameter of the brachial artery before (d1) and after (d2) transient vascular oc-clusion (300 mmHg for 4 minutes) with a sphygmomanometer (FMD= d2 – d1/d1).The intima-media thickness (IMT) >0.9 mm, extracranial carotid atherosclerosis,stroke severity (NIHSS score) at baseline and discharge, and modified Rankin Scale(mRS) at 3 months were also evaluated. Poor outcome was defined as mRS >2.FMD was categorized according to ROC analysis.Results: Median [quartiles] FMD was 8.9 [4.3, 13.9]. Median FMD values weresignificantly lower in patients with IMT >0.9 mm (p<0.0001), and extracranialcarotid atherosclerosis (p<0.0001). FMD negatively correlated to stroke severity,both at baseline (p=0.038) and discharge (p=0.034). Median FMD was significantlylower (4.5 [2.3, 10.3] vs 9.4 [5.6, 15.1], p=0.003) in patients with poor outcome(n=38). The adjusted odds ratio of poor outcome for FMD >4.5% was 9.69 (1.97,47.68; p=0.005).Conclusions: Impaired FMD in patients with acute ischemic stroke is associatedwith poor outcome.

4 Vascular biology

A SYSTEMATIC ASSESSMENT OF THE GENETIC INFLUENCES ONCAROTID INTIMA MEDIA THICKNESS (CIMT)L. Paternoster, N. Martínez-González, M. Chung, R. Charleton, S. Lewis,C. SudlowUniversity of Edinburgh, Division of Clinical Neuroscience, Edinburgh, UnitedKingdom

Background: CIMT is a measure of subclinical atherosclerosis, associated withincreased risk of stroke and myocardial infarction, and a heritability of around 50%.It should be informative in studying the genetics of vascular disease, particularlylarge artery ischaemic stroke.Studies of the association between various genes and CIMT have produced con-flicting results. We aimed to identify genes whose association with CIMT has beenstudied in >5000 subjects, and to perform meta-analyses to evaluate reliably theevidence for an association.Methods: For each relevant study, we extracted information on subjects, methodsand mean (&SD) CIMT per genotype. We calculated study-specific and pooledmean difference in CIMT between genotypes.Results: 8 genes were studied in >5000 subjects: angiotensin converting enzyme(ACE); apolipoprotein E (APOE); beta 2 adrenergic receptor; methylenetetrahy-drofolate reductase; endothelial nitric oxide synthase; factor V; interleukin 6;paraoxonase 1.Several relevant studies (accounting for 19% of subjects studied across all genes)had insufficient published data for inclusion.2 genes (ACE and APOE) showed a significant association with CIMT. The DDgenotype of ACE had a mean CIMT 0.02mm greater than the II genotype. e4allele-containing genotypes of APOE had a mean CIMT 0.07mm greater than e2allele-containing genotypes. For both genes, we found larger associations amongsmaller studies, Asian subjects and subjects at high vascular risk.Discussion: We have identified 2 genes likely to influence CIMT, but methodolog-ical issues such as small study bias and missing data make it difficult to estimatethe true size of the associations. To increase the reliability of our results, we areseeking additional data from studies with insufficient published data.


5 Vascular biology

THE EFFECT OF ACUTE HYPERHOMOCYSTEINAEMIA ON CEREBRALBLOOD FLOW OF HEALTHY ELDERLY VOLUNTEERSS.R. Hart, A.A. Mangoni, C. Swift, C. Deane, R. Sherwood, A. Wierzbicki,S.H. JacksonDiv. of Clinical Neuroscience, University of Edinburgh, United Kingdom

Background: Mildly increased plasma homocysteine is an independent risk factorfor ischaemic stroke. However, the underlying fundamental causal arterial mecha-nisms in vivo linking hyperhomocysteinaemia with cerebrovascular disease remainunclear.Objective: To test the hypothesis that acute increases in plasma homocysteineproduced by methionine are associated with an acute decrease in cerebral arterialblood flow velocity (CABFV) measured by transcranial Doppler (TCD) ultrasound.By contrast, the simultaneous response of peripheral arterial distensibility wasmeasured by pulse wave velocity (PWV) and digital volume pulse (DVP).DESIGN: A double-blind, cross-over, placebo controlled design was used andcerebral blood flow velocity and peripheral arterial distensibility and plasma homo-cysteine concentrations were measured between 12 and 20 hours after methionineloading or placebo.Results: Between 13 and 16 hours after initial exposure to a methionine loadingtest, mean CABFV showed a significant 5.1% decrease in mean blood flow velocity(34.1±0.3 m/s vs 36.0±0.3 m/s, p <0.01) compared to placebo (Sample size = 8).However, between 17 and 20 hours after methionine exposure, CABFV showed nosignificant sustained change, compared to placebo (36.1±1.0 m/s vs 35.3±0.3 m/s,p < 0.1). There was no significant change in peripheral arterial distensibility mea-sured by PWV during hyperhomocysteinaemia compared to placebo (9.9±0.2m/svs 10.1±0.2m/s, p<0.5) and no difference in DVP, stiffness index (83.7±1.8% vs83.7±1.6%, p<0.1).Conclusion: In healthy elderly volunteers, acute hyperhomocysteinaemia resultedin a significant initial decrease in CABFV but no sustained reduction in cerebralblood flow velocity. There was no significant simultaneous change in periph-eral arterial distensibility suggesting that elderly cerebral arterial response tohyperhomocysteinaemia is different to that of peripheral arteries.

6 Vascular biology

SUBMICROSCOPIC FEATURES OF SMALL VESSEL DISEASE IN SKINBIOPSIES OF PATIENTS WITH CHRONIC KIDNEY DISEASE ANDEARLY-ONSET (<50 YEARS) COGNITIVE IMPAIRMENT. PRELIMINARIESRESULTG. Arismendi-Morillo, M. Fernandez-Abreu, A. Castellano-RamirezLaboratory of Electron Microscopy, University of Zulia. Nephrology andPathology Department HGS, Maracaibo, Venezuela

Background and aims: Decline in cognitive function has been reported in patientswith advanced renal disease. In addition, end-stage renal disease has been associatedwith accelerated vascular disease of the cerebral circulation. Cerebral small vesseldisease is frequent in patients with cognitive impairment. Skin biopsy is hire in thestudy of leukoaraiosis since permit establish the responsible vascular pathology ofpossible brain disease. The aim of this study was illustrate the small vessel diseasein skin biopsies of patients with chronic kidney disease and early-onset cognitiveimpairment in Maracaibo city - Venezuela.Patients and methods: Two female patients with chronic kidney disease and early-onset (< 50 years) of cognitive impairment that showed signs of Leukoaraiosis werestudied. Punch skin biopsy was prepared for conventional transmission electronmicroscopy study and for haematoxylin/eosin, PAS and Red Congo stain.Results: Small vessels study by means electron microscope revealed an increase inmedia-lumen ratio, endothelial cells with hyperplasic nucleus, clear cytoplasm andscarce organelles, thickened and multilayered basal membrane with focal degener-ative changes and deposition of amorphous and electron-dense materials as wellas proliferation of collagen fibers. Smooth muscular cells exhibited hypertrophy.Pericytes showed phagocytoced material and residual bodies. In adventitia wasthickened with abundant collagen fibers, amorphous and electron-dense materialsand cell debris.Conclusion: The morphological changes in subcutaneous small vessel correspondto small vessel disease of type degenerative microangiopathy and, possibly corre-spond to the microvascular pathology in the brain. Added patients with chronickidney disease and early-onset cognitive impairment are needed to establish acomplete characterization of small vessel disease.

7 Vascular biology

ASSOCIATION BETWEEN STROKE SUB-TYPES AND INTERLEUKIN-1GENE POLYMORPHISM WITHOUT SALIVA INTERLEUKIN-1 BETAIMPLICATIONM. Caillier, Y. Bejot, G.V. Osseby, F. Contegal, D. Minier,R.M. Gueant-Rodriguez, M. GiroudStroke registry of Dijon, Dijon, France

Background: Ischemia-induced inflammation is characterised by early infiltra-tion of leucocytes in the ischaemic region and development of brain oedema.Interleukin-1 (IL-1) is one of the key modulators of the inflammatory response.The IL-1 gene cluster on chromosome 2q14 contains three related genes (IL-1alpha, IL-1 beta and IL-1 receptor antagonist, IL-1 ra). Clinical studies suggest anearly intrathecal IL-1 beta production and IL-1 beta mRNA expression in bloodmononuclear cells during stroke. We aimed to investigate the association betweenischemic stroke sub-types and IL-1 gene polymorphism as well as production ofsaliva IL-1 beta.Methods: The -889C/T IL-1A, -511C/T IL-1B and IL-1RN (VNTR) polymorphismwas genotyped in patients with stroke due to large vessel disease (n=22), car-dioembolism (n=33), lacunar stroke (n=24), other determined mechanism (n=17),undetermined cause (n=21), transient ischemic attack (n=19) and in control group(n=19) by PCR. IL-1 beta concentration was determined in saliva using ELISA.Results: There was no significant rise in the concentration of salivary IL-1 betain acute stroke compared to the control group’s results. Studied polymorphismsdid not influence concentration levels. Genotypes frequency of IL-1A CT and TTwere significantly higher in lacunar stroke with respect to the control group (62,5%and 31,6%, respectively; p=0,04), but not T allele. For IL-1RN (VNTR), IL-1RN4-4 genotype frequency was higher in cardioembolic stroke than in control group(85,7% and 46,2%, respectively; p=0,04). This result was also confirmed for theallele 4 (35,4% and 21,1% respectively, p=0,02).Discussion: This study suggests than IL-1A and IL-1RN gene polymorphism isrelated with respectively lacunar and cardioembolic stroke onset.

Poster SessionVascular biology
