THE RESCUE BRAIN TRIAL: A FRENCH MULTICENTER RANDOMIZED TRIAL ON NEUROPROTECTION WITH LOWER LIMB ISCHEMIC PER-CONDITIONING IN THE ACUTE PHASE OF CEREBRAL INFARCTION

NMSI’s College Readiness Program is expanding the number of students participating and succeeding in advanced math, science and English courses. AP exam scores of 3 or higher (on a 5-point scale) qualify students for credit at many colleges and universities and are key indicators of college preparedness and success. ONE-YEAR PERCENT INCREASE IN. AP Calculus BC Saturday Study Session #1: The “Big” Theorems (EVT, IVT, MVT, FTC) (With special thanks to Lin McMullin) On the AP Calculus Exams, students should be able to apply the following “Big” theorems though students need not know the proof of these theorems. The Extreme Value Theorem (EVT).

F. Pico¹, B. Lapergue², M. Ferrigno³, M.L. Chadenat¹, F. Bourdain², E. Meseguer⁴, M. Obadia⁵, C. Hirel¹, D.L. Duong¹, S. Deltour⁶, C. Rosso⁶, P. Aegerter⁷, A. Cattenoy⁸, D. Smadja⁹, H. Hosseini¹⁰, B. Guillon¹¹, V. Wolff¹², Y. Samson⁶, C. Cordonnier³ and P. Amarenco⁴

¹Versailles Mignot Hospital, Neurology and Stroke Center, Versailles, France

²Hopital Foch, Neurology and Stroke Center, Suresnes, France

³Centre Hospitalier Universitaire Lille, Department of Neurology, Lille, France

⁴AP-HP- Bichat Hospital, Neurology and Stroke center, Paris, France

⁵Fondation Ophtalmologique Adolphe de Rothschild, Neurology and stroke center, Paris, France

⁶AP-HP- Pitie-Salpetriere Hospital, Urgences Cerebro-Vasculaire, Paris, France

⁷AP-HP- A Paré Hospital, Unité Recherche Clinique Paris Ouest, Boulogne-Billancourt, France

⁸Versailles Mignot Hospital, Délégation à la Recherche Clinique DRCI, Versailles, France

⁹Centre Hospitalier Sud-Francilien, Stroke Unit, Corbeil-Essonnes, France

¹⁰Henri Mondor Hospital, Stroke Center, Créteil, France

¹¹University Hospital of Nantes, Neurology, Nantes, France

¹²Strasbourg University Hospital, Stroke Unit, Strasbourg, France

Remote ischemic per-conditioning—causing transient limb ischemia to induce ischemic tolerance in other organs—reduces final infarct size in animal stroke models and seems to have a neuroprotective effect in patients with acute ischemic stroke. Our aim was to evaluate whether remote ischemic per-conditioning during acute ischemic stroke (<6 h) reduces brain infarct size at 24 h.

Patients have been included at eleven French Stroke centers, within 6 h of symptom onset with MRI-confirmed carotid ischemic stroke (NIHSS: 5–25). Patients have been randomized 1:1 to lower limb remote ischemic per-conditioning or sham group. Remote ischemic per-conditioning consists of four cycles of electronic tourniquet inflation (5 min) and deflation (5 min) to a thigh within 6 h of symptom onset. Randomization has been stratified on IV thrombolysis. Mechanical thrombectomy was allowed. MRI was repeated 24 h after stroke onset. The design has been published in Int J Of Stroke.

A total of 200 patients have been randomized (97 in the experimental arm and 103 in the sham arm). Mean age was 67.2 years (SD: 15.7), 52% were male, mean NIHSS score on admission was 11.3 (SD: 5.6), 87% received IV thrombolysis and 39% have been treated by mechanical thrombectomy. Main results will be presented at the session.

The RESCUE BRAIN trial will inform on whether or not remote ischemic per-conditioning reduces final infarct size in acute ischemic stroke patients with pharmacological and/or mechanical recanalization strategies.

The ClinicalTrials.gov Identifier is NCT02189928. Funded by the French Ministry of Health (PHRC IR 2014-A00104-43).

PARAMEDIC ACUTE STROKE TREATMENT ASSESSMENT (PASTA) TRIAL: MAIN RESULTS

C.I. Price¹, L. Shaw¹, P. Dodd², C. Exley³, D. Flynn⁴, R. Francis¹, S. Islam⁵, M. Javanbakht³, R. Lakey⁶, J. Lally³, G. McClelland⁷, P. McMeekin⁸, H. Rodgers¹, H. Snooks⁵, L. Sutcliffe¹, P. Tyrell⁹, L. Vale³, A. Watkins⁵ and G.A. Ford¹⁰

¹Newcastle University, Stroke Research Group, Newcastle upon Tyne, United Kingdom

²Lay Investigator, Stroke Research Group, Newcastle upon Tyne, United Kingdom

³Newcastle University, Institute of Health and Society, Newcastle upon Tyne, United Kingdom

⁴Newcastle University, Insitute of Health and Society, Newcastle upon Tyne, United Kingdom

⁵Swansea University, College of Medicine, Swansea, United Kingdom

⁶Newcastle University, Newcastle Clinical Trials Unit, Newcastle upon Tyne, United Kingdom

⁷North East Ambulance Service, North East Ambulance Service, Newcastle upon Tyne, United Kingdom

⁸Northumbria University, Faculty of Health & Life Sciences, Newcastle upon Tyne, United Kingdom

⁹Manchester University, Stroke Medicine, Manchester, United Kingdom

¹⁰Oxford University, Medical Sciences Division, Oxford, United Kingdom

Despite evidence that intravenous thrombolysis (IVT) is an effective treatment for selected acute ischaemic stroke patients, large variations in delivery exist between stroke services. This study evaluated whether an enhanced care pathway delivered by trained paramedics (the Paramedic Acute Stroke Treatment Assessment (PASTA)) could increase the proportion of patients receiving IVT.

Study design: Multicentre cluster RCT.

Setting: Three regional NHS ambulance services and 15 hyperacute stroke units across England and Wales.

Randomisation: Ambulance stations within each region were the unit of randomisation; paramedics based at stations delivered the PASTA pathway (intervention) or continued with standard stroke care (control).

Study intervention: The PASTA pathway included additional pre-hospital information collection regarding IVT eligibility, systematic pre-notification, structured hospital handover; and assistance with simple tasks at hospital.

Study control: Standard stroke care.

Participants: Adults with confirmed stroke who were assessed by a study paramedic within 4 hours of onset.

Primary outcome: Proportion of participants receiving IVT.

Secondary outcomes include: Stroke severity and complications post IVT; mortality; dependency at discharge and 90 days; time intervals including time to scan and IVT.

Sample size: To detect a 10% difference in proportion of IVT treatments, it was estimated that 1297 participants were required. However, the final target was adjusted as some of the underlying assumptions altered during the study.

From December 2015 to July 2018, 1272 participants were enrolled. Final follow up data were collected in November 2018. The main results will be available for presentation at ESOC in 2019.

n/a

ISRCTN12418919

TAKING CHARGE AFTER STROKE: A RANDOMISED CONTROLLED TRIAL OF A PERSON-CENTRED, SELF-DIRECTED, REHABILITATION INTERVENTION IN COMMUNITY STROKE SURVIVORS

V. Fu¹, M. Weatherall², K. McPherson³, W. Taylor², A. McRae⁴, T. Thomson⁵, J. Gommans⁶, G. Green⁷, M. Harwood⁸, A. Ranta², C. Hanger⁹ and H. McNaughton¹

¹Medical Research Institute of New Zealand, Stroke Rehabilitation, Wellington, New Zealand

²University of Otago, Medicine, Wellington, New Zealand

³Health Research Council of New Zealand, Management, Wellington, New Zealand

⁴Auckland District Health Board, Physiotherapy, Auckland, New Zealand

⁵Hutt Valley District Health Board, Medicine, Lower Hutt, New Zealand

⁶Hawkes Bay District Health Board, Medicine, Hastings, New Zealand

⁷Counties Manukau District Health Board, Medicine, Auckland, New Zealand

⁸University of Auckland, Medicine, Auckland, New Zealand

⁹Canterbury District Health Board, Medicine, Christchurch, New Zealand

‘Take Charge’ is a novel, community-based, self-directed rehabilitation intervention which helps people with stroke take charge of their own recovery. Take Charge significantly improved independence and health-related quality of life in Māori and Pacific New Zealanders following stroke. We tested the same intervention in a larger study and in a broader population with stroke.

We randomised 400 people within 16 weeks of acute stroke, discharged to institution-free community living at seven centres in New Zealand to a single Take Charge session (TC1, n = 132), two Take Charge sessions 6 weeks apart (TC2, n = 138) or a control intervention (n = 130). The primary outcome was the Physical Component Summary (PCS) score of the Short Form 36 (SF-36) at 12 months following stroke, comparing any Take Charge intervention to control. The outcomes assessor was masked to allocation.

Twelve months following stroke, participants in either of the TC groups (ie TC1 + TC2) scored 2.9 (95% CI 0.95 to 4.9, p = 0.004) points higher (better) than control on the SF36 PCS and this difference remained significant when adjusted for pre-specified baseline variables. There was a dose effect with SF36 PCS scores increasing by 1.9 points (95% CI 0.8 to 3.1, p < 0.001) for each extra Take Charge session received. Some secondary outcomes were also significantly better for intervention groups than control at six and 12 months following stroke.

The Take Charge session – a simple, low cost, self-management intervention, improved health-related quality of life at 12 months following stroke.

ACTRN12615001163594

ROBOT ASSISTED TRAINING FOR THE UPPER LIMB AFTER STROKE (RATULS): MAIN RESULTS

H. Rodgers^1,2,3, H. Bosomworth¹, L. Shaw¹, N. Alvarado¹, D.L. Cohen⁴, J. Dawson⁵, C. Fernandez-Garcia¹, T. Finch¹, G.A. Ford⁶, R. Francis¹, D. Howel¹, N. Hughes⁷, H.I. Krebs⁸, C.I.M. Price^1,2, L. Ternent¹, D.L. Turner⁹, L. Vale¹, F. van Wijck¹⁰, S. Wilkes¹¹, N. Wilson¹; on behalf of the RATULS investigators

¹Newcastle University, Stroke Research Group, Newcastle upon Tyne, United Kingdom

²Northumbria Healthcare NHS Foundation Trust

³Newcastle Hospitals NHS Foundation Trust

⁴London North West Healthcare NHS Trust

⁵University of Glasgow

⁶Oxford University

⁷NHS Greater Glasgow and Clyde

⁸Massachusetts Institute of Technology

⁹University of East London

¹⁰Glasgow Caledonian University

¹¹University of Sunderland

Robot-assisted training is a promising treatment for upper limb recovery following stroke. The RATULS trial is the largest trial to date to determine whether robot-assisted training improves upper limb function post-stroke.

Study design: A pragmatic multicentre randomised controlled trial, cost analysis and process evaluation.

Study setting: Four study centres, each consisting of a hub hospital with an InMotion robotic gym system and adjacent primary and secondary care spoke sites.

Study participants: Adults with acute or chronic first ever stroke (1 week to 5 years post-stroke) causing moderate to severe upper limb functional limitation.

Study treatments: There are three randomisation groups:

i. Robot-assisted training using the InMotion robotic gym system delivered for 45 minutes, three times per week for 12 weeks.

ii. Enhanced upper limb therapy delivered for 45 minutes, three times per week for 12 weeks.

iii. Usual NHS care.

Randomisation: Central independent web based service.

Primary outcome: Upper limb function measured by the Action Research Arm Test at 3 months.

Secondary outcomes: Upper limb impairment, activities of daily living, quality of life, resource use and adverse events measured at 3 and 6 months.

Blinding: Outcomes assessments by blinded assessor.

Parallel process evaluation: Semi-structured interviews with a sub-sample of participants and staff.

Sample size: 762 participants.

From April 2014 to April 2018, 770 participants from 4 study centres were randomised. Intervention delivery was completed in July 2018 and final follow up data were collected in October 2018. The main results will be available for presentation at ESOC in 2019.

N/A

ISRCTN69371850

COMBINED USE OF CONTACT ASPIRATION AND THE STENT RETRIEVER TECHNIQUE VERSUS STENT RETRIEVER ALONE FOR RECANALIZATION IN ACUTE CEREBRAL INFARCTION: THE RANDOMIZED ASTER2 TRIAL

B. Lapergue¹, J. Labreuche², M. Piotin³; On behalf of the ASTER2 Trial Investigators

¹University of Versailles and Saint Quentin en Yvelines- Foch Hospital, Department of Stroke Center, suresnes, France

²Univ. Lille- CHU Lille- EA 2694 – Santé publique: Epidémiologie et Qualité des Soins- F-59000, Department of Biostatistics, Lille, France

³Rothschild Foundation- Paris, Department of Diagnostic and Interventional Neuroradiology, paris, France

Mechanical thrombectomy (MT) with a stent retriever (SR) device is now the standard intervention in anterior circulation ischemic stroke with large vessel occlusion. Combined Contact aspiration (CA) with SR is a new promising treatment.

We aim to ascertain whether the combined use of CA and SR is more efficient than SR alone as a first-line endovascular treatment.

Methods: With a two-sided test (alpha = 5%, power = 80%) and an anticipated rate of spontaneous recanalization and catheterization failures of 20%, we estimated that a sample size of 408 patients would be necessary to detect an absolute difference of 15% in primary outcome (superiority design).

The ASTER2 trial is a prospective, randomized, multicenter, open-label, blinded end-point clinical trial.

Patients admitted with suspected anterior circulation ischemic stroke secondary to large vessel occlusion, with onset of symptoms < 8 hours, were randomly assigned to treatment with combined CA and SR or SR alone in a 1:1 ratio.

The primary outcome is the rate of perfect reperfusion (modified Thrombolysis In Cerebral Infarction [mTICI] score 2c/3) at the end of the endovascular procedures. Major secondary outcomes include procedural outcomes, intracerebral hemorrhage at 24-hours, the modified Rankin Scale and all-cause mortality at 90-days.

No previous head-to-head randomized trials have directly compared the efficacy of combined use of CA and SR versus SR alone.

This prospective trial aims to provide further evidence of the synergistic effects of CA and SR devices in first-line endovascular treatment.

The final results of the ASTER2 Trial will be presented at the ESOC 2019.

Identifier NCT03290885

BURDEN OF STROKE IN EUROPE: 30-YEAR PROJECTIONS OF INCIDENCE, PREVALENCE, DEATHS AND DALYS IN EU COUNTRIES

H. Wafa¹, E. Emmet¹, C. Wolfe^1,2,3 and Y. Wang^1,2,3

¹King’s College London, School of Populaiton Health and Environmental Sciences, London, United Kingdom

²National Institute for Health Research NIHR, Collaboration for Leadership in Applied Health Research and Care CLAHRC South London, London, United Kingdom

³National Institute for Health Research NIHR Biomedical Research Centre, Guy’s and St Thomas’ NHS Foundation Trust and King’s College London, London, United Kingdom

To estimate 30-year projections of stroke incidence, prevalence, deaths, and disability-adjusted life years (DALYs) in the European Union (EU).

Data from the global burden of disease study between 1990 and 2017 were analysed. Separate projection models by age and sex groups were developed for 28 EU countries to estimate trends in incidence, prevalence, deaths, and DALYs rates. The Akaike information criterion (AIC) suggested exponential regressions to provide the best fit. Adjustments were made for two distal predictors: time, capturing medical advances effect; and gross domestic product per capita (GDP), indicating the impact of economic development. Future trajectories were based on the United Nations’ population projections and the World Bank’s GDP prospects.

In 2017, 1·12 million cases of incident stroke took place in the EU, 9·53 million prevalent cases, 0·46 million deaths, and 7·06 million DALYs lost. By 2047, there will be an additional 40,000 incident cases and 2·58 million prevalent cases. Conversely, there will be 80,000 fewer deaths and 2·31 million fewer DALYs. Largest increase in the age-adjusted incidence and prevalence rates are expected in Lithuania (14·4% and 21% respectively), whereas greatest reductions are estimated in Portugal (–47·2% and –39·4%). Change in mortality rates will range from –2·5% (Lithuania) to –85·9% (Estonia), and DALYs’ from –7% (Romania) to –83·5% (Estonia).

The number of people with stroke is estimated to increase by 27% between 2017 and 2047 in the EU. Improved life expectancy and prevention strategies will result in increased prevalence/burden of stroke in Europe, particularly in Eastern European countries.

N/A

THE ENHANCED CONTROL OF HYPERTENSION AND THROMBOLYSIS STROKE STUDY (ENCHANTED): THE INTERACTION BETWEEN ALTEPLASE DOSE AND BLOOD PRESSURE LOWERING TREATMENT

X. Wang¹, J. Chalmers¹, T. Robinson² and C. Anderson¹

¹The George Institute for Global Health, Neurological and Mental Health department, Sydney, Australia

²University of Leicester, Department of Cardiovascular Sciences, Leicester, United Kingdom

High blood pressure (BP) is present in 75% of patients with acute stroke, but there has been longstanding uncertainty over how best to treat high BP during the acute period after acute ischemic stroke (AIS) when thrombolysis treatment is being considered. The Enhanced Control of Hypertension and Thrombolysis Stroke Study (ENCHANTED) study is an international, quasi-factorial randomized controlled trial involving a package of 2 linked comparative treatment arms, the first of which was to assess the effects of low-dose compared with standard-dose alteplase, while the second arm was to examine the effects of early intensive BP lowering. Post-hoc analyses from alteplase dose arm shows that patients with lower baseline BP benefits more from low-dose alteplase. We aim to assess the relation of randomized BP lowering treatment effects by altepalse dose interaction.

The ENCHANTED BP arm compared a strategy of intensive (systolic BP [sbp] 130–140 mmHg) versus guideline-recommended (SBP < 180 mmHg) BP lowering in 2227 thrombolysis-treated AIS patients for superiority with respect to shift (‘improvement’) in 90-day modified Rankin score and the safety outcome of any sICH. The main results will be presented at the International Stroke Conference in February 2019. Logistic regression will be performed for analysis of patient subgroups defined by altepalse dose (low vs. standard) on efficacy and safety outcomes. The heterogeneity of BP lowering treatment effect between subgroups will be estimated by adding an interaction term to the statistical model.

Data will be presented on how altepalse dose modifies differences in effects of intensive- versus standard-BP lowering treatment.

N/A

N/A

INTERACT2/ATACH-II IPD POOLING PROJECT TO DEFINE OPTIMAL LEVELS OF SYSTOLIC BLOOD PRESSURE CONTROL IN ACUTE INTRACEREBRAL HEMORRHAGE

T. Moullaali¹, R. Martin², X. Wang³, V. Shipes², T. Robinson⁴, J. Chalmers³, J. Suarez⁵, A. Qureshi⁶, Y. Palesch² and C. Anderson³

¹University of Edinburgh, Centre for Clinical Brain Sciences, Edinburgh, United Kingdom

²Medical University of South Carolina, Department of Public Health Sciences, Charleston, USA

³The George Institute for Global Health, Neurological and Mental Health Division, Sydney, Australia

⁴University of Leicester, Department of Cardiovascular Sciences, Leicester, United Kingdom

⁵The Johns Hopkins University School of Medicine, Departments of Anesthesiology and Critical Care Medicine- Neurology- and Neurosurgery, Baltimore, USA

⁶University of Missouri, Department of Neurology, Columbia, USA

Uncertainty persists over the optimum approaches to early systolic blood pressure (SBP) control in spontaneous intracerebral hemorrhage (ICH).

Three post-randomization SBP parameters were analysed in generalized linear mixed models using pooled individual participant data from the two largest randomised controlled trials (INTERACT2 [NCT00716079]/ATACH-II [NCT01176565]) of intensive BP lowering in acute ICH: ‘magnitude’, reduction within 1 hr; and ‘achieved’ (mean) and ‘variability’ (standard deviation) at shared time-points over 1–24 hrs. Primary outcome was functional recovery (90-day ordinal mRS scores). Safety outcomes were neurologic deterioration and symptomatic hypotension within 24 hours, and serious adverse events within 90-days. Data presented are with adjusted odds ratios (aOR) and 95% confidence intervals (CI).

Among 3,829 participants (mean age 63 years, 37% female, 65% Asian) in these analyses, achieved SBP was associated with the primary outcome (aOR per 10 mmHg increase: 0.90, 95% CI 0.87–0.94, p < 0.0001). Positive linear associations existed for levels of achieved SBP and SBP variability with neurologic deterioration, death, and any serious adverse event. Symptomatic hypotension and renal serious adverse events were uncommon (<1%). Trends toward unfavourable outcomes were evident with a large reduction in SBP (≥60 mmHg) within one hour.

Our unified analytical approach indicates that achieving and sustaining lower SBP are associated with favorable outcomes in acute ICH.

INTERACT2 [NCT00716079]/ATACH-II [NCT01176565]

THE TRANEXAMIC ACID FOR INTRACEREBRAL HAEMORRHAGE-2 (TICH-2) TRIAL: RESULTS OF ONE YEAR FOLLOW UP DATA

Z.K. Law^1,2, P. Scutt¹, L. Woodhouse¹, R. Gray¹, J. Appleton¹, C. Lysons¹, N. Frowd¹, P.M. Bath¹ and N. Sprigg¹; the TICH-2 investigators

¹University of Nottingham, Stroke Trials Unit, Nottingham, United Kingdom

²UKM Medical Centre, Department of Medicine, Kuala Lumpur, Malaysia

The Tranexamic acid for IntraCerebral Haemorrhage-2 (TICH-2) trial reported no significant improvement in functional outcome (shift in modified Rankin Scale) at day 90 despite reduction in haematoma expansion and early death (day 7). However, significant recovery after stroke, particularly intracerebral haemorrhage may take more than 3 months. Here we will report the participant outcomes one year after stroke.

The TICH-2 was a prospective randomised controlled trial that tested the efficacy and safety of tranexamic acid in spontaneous intracerebral haemorrhage when given within 8 hours of onset. Centralised blinded telephone follow up was performed for patients from the United Kingdom at Day 365. Outcome assessments included modified Rankin Scale, Barthel index, Telephone Interview Cognitive Status-modified, EuroQoL-5D and Zung Depression Scale.

2325 patients were recruited into the trial (age 68.9 ± 13.8 years; 1301 male, 56%). 1910 participants (82.2%) were eligible for day 365 follow up. 120 patients (6.3%) were lost to follow up. There was no significant difference in proportion of death in tranexamic acid or placebo group (260, 28% vs 276, 29%; OR 0.82, 95% CI 0–65.05; p = 0.11). Tranexamic acid did not reduce the risk of poor functional outcome (modified Rankin Scale >3; OR 1.01, 95% CI 0–80.28; p = 0.95). However, Cox proportional hazard analysis revealed significant survival benefit in the tranexamic acid group (adjusted HR 0.83, 95% CI 0–70.99; p = 0.038; figure).

Tranexamic acid did not significantly improve functional outcome but improved survival at one year in patients with intracerebral haemorrhage.

EudraCT number 2012–004108-37

DISENTANGLING THE INFLUENCE OF ATRIAL SEPTAL ANEURYSM AND SHUNT SIZE ON THE RISK OF RECURRENT STROKE ASSOCIATED WITH PATENT FORAMEN OVALE

G. Turc¹, J.Y. Lee², C. Perrin³, J.K. Song², J.L. Mas¹; on behalf of the CLOSE and DEFENSE-PFO investigators

¹Hopital Sainte-Anne- Paris- France & Université Paris Descartes & INSERM U1266, Neurology, Paris, France

²Asan Medical Center- Seoul- Republic of Korea, Cardiology, Seoul, Republic of Korea

³Hopital Sainte-Anne- Paris- France & INSERM U1266, Neurology, Paris, France

In patients with patent foramen ovale (PFO)-associated stroke, presence of large interatrial shunt or atrial septal aneurysm (ASA) has been suggested to convey a high risk of stroke recurrence. We assessed the respective influence of shunt size and ASA status on stroke recurrence under medical therapy in patients with recent PFO-associated stroke without alternative cause.

We pooled individual patient data from two prospective observational studies and the medical arms of two randomized trials, in which shunt size and ASA status was assessed by independent reading of echocardiography images. Associations between PFO anatomical features and recurrent ischemic stroke were assessed by mixed effects Cox models.

Of 894 patients (mean age 45.4 years), 188 (21%) had ASA plus large PFO, 58 (6.5%) ASA without large PFO, 418 (46.8%) large PFO without ASA and 230 (25.7%) non-large PFO without ASA. Over a median follow-up of 4.0 (IQR = 2–8.7) years, 47 (5.3%) patients experienced a recurrent stroke. There was no evidence of a synergistic effect between ASA and shunt size regarding stroke recurrence (Pinteraction = 0.29). In a model accounting for age, hypertension, antithrombotic therapy and PFO anatomy, ASA was independently associated with recurrent stroke (adjusted HR = 3.56, 95% CI:1–94.52; P < 0.0001), while large shunt was not (adjusted HR = 1.62, 95% CI:0–75.50; P = 0.22).

In patients with PFO-associated stroke, ASA is a more important predictor of recurrent stroke than shunt size. These results can help to better identify those patients with a high risk of stroke recurrence under medical therapy, who may derive the most benefit from PFO closure.

NA

BLOOD PRESSURE OVER THE LIFE COURSE AND LATER-LIFE COGNITION IN BLACKS, HISPANICS, AND WHITES (BP-COG): POOLED COHORT ANALYSIS OF ARIC, CARDIA, CHS, FOS, AND NOMAS

D. Levine¹, A.L. Gross², E.M. Briceño³, N. Tilton¹, M. Kabeto¹, D. Okullo¹, S.M. Hingtgen¹, B.J. Giordani⁴, J.B. Sussman¹, R.A. Hayward¹, J.F. Burke⁵, J.J. Manly⁶, M.S.V. Elkind⁶, E. Kulick⁷, R.F. Gottesman⁸, Y. Yano⁹, D.J. Gaskin¹⁰, S. Sidney¹¹, K. Yaffe¹² and A.T. Galecki¹³

¹University of Michigan, Internal Medicine, Ann Arbor- MI, USA

²Johns Hopkins University, Epidemiology, Baltimore- MD, USA

³University of Michigan, Physical Medicine and Rehabilitation, Ann Arbor- MI, USA

⁴University of Michigan, Psychiatry, Ann Arbor- MI, USA

⁵University of Michigan, Neurology, Ann Arbor- MI, USA

⁶Columbia University, Neurology, New York- NY, USA

⁷Brown University, Institute at Brown for Environment & Society, Providence- RI, USA

⁸Johns Hopkins University, Neurology, Baltimore- MD, USA

⁹Duke University, Community and Family Medicine, Durham- NC, USA

¹⁰Johns Hopkins University, Health Policy and Management, Baltimore- MD, USA

¹¹Kaiser Permanente, Division of Research, Oakland- CA, USA

¹²University of California San Francisco, Psychiatry, San Francisco- CA, USA

¹³University of Michigan, Biostatistics, Ann Arbor- MI, USA

Older Blacks are 2 times more likely and older Hispanics are 1.5 times more likely than older non-Hispanic Whites to have late-life cognitive impairment and dementia. We hypothesized that higher blood pressure (BP) levels explain race-ethnic differences in cognitive decline.

Determine associations between BP over the life course and cognitive decline in blacks, Hispanics, and whites.

We pooled 20,385 participants 18 and older free of stroke and dementia across 5 longitudinal cohorts from 1971 to 2017: Atherosclerosis Risk in Communities Study (ARIC), Coronary Artery Risk Development in Young Adults Study (CARDIA), Cardiovascular Health Study (CHS), Framingham Offspring Study (FOS), and Northern Manhattan Study (NOMAS). The primary outcome was change in global cognitive performance (GCP) harmonized across all cohorts. Mean follow-up was 13.5 + 8.0 years.

Compared with whites, blacks had significantly faster, and Hispanics significantly slower, declines in GCP (Model A, Table 1). After adjusting for time-dependent cumulative mean systolic BP, black-white differences in cognitive decline attenuated, but Hispanic-white differences in cognitive decline did not (Model B). Further adjustment for antihypertensive medication use did not change the race-ethnic differences in cognitive decline (Model C).

Blacks’ higher BP levels and worse control over the life course contribute to black-white differences in cognitive decline. We found no evidence that BP disparities explain Hispanic-white differences in cognitive decline.

N/A

CEREBRAL MICROBLEEDS AND FUTURE STROKE RISK AFTER ISCHAEMIC STROKE OR TRANSIENT ISCHAEMIC ATTACK: POOLED ANALYSIS OF INDIVIDUAL PATIENT DATA FROM COHORT STUDIES

D. Werring¹; on behalf of The Microbleeds International Collaborative Network

¹PhD FRCP FESO, Stroke Research Centre- UCL Queen Square Institute of Neurology, London, United Kingdom

Cerebral microbleeds (CMBs) are a promising biomarker of increased intracranial haemorrhage risk after ischaemic stroke (IS) or TIA, but whether they might identify patients at risk of net harm from antithrombotic therapy is unknown.

In an individual patient data pooled analysis, we determined the association of CMBs with intracranial haemorrhage (ICH) and IS risk using multivariable Cox regression, adjusted for pre-specified prognostic variables. We developed and validated stroke risk prediction models including CMBs.

In 20,322 individual patients from 38 cohorts (with over 35,000 patient-years (median 1.98 years) follow-up), CMBs were associated with increased adjusted hazard ratios (HRs) (95% CI) of all-stroke (1.39 (1–24.56)), ICH (2.56 (1–89.47)), and IS (1.27 (1–10.45)). As CMB burden increased, the HR rose more steeply for ICH (for ≥5 CMBs, HR 4.86 (3–27.23)) than IS (for ≥5 CMBs, HR 1.53 (1–23.90)). CMB distribution had little effect on HRs. However, regardless of CMB burden or distribution, the absolute rate of IS exceeded that of ICH (e.g. for ≥10 CMBs, rate 64 (48-84) per 1000 patient-years vs. 27 (17-41) per 1000 patient-years, respectively). Findings were similar in patients taking antiplatelet drugs or oral anticoagulants. Risk prediction models including CMBs showed good discrimination for ICH (C-index 0.71 (0–62.79)) and IS (C-index 0.65 (0–59.69)).

CMBs are associated with a greater hazard ratio for ICH than IS; predictive models have good discrimination. However, the absolute risk of IS is consistently much higher than that of ICH, suggesting that CMBs do not identify patients at risk of net harm from antithrombotic use.

N/A

CILOSTAZOL VS ASPIRIN IN ISCHEMIC STROKE WITH CEREBRAL MICROBLEEDS VS PRIOR INTRACEREBRAL HEMORRHAGE: SUBGROUP ANALYSIS OF A RANDOMIZED CONTROLLED TRIAL

K.S. Hong¹, S.U. Kwon², H.K. Park¹, J.S. Lee³, B.J. Kim⁴, J.H. Park⁵, Y.J. Kim⁶, S. Yu⁷, Y.H. Hwang⁸, J.H. Rha⁹, S.H. Heo⁴, S.H. Ahn¹⁰, W.K. Seo¹¹, J.M. Park¹², J.H. Lee¹³, J.H. Kwon¹⁴, S.I. Sohn¹⁵ and J.M. Jung¹⁶

¹Ilsan Paik Hospital Inje Univeristy, Neurology, Goyang, Republic of Korea

²Asan Medical Center Ulsan University, Neurology, Seoul, Republic of Korea

³Asan Medical Center Ulsan University, Clinical Research Center, Seoul, Republic of Korea

⁴Kyung Hee University Medical Center, Neurology, Seoul, Republic of Korea

⁵Myongji Hospital- Hanyang University College of Medicine, Neurology, Goyang, Republic of Korea

⁶Ewha Woman’s University Hospital, Neurology, Seoul, Republic of Korea

⁷Anam Hospital Korea University, Neurology, Seoul, Republic of Korea

⁸Kyungpook National University Hospital, Neurology, Daegu, Republic of Korea

⁹Inha University Hospital, Neurology, Incheon, Republic of Korea

¹⁰Chosun University Hospital, Neurology, Gwangju, Republic of Korea

¹¹Samsung Medical Center Sunkyunkwan University, Neurology, Seoul, Republic of Korea

¹²Eulji General Hospital Eulji University, Neurology, Seoul, Republic of Korea

¹³Sacred Heart Hospital Hallym University, Neurology, Seoul, Republic of Korea

¹⁴Ulsan University Hospital, Neurology, Ulsan, Republic of Korea

¹⁵Dongsan Medical Center Keimyung University, Neurology, Daegu, Republic of Korea

¹⁶Ansan Hospital Korea University, Neurology, Ansan, Republic of Korea

In the PICASSO trial (ClinicalTrials.gov Identifier: NCT01013532) enrolling recent ischemic stroke patients with multiple cerebral microbleeds (CMBs) or prior intracerebral hemorrhage (ICH), cilostazol and aspirin were comparable for the co-primary efficacy and safety endpoints. We explored a treatment-by-underlying cerebral hemorrhage-prone lesion interaction.

This was a post hoc analysis of PICASSO. Patients were divided into the CMBs and prior ICH subgroups. The primary safety endpoint was the time to first occurrence of cerebral hemorrhage (ICH or subarachnoid hemorrhage). The primary efficacy endpoint was the time to first occurrence of composite of stroke, myocardial infarction, or vascular death.

Of 1512 patients, 903 (59.7%) had multiple CMBs and 609 (40.3%) had prior ICH. The risk of cerebral hemorrhage was lower with cilostazol vs aspirin (0.12%/year vs 1.49%/year; hazard ratio [HR], 0.08 [95% CI 0–01.60]; p = 0.015) in the CMBs subgroup, but was not different (1.26%/year vs 0.79%/year; HR, 1.60 [0–52.90]; p = 0.408) in the prior ICH subgroup. The interaction of treatment-by-hemorrhagic prone lesion was significant (Pinteraction = 0.011). For the composite of major vascular events, there was a trend toward a lower risk with cilostazol vs aspirin (3.56%/year vs 5.53%/year; HR, 0.64 [0–41.01]; p = 0.056) in the CMBs subgroup, but was comparable (5.21%/year vs 5.05%/year; HR, 1.03 [0–63.67]; p = 0.913) in the prior ICH subgroup. There was no significant interaction (Pinteraction = 0.165).

In ischemic stroke with multiple CMBs, cilostazol might be a better option over aspirin with a lower risk of cerebral hemorrhage and a trend of reduced risk of major vascular events.

ClinicalTrials.gov Identifier: NCT01013532

EFFECTS OF RIVAROXABAN VS ASPIRIN ON THE OCCURRENCE OF CLINICAL AND MRI-DEFINED INFARCTS AND MICROBLEEDS IN NAVIGATE ESUS – RESULTS OF NAVIGATE MIND MRI SUBSTUDY

M. Sharma¹, S. Kasner², B.W. Yoon³, E. Smith⁴, P. Sheridan⁵, S. Ameriso⁶, J. Fiebach⁷, J. Puig⁸, D. Damgaard⁹, K. Muir¹⁰, R. Veltkamp¹¹, D. Toni¹², R. Gagliardi¹³, N. Shamalov¹⁴, R. Mikulik¹⁵, D. Bereczki¹⁶, S. Engelter¹⁷, M. O’Donnell¹⁸, A. Shoamanesh¹, H. Mundl¹⁹ and R. Hart²⁰

¹McMaster University/Population Health Research Institute, Department of Medicine Neurology, Hamilton, Canada

²University of Pennsylvania, Department of Neurology, Philadelphia, USA

³Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Republic of Korea

⁴Cumming School of Medicine- University of Calgary, Department of Medicine Neurology, Calgary, Canada

⁵McMaster University/Population Health Research Institute, Statistics, Hamilton, Canada

⁶Institute for Neurological Research FLENI, Department of Neurology, Buenos Aires, Argentina

⁷Center for Stroke Research Berlin CSB- Charité – Universitätsmedizin Berlin Campus Benjamin Franklin, Dept. of Neurology, Berlin, Germany

⁸Girona Biomedical Research Institute IDIBGI- University of Girona, Department of Neurology, Girona, Spain

⁹Aarhus University Hospital, Department of Neurology, Aarhus, Denmark

¹⁰University of Glasgow- Queen Elizabeth University Hospital, Institute of Neuroscience and Psychology, Glasgow, United Kingdom

¹¹Imperial College London, Department of Medicine, London, United Kingdom

¹²“Sapienza” University of Rome, Department of Neurology and Psychiatry, Rome, Italy

¹³Irmandade da Santa Casa de Misericórdia de São Paulo, Department of Medicine, Sao Paulo, Brazil

¹⁴Pirogov Russian National Research Medical University, Department of Medicine, Moscow, Russia

¹⁵St. Anne’s University Hospital, International Clinical- Research Center and Neurology Department, Brno, Czech Republic

¹⁶Semmelweis University, Department of Neurology, Budapest, Hungary

¹⁷Universitätsspital Basel, Innere Medizin, Basel, Switzerland

¹⁸HRB-Clinical Research Facility- NUI Galway, Department of Neurology, Galway, Ireland

¹⁹Bayer AG, na, Wuppertal, Germany

²⁰Population Health Research Institute, na, Hamilton, Canada

NAVIGATE ESUS randomized participants with nonlacunar cryptogenic stroke to rivaroxaban 15 mg or aspirin 100 mg. We sought to determine the effect of treatment on the primary substudy outcome of the combination of incident clinical or covert infarcts.

Consenting NAVIGATE ESUS participants had a study specific MRI comprising T1, T2, FLAIR, T2* (SWI or GRE) sequences performed soon after randomization and at study termination provided at least 3 months had elapsed since randomization. Participants were recruited from 80 sites in 15 countries and imaged on 1.5T or 3T scanners. Incident infarcts and microbleeds were ascertained by trained readers blinded to treatment assignment and characterized by location and size. Covert infarcts were defined as infarcts on MRI in the absence of an adjudicated stroke between the baseline and end-study MRI. Treatment effect was determined by logistic regression using the intention to treat principle.

Baseline and follow up MRIs were obtained on 728 participants with a median (IQR) 336(182-533) days between scans. Mean age was 66.8 years and 60% were men with 58(25-119) days between the index stroke and randomization. At baseline, 631 (87%) had chronic infarcts and 51 (7%) had petechial hemorrhagic transformation of the index stroke. Incident clinical strokes occurred in 29 participants while 78 had an infarct detected on MRI. Final results will be presented.

NAVIGATE ESUS MIND provides insight into incidence and topography and response to treatment of MRI defined lesions in patients with ESUS.

NCT02313909

SPHENOPALATINE GANGLION STIMULATION IMPROVES OUTCOME FROM ACUTE ISCHEMIC STROKE IN A DOSE-DEPENDENT MANNER: FURTHER INSIGHTS FROM THE PIVOTAL IMPACT-24B TRIAL

J. Saver¹, N. Bornstein², H.C. Diener³, P. Gorelick⁴, A. Shuaib⁵, Y. Solberg⁶, M. Savic⁷, T. Janelidze⁸, N. Zarqua⁹, D. Yarnitsky¹⁰, C. Molina¹¹; for the ImpACT-24B Investigators

¹University of California- Los Angeles, Neurology, Los Angeles, USA

²Shaarei Zedek Medical Center, Neurology, Jerusalem, Israel

³University of Duisburg-Essen, Neurology, Essen, Germany

⁴Northwestern University, Neurology, Chicago, USA

⁵University of Alberta, Neurology, Edmonton, Canada

⁶BrainsGate Ltd, Medicine and Science, Caesarea, Israel

⁷Special Hospital for Cerebrovascular Diseases 'Sveti Sava', Neurology, Belgrade, Serbia

⁸Kutaisi Referral Hospital, Neurology, Kutaisi, Georgia

⁹Zugdidi Referral Hospital, Neurology, Zugdidi, Georgia

¹⁰Rambam Medical Center, Neurology, Haifa, Israel

¹¹Hospital Universitari Vall d’Hebron, Neurology, Barcelona, Spain

Stimulation of the sphenopalatine ganglion (SPG) increases cerebral collateral blood flow, stabilizes the blood-brain barrier, reduces infarct size in preclinical acute ischaemic stroke (AIS) models, and showed evidence of benefit in the co-primary confirmed-cortical-infarct (CCI) patient population in the recent pivotal ImpACT-24B trial. Dose-response analysis would provide further treatment effect confirmation and guide dose-selection for clinical practice.

ImpACT-24B was a randomized, sham-controlled, double-masked trial, performed at 73 centers in 18 countries. Patients with anterior circulation AIS, 8–24h post-onset, NIHSS 7–18, were randomized to active/sham pterygopalatine fossa neural stimulator placement, and active/sham stimulation 4h/d for 5d. Attained stimulation intensities varied, reflecting stimulator intensity profile and inter-individual differences in somatosensory sensitivity.

The confirmed-cortical-infarct co-primary population comprised 520 of 1000 enrolled (randomized 244 SPG-stimulation, 276 sham-control). Across all stimulation doses, benefit on the primary outcome, improvement beyond expectations on the 90d mRS, was 49.6% vs 39.9%, p = 0.0258. Dose-response analysis revealed a strong dose effect (p = 0.0006) with an inverted-U shape (medium-intensity better than low or high).(Figure) Among the 25% of patients receiving stimulation in the optimum (medium) intensity range, benefit was substantially magnified (68.9% vs 39.9%, p < 0.0001). Similar dose-response curves were seen for all 4 prespecified secondary efficacy endpoints: functional independence (mRS 0–2), self-care or better (mRS 0–3), improved stroke-specific-quality of life (SIS-16), and degree of disability (UW-mRS). (Table)

SPG stimulation improved the outcome of acute ischemic stroke patients with confirmed cortical infarction in a strongly dose-dependent manner. A simple stimulation protocol has been developed that reliably achieves optimum dose delivery for clinical implementation.

NCT00826059

VESSEL OCCLUSION DOES NOT MODIFY TREATMENT EFFECT OF INTRAVENOUS ALTEPLASE IN THE WAKE-UP TRIAL

J. Fiebach¹, I. Galinovic¹, F. Boutitie², K. Villringer¹, B. Cheng³, A. Wouters⁴, M. Ebinger⁵, M. Endres⁶, J. Fiehler⁷, V. Thijs⁸, R. Lemmens⁴, K.W. Muir⁹, N. Nighoghossian¹⁰, C.Z. Simonsen¹¹, C. Gerloff³, G. Thomalla³, S. Pedraza¹²; on behalf of the WAKE-UP investigators

¹Charité Universitätsmedizin Berlin, Center for Stroke Research – Neuroradiology, Berlin, Germany

²Hospices Civils de Lyon, Service de Biostatistique, Lyon, France

³University Medical Center Hamburg-Eppendorf, Neurology, Hamburg, Germany

⁴University Hospitals Leuven, Department of Neurology, Leuven, Belgium

⁵Medical Park Humboldtmühle, Neurology, berlin, Germany

⁶Charité Universitätsmedizin Berlin, Neurology, Berlin, Germany

⁷University Medical Center Hamburg-Eppendorf, Diagnostic and Interventional Neuroradiology, Hamburg, Germany

⁸The University of Melbourne, Florey Department of Neuroscience and Mental Health, Melbourne, Australia

⁹University of Glasgow, ICE & Queen Elizabeth University Hospital, Glasgow, United Kingdom

¹⁰Université Claude Bernard Lyon, Department of Stroke Medicine, Lyon, France

¹¹Aarhus University Hospital, Neurology, Aarhus, Denmark

¹²Hospital Universitari Dr. Josep Trueta, Institut d’investigació Biomèdica de Girona IDIBGI, Girona, Spain

Vessel occlusion (VO) of an intracranial artery is the therapeutic target for fibrinolysis. In this pre-planned subgroup analysis, we studied whether VO modified treatment effect of intravenous alteplase in the WAKE-UP trial.

In the WAKE-UP study, patients with DWI-FLAIR mismatch regardless of vessel occlusion were randomized to either alteplase or placebo treatment within 4.5 h of symptom recognition. VO was determined on baseline MR angiography (MRA). We compared clinical and imaging parameters as well as treatment effect on the primary endpoint (modified Rankin Scale 0–1 at 90 days) between patients with and without VO.

In the intention-to-treat population of 503 patients, 485 patients had MRA of diagnostic quality and 185 (32%) presented with VO at baseline. Strokes were more severe, median NIHSS 9 (IQR 5–15) versus 5 (IQR 3–7) and baseline DWI lesion volumes were bigger (mean volume 15 ml SD 17 ml versus 3 ml SD 7 ml) in the VO group. The intervals between symptom recognition and treatment did not differ between patients with and without VO. There was no interaction of VO and treatment effect (p = 0.56). Among patients with VO, 30/94 (31.9%) alteplase treated patients had good clinical outcome (mRS 0–1) as compared to 18/91 (19.8%) placebo treated patients (OR 2.0, 95% CI 1.0-4.2; p = 0.05).

The presence of a VO did not modify the treatment effect of intravenous alteplase in WAKE-UP. Although WAKE-UP was not powered to demonstrate treatment efficacy in subgroups, intravenous alteplase appears to result in better functional outcome in patients with VO than placebo.

ClinicalTrials.gov number NCT01525290

PRACTISE TRIAL: PENUMBRA AND RECANALISATION ACUTE COMPUTED TOMOGRAPHY IN ISCHAEMIC STROKE EVALUATION

S. Eltawil¹, A. Murray², J. Wardlaw³, K. Lalit⁴, I. Ford⁵, T. Robinson⁶, E. Warburton⁷, J. Freeman⁸ and K. Muir⁹

¹Univesrsity of Glasgow, Neurology, Gla, United Kingdom

²Univeristy of Glasgow, Instiute of Neuroloical sciences, Glasgow, United Kingdom

³Univeristy of Edinburgh, Division of Neuroimaging Sciences, Edinburgh, United Kingdom

⁴King’s college london, Department of Basic and Clinical Neurosciences, London, United Kingdom

⁵University of Glasgow, Robertson Center for biostatistics, Glasgow, United Kingdom

⁶Department of Cardiovascular Sciences- Ageing and Stroke Medicine Group, University of Leicester, Leicester, United Kingdom

⁷University of Cambridge, Department of Clinical Neurosciences Neurology Unit, Cambridge, United Kingdom

⁸Not applicable, Not apllicable, Cornwal, United Kingdom

⁹University of Glasgow, Institute of Neuroscience & Psychology, Glasgow, United Kingdom

It is currently unknown whether benefits of using multimodal CT imaging (including CTP and CTA), outweigh the disadvantages of additional resource utilisation, radiation and contrast exposure, and treatment delay associated with the use of additional multimodal CT imaging. This study aims to evaluate the effect of additional CT imaging on the number of acute stroke patients treated with IV rtPA and their outcomes.

PRACTISE is a prospective, multicentre randomised controlled trial (RCT).

Patients with acute ischaemic stroke, ≥18 years, and clinically eligible for IV rtPA treatment are randomised in a 1:1 ratio to control (NCCT alone) or multimodal CT (CT+CTA+CTP). The primary endpoint is the proportion treated with rtPA. Secondary endpoints evaluate times to decision making, comparison of different image processing software and clinical outcomes at 3 months.

A total of 272 patients were randomised to PRACTISE study, Detailed results will be reported later

analysis is on going. Conclusions will be reported later

NCT02360670

Scientific Communications (Oral Abstract Presentation Sessions)

Scientific Communications 01-Endovascular Stroke Treatment – Clinical Practice

STROKE THROMBECTOMY IN CLINICAL PRACTICE: RESULTS FROM THE GERMAN STROKE REGISTRY – ENDOVASCULAR TREATMENT

F. Wollenweber¹, J. Fiehler²; on behalf of the German Stroke Registry – Endovascular Treatment (GSR-ET) Collaborators

¹Institute for stroke and dementia ISD, Neurology, Munich, Germany

²University Medical Center Hamburg-Eppendorf, Department of Neuroradiology, Hamburg, Germany

Endovascular treatment (EVT) for large vessel occlusion in ischemic stroke has proven to be highly efficient and safe in large clinical trials. However, industry independent, real world data remains limited within this still evolving field.

2,794 patients treated with EVT for large vessel occlusion in ischemic stroke have been included into an academic, independent, prospective, multicenter register in 25 participating sites in Germany between June 2015 and April 2018. All patients received EVT as per standard of care. The primary outcome was the modified Rankin scale (mRS) at 3 months after EVT. Secondary outcomes included reperfusion success measured by the Thrombolysis In Cerebral Infarction (TICI) scale. After a central data quality control, 2611 patients entered the final analysis.

Median age was 75 years (IQR 65–82) and median NIHSS score before EVT was 15 (IQR 10–19). Vessel occlusion was located in the anterior circulation in 88%. Successful reperfusion (TICI 2b/3) was achieved in 83%. After 3 months, 26% (568/2237) of the patients showed an excellent outcome (mRS 0–1) and 37% (822/2237) a good outcome (mRS 0–2); mortality was 29% (640/2237). Functional outcome was comparable between anterior and posterior circulation but strongly associated with stroke severity (NIHSS) and reperfusion success (TICI) (see figure).

This large registry trial demonstrates high rates of reperfusion and favorable outcome after EVT in clinical routine for both anterior and posterior circulation large vessel strokes.

ClinicalTrials.gov Identifier: NCT03356392

INTRAVENOUS TROMBOLYSIS DOES NOT DELAY TIME TO ENDOVASCULAR TREATMENT, SUBSTUDY RESULTS FROM THE MR CLEAN REGISTRY

W. Hinsenveld¹, I. de Ridder¹, R. van Oostenbrugge¹, W. van Zwam², J.A. Vos³, G. Lycklama à Nijeholt⁴, J. Boiten⁵, W. Schonewille⁶; the MR CLEAN Registry Investigators

¹Maastricht University Medical Center, Neurology, Maastricht, The Netherlands

²Maastricht University Medical Center, Radiology, Maastricht, The Netherlands

³St. Antonius Hospital, Radiology, Nieuwegein, The Netherlands

⁴Haaglanden Medical Center, Radiology, The Hague, The Netherlands

⁵Haaglanden Medical Center, Neurology, The Hague, The Netherlands

⁶St. Antonius Hospital, Neurology, Nieuwegein, The Netherlands

Endovascular treatment (EVT) is effective and safe in ischemic stroke caused by large vessel occlusion, with or without intravenous thrombolysis (IVT). EVT alone may be more effective, as prior IVT may cause a delay in EVT initiation and may increase the risk of hemorrhage. We sought to determine the effect of prior IVT on time to treatment and risk of intracranial hemorrhage in patients treated with EVT.

We analyzed data from the MR CLEAN Registry, a prospective, multicenter, observational study in the Netherlands and included patients with an anterior circulation occlusion treated with EVT who presented directly to intervention centers between March 2014 and June 2016. Primary endpoint was door to groin time (DGT). Secondary outcomes were workflow time intervals and safety outcomes. We log-transformed time intervals, adjusted for baseline factors and used linear and ordinal regression models.

We included 665 patients of which 500 (75%) were treated with EVT+IVT. Adjusted DGT was similar in both groups with a 2.1% difference (95% CI: -5.4 – 9.1) in favor of EVT only. Door to CT times were reduced in the EVT+IVT group (32.6% difference, 95% CI: 21.3 – 42.2), while CT to groin times were increased (11.0% difference, 95% CI: 0.2 – 19.8) and DSA more often showed no treatable target occlusion (10.5% versus 2.3%, p = 0.010). Rates of symptomatic intracranial hemorrhage were similar for both groups.

Prior intravenous thrombolysis does not delay time to endovascular treatment and does not seem to increase the risk of intracranial hemorrhage.

N/A

BRIDGING THERAPY VERSUS DIRECT MECHANICAL THROMBECTOMY IN ACUTE ISCHEMIC STROKE PATIENTS WITH LARGE VESSEL OCCLUSION: A SYSTEMATIC REVIEW AND META-ANALYSIS

G. Tsivgoulis^1,2, A. Katsanos^1,3, K. Malhotra⁴, N. Goyal², A. Arhur⁵, P. Schellinger⁶, M. Köhrmann⁷, C. Krogias³, G. Turc⁸, G. Magoufis⁹, D. Leys¹⁰, N. Ahmed^11,12, P. Khatri¹³, M. Goyal¹⁴ and A.V. Alexandrov²

¹National and Kapodistrian University of Athens, Second Department of Neurology, Athens, Greece

²University of Tennessee Health Science Center, Department of Neurology, Memphis, USA

³St. Josef-Hospital- Ruhr University, Department of Neurology, Bochum, Germany

⁴West Virginia University- Charleston Division, Department of Neurology, Charleston, USA

⁵University of Tennessee Health Science Center, Department of Neurosurgery, Memphis, USA

⁶Johannes Wesling Medical Center- Ruhr University Bochum, Department of Neurology and Neurogeriatry, Minden, Germany

⁷Universitätsklinikum Essen, Department of Neurology, Essen, Germany

⁸Hôpital Sainte-Anne, Department of Neurology, Paris, France

⁹Metropolitan Hospital, Stroke Unit, Piraeus, Greece

¹⁰Univ-Lille, Inserm U1171- CHU Lille, Lille, France

¹¹Karolinska University Hospital Solna, Department of Neurology, Stockholm, Sweden

¹²Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden

¹³University of Cincinnati, Department of Neurology, Cincinnati, USA

¹⁴University of Calgary, Departments of Radiology and Clinical Neurosciences, Calgary, Canada

The substantial clinical improvement in acute ischemic stroke (AIS) patients treated with mechanical thrombectomy (MT) combined with the poor response of proximal intracranial occlusions to intravenous thrombolysis (IVT), resulted in questioning the utility of bridging therapy (BT; IVT followed by MT), compared to direct mechanical thrombectomy (dMT) for AIS patients with large vessel occlusions (LVO).

We performed a systematic review and random-effects meta-analysis of observational studies to assess unadjusted and adjusted for potential confounders outcomes between BT and dMT.

We identified 36 eligible studies including a total of 11,295 LVO patients (mean age 68 years, median NIHSS score 16, 51% women, 57% treated with BT). In unadjusted analyses BT was associated with higher likelihood of three-month functional independence (mRS-scores 0–2; OR = 1.51, 95% CI: 1.30-1.75), three-month functional improvement (cOR for 1-point decrease in mRS-score = 1.52, 95% CI: 1.18-1.97). BT was also related to lower likelihood of three-month mortality (OR = 0.64, 95% CI: 0.56-0.73), with no significant differences in the risks of symptomatic intracranial hemorrhage (OR = 0.94, 95% CI: 0.77-1.14). In the adjusted analyses, BT was independently associated with higher likelihood of three-month functional independence (adjusted OR = 1.55, 95% CI: 1.25-1.92) and lower odds of three-month mortality (adjusted OR = 0.79, 95% CI: 0.65-0.97) compared to dMT. The two groups did not differ in in functional improvement (adjusted cOR = 1.24, 95% CI: 0.89-1.74) and symptomatic intracranial hemorrhage (adjusted OR = 0.87, 95% CI: 0.61-1.25).

BT appears to be associated with improved functional independence and no safety concerns, compared to dMT, for AIS patients with LVO.

N/A

THROMBECTOMY IN ANDALUCIA USING ASPIRATION (TRIANA) VS STANDARD STENTRIEVER: POTENCY AND QUICKNESS OF THE THERAPY

A. De Alboniga-chindurza^1,2, J. Ortega-Quintanilla¹, F. Delgado³, J. Alcalde-Ló pez¹, R. Oteros³, E. Zapata-Arriaza^1,2, I. Bravo³, F. Moniche^2,4, E. Jiménez-Gómez³, I. Escudero-Martínez^2,4, R. Valverde⁵, M. Zapata², S. Pérez^2,6, M.Á. Gamero⁶, P. Baena⁴, L. Lebrato^2,4, A. Cayuela⁷, J.A. Cabezas^2,4, J. Montaner^2,6 and A. González^1,2

¹Virgen del Rocío University Hospital, Radiology- Interventional Neuroradiology, Seville, Spain

²Institute of Biomedicine of Seville, Neurovascular Research Laboratory, Seville, Spain

³Reina Sofía University Hospital, Radiology- Interventional Neuroradiology, Cordoba, Spain

⁴Virgen del Rocío University Hospital, Neurology, Seville, Spain

⁵Reina Sofía University Hospital, Neurology, Cordoba, Spain

⁶Virgen Macarena University Hospital, Neurology, Seville, Spain

⁷Area of Sanitary Management South of Seville, Unit of Clinical Management of Public Health- Prevention and Promotion of Health, Seville, Spain

Interventions below 60 minutes could achieve better functional prognosis. Therefore, we compare capacity and quickness of Direct Aspiration First Pass Technique (ADAPT) vs stentriever, in order to be able to adjust to the resources of our health system.

We design a multicenter, prospective, non-randomized study in patients with anterior circulation acute ischemic stroke, large vessel occlusion and less than 8 hours of evolution treated with endovascular rescue techniques: ADAPT using SOFIA® catheter in comparison with the stentriever.

Of 532 patients, 370 (69.5%) were treated with ADAPT and 162 (30.5%) with stentriever as first choice. For primary outcome, the proportion of patients with successful revascularization was 93.5% in ADAPT vs 80.2% in stentriever (p = 0.003). When another treatment was not required (76.5% in aspiration, 95.7% in stentriever), in ADAPT group lower number of passes were needed: 1 (IQR: 1–2) vs 2 (IQR: 1–3) stentriever (p < 0.01), and intervention was faster: median 19 vs 60 min stentriever. Whereas in all patients in whom aspiration was used, 30.8% TICI 3 rate was achieved in first pass; when a rescue technique was not necessary, TICI 3 improved to 40.4% (12.9% and 13.5% respectively for stentriever, p < 0.01). If performing a rescue treatment was not required, a better outcome at 3 months was achieved in aspiration group: mRS 1 (IQR: 0–3) vs mRS 2 (IQR: 1–5) (p = 0.003).

Although further randomized clinical trials are needed, ADAPT could be quicker and more resolutive when another technique is not required.

ClinicalTrials.gov Identifier: NCT03407092

MEDICAL MANAGEMENT VERSUS MECHANICAL THROMBECTOMY FOR MILD DEFICITS STROKES: INTERNATIONAL MULTICENTER STUDY AND META-ANALYSIS

N. Goyal¹, G. Tsivgoulis¹, M.F. Ishfaq¹, M.T. Frohler², A.M. Spiotta³, M. Psychogios⁴, A. Siddiqui⁵, M. Groen⁶, C. Krogias⁷, M. Saqqur⁸, M. Mokin⁹, R. Leker¹⁰, A.H. Katsanos¹¹, V. Lioutas¹², V. Sharma¹³, M. Paciaroni¹⁴, A. Rentzos¹⁵, J.D. Mocco¹⁶, A.V. Alexandrov¹ and A.S. Arthur¹⁷

¹University of Tennessee Health Science Center, Neurology, Memphis, USA

²Vanderbilt University, Cerebrovascular Program, Nashville, USA

³Medical University of South Carolina, Department of Neurosurgery, Charleston, USA

⁴University Medical Center Göttingen, Department of Neuroradiology, Göttingen, Germany

⁵University at Buffalo, Departments of Neurosurgery and Radiology, New York, USA

⁶Johannes Wesling Medical Center Minden, Department of Neurology and Neurogeriatry, Minden, USA

⁷St. Josef-Hospital- Ruhr University of Bochum- Germany, Department of Neurology, Bochum, Germany

⁸Hammad Medical Center, Department of Neurology, Doha, Qatar

⁹University of South Florida, Department of Neurosurgery, Tampa, USA

¹⁰The Agnes Ginges Center of Neurogenetics- Hadassah-Hebrew University Medical Center, Department of Neurology, Jerusalem, Israel

¹¹Metropolitan Hospital, Acute Stroke Unit-, Piraeus, Greece

¹²Beth Israel Deaconess Medical Center- Harvard Medical School, Department of Neurology, Boston, USA

¹³Yong Loo Lin School of Medicine- National University of Singapore, Division of Neurology, Singapore, Singapore

¹⁴Divisione di Medicina Cardiovascolare- Università di Perugia-, Stroke Unit, Perugia, Italy

¹⁵Department of Interventional and diagnostic Neuroradiology, Department of Interventional and diagnostic Neuroradiology, Gothenburg, Sweden

¹⁶Mount Sinai Medical Center-, Department of Neurosurgery-, New York, USA

¹⁷University of Tennessee Health Science Center, Neurosurgery, Memphis, USA

Mechanical-thrombectomy (MT) benefits for mild deficits emergency-large-vessel-occlusions (mELVO-NIHSS < 6) are uncertain. We sought to evaluate outcomes of MT in comparison to best-medical-management (bMM) in mELVO patients.

A total of 251 anterior circulation mELVO (MT: 138, median NIHSS 4, bMM: 113, median NIHSS 3) presenting within 24 hours were reviewed and compared. A meta-analysis of studies reporting outcomes of MT in comparison to bMM was conducted.

The bMM group had a lower rate of asymptomatic ICH (4.6% vs. 17.5%, p = 0.002) compared to the MT group. After imputation of missing data, the univariable-analysis showed higher rates of 3-months functional-independence in the bMM group (mRS 0–2: 88.5% vs. 77.4%, p = 0.021) but in multivariable analyses, this difference was non-significant. Otherwise the two groups did not differ in terms of 3-months FFO (favorable-functional-outcome: mRS 0–1), sICH, or 3-month-mortality. In multivariable-analyses, MT was associated with higher odds of asymptomatic ICH (OR 4.13, 95% CI; 1.50-11.40, p = 0.006), but was not related to sICH, 3-month functional outcomes or mortality. Five studies (890 patients) met the inclusion criteria of our meta-analysis. The odds of sICH were higher in the MT arm when compared with the bMM group (OR 5.22,95% CI; 1.85-14.72, p = 0.002), however this difference failed to reach significance in adjusted analyses. Otherwise, the MT and bMM groups did not differ in terms of mortality, 3-month FFO, and 3-month FI.

Our study suggests no improvement in functional-outcomes in mELVO patients receiving MT compared to bMM. MT was associated with higher rates of asymptomatic ICH compared to bMM.

N/A

ISOLATED DISTAL INTERNAL CAROTID ARTERY OCCLUSIONS WITH PATENT INTRACRANIAL CIRCULATION IN ACUTE ISCHEMIC STROKE: NO REASON TO REFRAIN FROM ENDOVASCULAR TREATMENT

J.W. Hoving¹, M. Kappelhof¹, M. Schembri¹, B. Emmer¹, O. Berkhemer A¹, A. Groot², D. Dippel³, W. van Zwam⁴, J. Coutinho², C. Majoie¹; for the MR CLEAN Registry Investigators

¹Amsterdam UMC- University of Amsterdam, Department of Radiology, Amsterdam, The Netherlands

²Amsterdam UMC- University of Amsterdam, Department of Neurology, Amsterdam, The Netherlands

³Erasmus MC- University Medical Center Rotterdam, Department of Neurology, Rotterdam, The Netherlands

⁴Maastricht UMC, Department of Radiology, Maastricht, The Netherlands

Patients with acute occlusions of the distal internal carotid artery with an open ICA-terminus and intact collateral flow via the circle of Willis may achieve functional independence after stroke even without treatment. However, the optimal treatment for these ‘carotid I-occlusions’ is not known. We aimed to investigate the results of endovascular treatment (EVT) compared to conservative treatment in patients with carotid I-occlusions.

We compared EVT-treated patients with carotid I-occlusions from the MR CLEAN Registry to non-EVT-treated patients from our center. CT angiography images were reviewed on occlusion pattern, collateral structures, and plexus enhancement. Our primary outcome was 90-day functional independence (mRS 0–2). Secondary outcomes were 90-day mortality and delta-NIHSS.

Twenty-nine patients received EVT and nine patients did not. IVT was administered in 79% (23/29) in EVT-treated patients versus 67% (6/9; p = 0.66) in non-EVT-treated patients. Median baseline NIHSS was 12 versus 16 (p = 0.21). Functional independence occurred in 52% (14/27) versus 22% (2/9) in the respective groups (aOR 0.48;95% CI 0.1-3.4). Mortality for EVT versus non-EVT treated patients was 19% (5/27) versus 33% (3/9)(aOR 0.96; 95% CI 0.1-9.8). Respective median delta-NIHSS was 7 (IQR1-9) versus 8 (IQR4-11)(b0.1; 95% CI -3.9-7.3). Patency of the anterior choroidal artery, anterior communicating artery, posterior artery, and plexus enhancement were positively related with functional independence.

Our data suggest improved outcomes in carotid I-occlusion patients treated with EVT, similar to other patients in the thrombectomy trials. Good collateral flow with a patent middle cerebral artery is no reason to refrain from EVT in acute stroke due to carotid I-occlusion.

N/A

RECANALISATION TREATMENT IN CHILDHOOD AND YOUNG ADULT STROKE – COMPARISON OF SAFETY AND OUTCOME

U. Fischer¹, M. Steinlin², J. Gralla³, C. Bernasconi¹, A. Datta⁴, M. Arnold¹, A. Dulcey², J. Kaesmacher¹, J. Fluss⁵, A. Hackenberg⁶, O. Maier⁷, J. Weber⁸, C. Poloni⁹, S. Bellwald¹, S. Bigi²

¹University Hospital of Bern- University of Bern, Department of Neurology, Bern, Switzerland

²Universtiy Children’s Hospital- University of Bern, Department of Pediatrics- Division of Child Neurology, Bern, Switzerland

³University Hospital of Bern- University of Bern, Department of Neuroradiology, Bern, Switzerland

⁴University Children’s Hospital Basel- Universty of Basel, Department of Pediatrics- Division of Child Neurology, Basel, Switzerland

⁵University Children’s Hospital Geneva- University of Geneva, Department of Pediatrics- Division of Child Neurology, Geneva, Switzerland

⁶University Children’s Hospital Zurich- Universtiy of Zurich, Department of Pediatrics- Division of Child Neurology, Zurich, Switzerland

⁷Children’s Hospital St. Gallen, Department of Pediatrics, St. Gallen, Switzerland

⁸Cantonal Hospital St. Gallen, Department of Neuroradiology, St. Gallen, Switzerland

⁹University Children’s Hospital Lausanne- University of Lausanne, Department of Pediatrics- Division of Child Neurology, Lausanne, Switzerland

Intravenous thrombolysis and endovascular therapy (IVT/EVT) are increasingly used in the pediatric arterial ischemic stroke (AIS) population. However, it remains unknown if these treatments are as safe and effective as in adult AIS patients. This study aimed to compare safety and outcome of IVT/EVT in children and young adults with AIS.

Retrospective study (01/2000–12/2017) of patients aged 1 month – 45 years treated with IVT/EVT for AIS. Clinical and radiological data of children (1 month – 16 years) were compared to young adults (>16 years – 45 years) by univariate analysis. Delayed treatment initiation was defined >4.5 hours and >6 hours after symptom onset for IVT and EVT respectively. Outcome was assessed using the mRS 3–6 months after AIS.

19 children (age 11 +/- 3.9 years, 6 (31.6%) females) and 175 adults (age 36 +/- 8.1 years, 85 (48.6%) females) were included. Overall, children showed different etiologies (P = 0.001) and lower recanalisation rates (p = 0.019), whereas initial pedNIHSS, bleeding complications, mortality and outcome did not differ between the two groups. Comparing IVT to EVT, children receiving IVT showed a higher initial pedNIHSS (p = 0.003), delayed treatment initiation (p = 0.001) and worse outcome (p = 0.003) compared to adults treated with IVT. No major differences were observed in children and adults treated with EVT.

Safety and outcome of recanalization treatment in children is comparable to findings observed in young adults. Delayed treatment initiation is a serious concern in the pediatric population treated with IVT and is probably responsible for the observed differences in outcome parameters.

N/A

COST-EFFECTIVENESS OF TENECTEPLASE BEFORE THROMBECTOMY FOR ISCHEMIC STROKE: ECONOMIC EVALUATION OF THE EXTEND-IA TNK RANDOMISED CONTROLLED TRIAL

L. Gao^1,2, M. Moodie^1,2, P. Mitchell³, L. Churilov⁴, T. Kleinig⁵, N. Yassi^6,7, M. Parsons⁶, G. Donnan⁷, S. Davis⁶, B. Campbell⁶; EXTEND-IA TNK Investigators

¹Deakin University, Deakin Health Economics, Melbourne, Australia

²Deakin University, Global Obesity Centre- Centre for Population Health Research, Melbourne, Australia

³Royal Melbourne Hospital, Department of Radiology, Melbourne, Australia

⁴Florey Institute of Neuroscience and Mental Health- University of Melbourne, NHMRC Research Excellence in Stroke Rehabilitation and Brain Recovery, Melbourne, Australia

⁵Royal Adelaide Hospital, Department of Neurology, Adelaide, Australia

⁶Royal Melbourne Hospital, Department of Medicine and Neurology, Melbourne, Australia

⁷University of Melbourne, Florey Institute of Neuroscience and Mental Health, Melbourne, Australia

Tenecteplase reduced disability and the requirement for endovascular thrombectomy in ischemic stroke patients with large vessel occlusion. We assessed the cost-effectiveness of tenecteplase versus alteplase.

Within-trial economic analysis of the EXTEND-IA TNK randomised controlled trial (RCT) was performed alongside long-term modelled cost-effectiveness analysis. The trial evaluation included costs relating to the index emergency department and inpatient stroke hospitalisation, rehabilitation/subacute care and rehospitalisation due to stroke within 90 days. Quality-adjusted life years (QALY) were estimated using utility scores derived from the modified Rankin Score (mRS) at Day 90. In the modelled economic evaluation, a Markov model with seven health states corresponding to seven mRS scores was adopted to simulate the patient outcomes and costs for the lifetime. Model parameters including resource use, unit cost and utility weights were sourced from published literature or government websites.

In the RCT, the total cost was nominally lower in the tenecteplase-treated patients (AUD64,539) compared to alteplase-treated patients (AUD71,473). Tenecteplase was the dominant treatment strategy in the short-term, with lower cost (AUD-7302, 95% CI: -27048-12444, p = 0.469) and higher benefits (QALY 0.099, 95% CI: 0.001- 0.1967, p = 0.048). In the long-term, tenecteplase was associated with less cost (AUD37,761 vs AUD42,050) and greater benefits (QALY 7.18 vs 5.99), and had a probability of 100% being cost-effective.

Both within trial and long-term economic analyses showed that tenecteplase was highly likely to be cost-effective for acute stroke patients prior to thrombectomy.

N/A

PREDICTORS OF SELF-REPORTED QUALITY OF LIFE AFTER STROKE THROMBECTOMY

M. Deb-Chatterji¹, F. Flottmann², H. Leischner², A. Alegiani¹, C. Brekenfeld², J. Fiehler², C. Gerloff¹ and G. Thomalla¹

¹University Medical Center Hamburg-Eppendorf, Neurology, Hamburg, Germany

²University Medical Center Hamburg-Eppendorf, Neuroradiological Diagnostics and Intervention, Hamburg, Germany

Several randomized controlled trials (RCTs) demonstrated improved outcome in patients with large vessel occlusion (LVO) treated by endovascular thrombectomy (ET) compared to standard medical care. However, little is known about self-reported quality of life (QoL) in these patients. We assessed whether the treatment benefit in patients with ET ascertained by the modified Rankin Scale (mRS) at 90 days in prior RCTs is also reflected in QoL of these patients.

In a single-center prospective study we collected data from stroke patients with LVO treated by ET. Baseline characteristics, procedural and outcome parameter were evaluated. Health-related QoL was measured by the European QoL-five dimensions questionnaire (EQ-5D) utility index (ranging from –0.594 to 1, higher values indicate better QoL) three months after stroke. Multivariate linear regression analysis was applied to identify predictors of better QoL.

Four hundred thirty-five patients were included in the study (median age: 76 years, 52% female), median baseline NIHSS was 15. Successful recanalization (defined by Thrombolysis in Cerebral Infarction (TICI) score ≥ 2b) was achieved in 73%. Mean EQ-5D utility index was 0.38 [SD 0.44]. A lower age (p < 0.001), female sex (p = 0.023), lower pre-stroke mRS (p = 0.001), lower baseline NIHSS score (p < 0.001), higher ASPECT score (p < 0.001), concomitant IVT (p = 0.032) and a successful recanalization (p < 0.001) were associated with better QoL.

Predictors of better health-related quality of life after stroke thrombectomy closely match the predictors of better functional outcome in these patients that were identified in previous trials.

N/A

TEMPORAL TRENDS IN STROKE INCIDENCE IN HIGH-INCOME COUNTRIES IN THE 21ST CENTURY: POPULATION-BASED STUDY AND SYSTEMATIC REVIEW

L. Li¹, C.A. Scott¹, P.M. Rothwell; On behalf of the Oxford Vascular Study¹

¹Centre for Prevention of Stroke and Dementia, Nuffield Department of Clinical Neurosciences, Oxford, United Kingdom

Population-based studies provide most reliable data on stroke incidence. A previous systematic review of population-based studies suggested that stroke incidence in high-income countries decreased by 42% between the 1970s and early 2000s. However, it is uncertain whether this trend of steady decline has been maintained in more recent periods.

All published population-based stroke incidence studies that reported temporal trends of stroke incidence since 1990 in high-income countries were included along with unpublished data from the Oxford Vascular Study (OXVASC; 2002–2017). Age-standardised relative incidence rate ratios (IRR) were calculated for each study and then pooled with inverse variance weighted random effects meta-analysis.

Of 12 population-based studies from 9 high-income countries, 30,321 incident stroke cases occurred during 17,903,041 person-years of observation. Despite variation in age-standardised absolute stroke incidence rates between studies, stroke incidence declined steadily between the 1990s and 2010s within each study, resulting in a 23% decline over an average period of 16 years (7 studies; pooled IRR = 0.77, 95% CI 0.71-0.83, p < 0.0001). The trend was the same for men (6 studies; 0.72; 0.63-0.82, p < 0.0001) and for women (6 studies; 0.71; 0.63-0.79, p < 0.0001), and remained consistent after year 2010 in OXVASC. Proportion of disabling or fatal stroke also decreased over time (2 studies; early vs. later period 53.6% vs. 46.1%, OR = 0.77, 0.62-0.96, p = 0.02). However, no decrease was observed for individuals aged < 55 years (6 studies; 1.03, 0.82-1.30, p = 0.79).

Stroke incidence is continuing to decline with steady rate in Oxfordshire and in other high-income settings. However, no decrease is observed at younger ages.

N/A

20-YEAR TIME TRENDS IN LONG-TERM CASE-FATALITY AND RECURRENCE RATES AFTER ISCHEMIC STROKE STRATIFIED BY ETIOLOGY – RESULTS FROM THE POPULATION-BASED ERLANGER STROKE PROJECT

V. Rücker¹, P.U. Heuschmann^1,2,3, M. O’Flaherty⁴, M. Weingärtner⁵, M. Hess⁵, C. Sedlak⁵, S. Schwab⁶ and P.L. Kolominsky-Rabas⁵

¹University of Würzburg, Institute of Clinical Epidemiology and Biometry, Würzburg, Germany

²University Hospital Würzburg, Clinical Trial Centre, Würzburg, Germany

³University Würzburg, Comprehensive Heart Failure Centre, Würzburg, Germany

⁴University of Liverpool, Institute of Population Health Sciences, Liverpool, United Kingdom

⁵Friedrich-Alexander-University of Erlangen-Nürnberg, Interdisciplinary Centre for Health Technology Assessment HTA and Public Health, Erlangen, Germany

⁶University Hospital Erlangen, Department of Neurology, Erlangen, Germany

Data on long-term survival and recurrence patterns after stroke are lacking. We investigated time trends in ischemic stroke case-fatality and recurrence rates from 1996–2015 stratified by etiological subtype within a population-based stroke register in Germany.

Data was collected within the Erlangen Stroke Project, a prospective, population-based stroke register covering a population of 105,164 inhabitants (2010). Time trends for case-fatality and recurrence rates up to 5-years after index event were estimated with Kaplan-Meier and Cox Regression. We adjusted for age, sex and year of event and stratified for etiological subtypes according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification.

3,343 patients with first ischemic stroke were included; age-standardized incidence per 100,000 was 75.8 in women and 131.6 in men (2015). Overall, 1-year and 5-year survival probabilities were 75.9% and 50.4% in women and 82.8% and 59.2% in men; 5-year survival was highest in patients with stroke due to small-artery occlusion (women: 71.8%; men: 75.9%) and lowest in cardioembolic stroke (women: 35.7%; men: 47.8%). 1-year and 5-year recurrence rates were 7.5% and 21.1%; 5-year recurrence rate was lowest in women in stroke due to small artery occlusion (16.0%) and in men in large-artery atherosclerosis (16.6%); highest risk of recurrence was observed in undefined strokes (women: 22.3%; men: 21.4%). Cox regression revealed improved case-fatality rates over time. No statistically significant time trends were observed for stroke recurrence.

Long-term survival and recurrence varied by stroke etiology. Survival probabilities improved, but no major trends in recurrence rates were observed.

N/A

LONG-TERM TRENDS IN MORTALITY AND DEPENDENCE AFTER ISCHAEMIC STROKES: PROSPECTIVE POPULATION STUDY OF THE SOUTH LONDON STROKE REGISTER

H. Wafa¹, C. Wolfe^1,2,3, A. Rudd^1,3 and Y. Wang^1,2,3

¹King’s College London, School of Populaiton Health and Environmental Sciences, London, United Kingdom

²National Institute for Health Research NIHR, Collaboration for Leadership in Applied Health Research and Care CLAHRC South London, London, United Kingdom

³National Institute for Health Research NIHR Biomedical Research Centre, Guy’s and St Thomas’ NHS Foundation Trust and King’s College London, London, United Kingdom

To estimate changes in mortality and functional dependence after ischaemic strokes (ISs) over a 16-year period, and to ascertain demographic disparities.

We identified all first-ever IS cases from the population-based South London Stroke Register (SLSR) between 2000 and 2015. IS was classified into large-artery atherosclerosis (LAA), cardio-embolism (CE), small-vessel occlusion (SVO), other determined aetiologies (OTH), and undetermined aetiologies (UND). Using proportional-hazards survival modelling, adjusted for demographic factors and case mix variables, we examined temporal trends in one-year mortality rates. Additional regression analyses explored whether trends occurred at the expense of greater functional dependence (Barthel index < 15) among survivors.

A total of 3,128 patients with first ISs were registered. Risk-adjusted rates of one-year mortality decreased from 30·1% in 2000–2003 to 22·8% in 2012–2015 (adjusted hazard ratio per year, 0·97; 95% confidence interval [CI], 0·96 to 0·99; P = 0·0002 for trend). Improvements were noted in all IS subtypes and ethnic groups, however, were only significant in CE and UND and in those of white ethnic origins. Functional dependence has decreased between 2000–2003 and 2012–2015 at seven days poststroke (adjusted odds ratio [OR], 0·61; 95% CI, 0·44 to 0·84) but not at one-year poststroke (OR, 1·15; 95% CI, 0·8 to 1·64).

Both mortality and functional dependence after ISs have decreased during the past 16 years in association with improvements in the acute processes of care. Reductions in mortality are particularly evident in CE and UND subtypes and in the white population.

N/A

PRESENCE, DURATION, AND CONTROL OF VASCULAR RISK FACTORS AND INTRACRANIAL STENOSIS: THE NORTHERN MANHATTAN STUDY (NOMAS)

J. Gutierrez¹, F. Khasiyev¹, A. Gil-Guevara², S. Tom¹, M. Santiago¹, C. Wright³, T. Rundek⁴, R. Sacco⁴ and M. Elkind¹

¹Columbia University, Neurology, New York, USA

²University of Texas, Medicine, Galveston, USA

³National Institute of Health, Neurology, Bethesda, USA

⁴University of Miami, Neurology, Miami, USA

To test the hypothesis that that prolonged exposure to vascular risks associates with intracranial stenosis in stroke-free individuals.

The Northern Manhattan Study recruited randomly selected stroke-free individuals ≥ 40-years living in Northern Manhattan between in 1993 and 2001. Participants who remained stroke-free had a brain MRA between 2003 and 2008. During the interval between enrollment and the brain MRI, participants were assessed annually for vascular risk evaluation. Intracranial stenosis categorized as no stenosis, < 50%, 50–80% and > 80% stenosis. We used fixed effects from adjusted hierarchical generalized linear models to examine the relationship between vascular risks and intracranial stenosis.

1029 participants (58% women, 63% Hispanic, mean age 64 ± 8 years were followed for an average of 7 years (IQR 5–9). Forty percent had intracranial stenosis. Any intracranial stenosis was associated with older age (B = 0.04, P < 0.001) and with vascular risks at enrollment (hypertension B = 0.55, P = 0.002; diabetes B = 0.62, P < 0.001; dyslipidemia B = 0.63, P = 0.02) but not with incident vascular risks during follow-up. Intracranial stenosis was associated with repeated measures of systolic blood pressure (B = 0.11 per 10 mmHg increment, P = 0.005), fasting glucose (B = 0.05 per 10 mg/dl increment, P = 0.03), low-density lipoprotein cholesterol (B = 0.04 per 10 mg/dl increment, P = 0.03) and lifetime pack-years of smoking (B = 0.002 per each 100 packs, P = 0.01).

Intracranial stenosis is associated with longer exposure and poorer control of vascular risks. Early diagnosis and aggressive medical management of vascular risk may reduce intracranial stenosis and subsequent stroke risk. Evaluating a role for adjunctive therapies should be considered a research priority in this field.

NA

HOW ARE COMORBIDITIES RELATED TO RECURRENT STROKE? AN ELECTRONIC DATA LINKAGE STUDY

M.J. Macleod¹, P. Langhorne² and M. Turner³

¹University of Aberdeen, Institute of Medical Sciences, Aberdeen, United Kingdom

²University of Glasgow, Division of Cardiovascular and Medical Sciences, Glasgow, United Kingdom

³Institute of Applied Health Sciences, Aberdeen, United Kingdom

60% of patients with stroke are reported to have three or more comorbid conditions. For those who survive an initial event, recurrent stroke is an important complication. We explored the relationship of comorbidity with recurrent stroke, defined as a new event requiring hospital admission.

Data was extracted from the Scottish Stroke Care Audit and linked to hospital admissions for the ten years prior to the initial stroke event.

77779 patients were included in the analysis. 6.6% had a recurrent stroke within one year. Patients with myocardial infarction, peripheral vascular disease, dementia, chronic pulmonary disease, diabetes, and renal disease had an increased likelihood of recurrent stroke within 1 year as did those with an increasing Charlson comorbidity index (CCI) score.

Table 1. Relationship of recurrent stroke within 1 year with comorbidities. Outcomes were adjusted for six simple variables, stroke type, stroke unit admission, admission year, and hospital.

Table

In this population based study, several specific comorbidities plus a higher CCI were associated with an increased odds of recurrent stroke within 1 year of an initial stroke. Further work is required to explore the presentation of patients with comorbidities.

N/A

STATISTICAL METHODS TO GENERATE VIRTUAL PATIENT POPULATIONS: A STROKE EXAMPLE

N. Samuels¹, D. Nieboer², A.G. Hoekstra³, H.A. Marquering⁴, C.B.L.M. Majoie⁵, D.W.J. Dippel⁶, D. Kurowicka⁷, A. van der Lugt⁸ and H.F. Lingsma⁹

¹Erasmus University Medical Center, Neurology, Rotteram, The Netherlands

²Erasmus University Medical Center, Public Health, Rotterdam, The Netherlands

³Universiteit van Amsterdam, Informatics Institute, Amsterdam, The Netherlands

⁴Amsterdam University Medical Centers- location AMC, Radiology and Nuclear Medicine- Biomedical Engineering and Physics, Amsterdam, The Netherlands

⁵Amsterdam University Medical Centers- location AMC, Radiology and Nuclear Medicine, Amsterdam, The Netherlands

⁶Erasmus University Medical Center, Neurology, Rotterdam, The Netherlands

⁷Technical University of Delft, Applied Mathematics, Delft, The Netherlands

⁸Erasmus University Medical Center, Radiology and Nuclear Medicine, Rotteram, The Netherlands

⁹Erasmus University Medical Center, Public Health, Rotteram, The Netherlands

There is a growing interest in performing in silico clinical trials (ISCT) to support the design of clinical trials for new therapies. ISCT require cohorts of virtual patients (VP) to perform simulations. An advantage of a VP cohort is that data sharing is not limited by privacy issues. An important requirement for an ISCT to obtain realistic simulation results is that VPs are similar to real patients. We present three statistical methods to develop a representative ischemic stroke VP cohort.

Eleven prognostic clinical and imaging parameters were selected from patients in the MR CLEAN Registry database. To generate VPs we used a method similar to Multiple Imputation By Chained Equations (MICE), a Conditional Regression (CR) and a Normal Copula (NC) method. Performance of the methods was assessed graphically by comparing distributions and correlations of simulated and real patient data. Furthermore, discriminative ability of a model predicting whether a patient was from the original or simulated dataset was evaluated (area under ROC-curve).

We generated 1500 virtual stroke patients per method. The discrimination between the original and simulated set was poor, which implies adequate performance of all three methods, MICE = 0.55 (95%Confidence Interval (CI) 0.54-0.57), CR = 0.52 (95% CI 0.50-0.54), NC = 0.56 (95% CI 0.54-0.58). Correlations between variables in the simulated data were comparable to the ones in the original data.

All three statistical methods are appropriate to generate a representative cohort of virtual stroke patients.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 777072

DERIVING AN INDICATOR OF STROKE SEVERITY FROM ADMINISTRATIVE DATA

A.Y.X. Yu¹, P.C. Austin², M. Rashid², J. Fang², J. Porter², M.D. Hill³ and M.K. Kapral⁴

¹University of Toronto-SHSC, Medicine, Toronto, Canada

²ICES, Cardiovascular research program, Toronto, Canada

³University of Calgary, Department of Clinical Neurosciences, Calgary, Canada

⁴University of Toronto-UHN, Medicine, Toronto, Canada

Baseline stroke severity is an important prognostic variable that is not routinely collected in administrative data, but necessary for population-based outcome studies. We aimed to derive an indicator of stroke severity using administrative data.

We included all patients with stroke admitted to acute care hospitals between April 2002 and March 2013 in the Ontario Stroke Registry. The registry collected detailed clinical information, including stroke severity. We linked the registry to administrative databases and identified variables at or within the first 48 hours of admission that might be associated with severe stroke (Canadian Neurological Scale score < 4 or NIHSS >15). We used multivariable logistic regression models with variables from administrative data to predict severe strokes based on information in the registry and used bootstrap resampling to calculate C-statistics. We included all covariates significantly associated with the outcome in univariable analysis.

We included 39,972 patients (49% female, median age 75 [64,83] years). Variables in administrative data that were associated with severe stroke included female sex, higher age, high acuity triage level, arrival by ambulance, transfer to higher-level care, early admission to intensive care unit, early mechanical ventilation, motor weakness, speech disturbance, or decreased level of consciousness (p < 0.001). We are in the process of completing the multivariable analyses.

An indicator of stroke severity is crucial for risk-adjustment when evaluating stroke outcomes in population health studies. The results of this study will provide information on whether a measure of stroke severity can be derived from existing administrative databases.

N/A

ASSOCIATION BETWEEN FIRST-EVER ISCHAEMIC STROKE AND INCIDENT HEART DISEASE IN A POPULATION-BASED COHORT STUDY OF 93,627 OLDER ADULTS

L. Sposato¹, M. Lam², B. Allen², L. Richard², S. Shariff² and G. Saposnik³

¹Western University, Clinical Neurological Sciences, London, Canada

²ICES, Epidemiology, London, Canada

³University of Toronto, Department of Medicine- Division of Neurology, Toronto, Canada

Cardiac complications are common after ischaemic stroke. It remains unknown whether this is because of shared risk factors and pre-existing heart disease or, alternatively, due to stroke-induced heart injury. We assessed whether first-ever ischaemic stroke is associated with increased risk of incident heart disease in patients without pre-existing cardiovascular comorbidities.

This population-based cohort study included 93,627 residents in Ontario between 2002 and 2012, aged >65 years, without prior cardiovascular disease. We compared the incident risk of MACE (myocardial infarction, unstable angina, congestive heart failure, coronary artery disease, coronary artery revascularization, or cardiovascular death) at 1 year in patients with first-ever ischaemic stroke vs. propensity-matched individuals without stroke history or heart disease (4:1 matching, using 31 variables). We used Cox regression adjusted for variables with residual weighted standardized differences >10% or known to influence the risk of heart disease to estimate cause-specific hazard ratios (HR).

Overall, 21,931 patients with first-ever ischaemic stroke and 71,696 propensity-matched individuals were well-balanced on all variables used for propensity-matching (weighted differences ≤5%). First-ever ischaemic stroke was associated with increased risk of incident cardiovascular events (HR 4.5, 95% CI 4.3-4.8) after propensity-matching. The adjusted risk of MACE was highest in the first 30 days (HR 25.0, 95% CI 20.5-30.5) and declined thereafter: HR 4.8, 95% CI 4.1-5.7 between days 31 and 90, and HR 2.2, 95% CI 2.0-2.4 between days 91 and 365.

In this large population-based study, ischaemic stroke was independently associated with increased risk of incident heart disease, suggesting that stroke-induced heart injury may explain post-stroke cardiovascular complications.

N/A

WHAT ARE THE CAUSES OF RECURRENT ISCHEMIC STROKE AFTER EMBOLIC STROKE OF UNDETERMINED SOURCE (ESUS)? – INSIGHTS FROM THE NAVIGATE-ESUS TRIAL

R. Veltkamp¹, M. Sharma², P. Sheridan³, S. Kasner⁴, D. Toni⁵, S. Ameriso⁶, H. Mundl⁷, T. Tatlisumak⁸, G. Hankey⁹, A. Lindgren¹⁰, S. Berkowitz¹¹, A. Arauz¹², S. Ozturk¹³, K. Muir¹⁴, G. Peters¹⁵, A. Chamorro¹⁶, K. Perera², A. Shuaib¹⁷, S. Rudilosso¹⁶, A. Shoamanesh²; the NAVIGATE.ESUS investigators

¹Alfried-Krupp Krankenhaus, Neurology, Essen, Germany

²PHRI, Neurology, Hamilton, Canada

³PHRI, Statistics, Hamilton, Canada

⁴Univ of Pennsylvana, Neurology, Philadelphia, USA

⁵University of Rome, Neurology, Rome, Italy

⁶Univ of Buenos Aires, Neurology, Buenos Aires, Argentina

⁷Bayer, Clinical Research, Berlin, Germany

⁸Univ. of Gothenburg, Neurology, Gothenburg, Sweden

⁹Univ of Perth, Neurology, Perth, Australia

¹⁰University of Lund, Neurology, Lund, Sweden

¹¹Bayer, Clinical Research, Berlin, USA

¹²National Institute of Neurology, Neurology, Mexico City, Mexico

¹³Selcuk University, Neurology, Selcuk, Turkey

¹⁴Univ of Glasgow, Neurosciences, Glasgow, United Kingdom

¹⁵Johnson&Johnson, Research, New York, USA

¹⁶Hospital Clinico de Barcelona, Neurology, Barcelona, Spain

¹⁷University of Alberta, Neurology, Calgary, Canada

ESUS is a recent concept that identifies a subgroup of cryptogenic strokes based on neuroimaging, a defined set of diagnostic tests and exclusion of certain etiologies. Despite the high annual stroke recurrence rate after ESUS, little is known about the etiology underlying recurrent strokes. We analysed recurrent ischemic strokes (IS) reported in NAVIGATE-ESUS.

NAVIGATE-ESUS was a randomised, controlled trial comparing efficacy and safety of rivaroxaban 15 mg versus aspirin 100 mg in patients with ESUS. IS were validated by adjudicators blinded to group assignment and classified by local investigators into the categories ESUS or non-ESUS (comprising lacunar, cardioembolic, atherosclerotic, other categories or insufficient testing).

309 IS were identified during the mean follow-up period of 11.0 months. The annual recurrence rate of IS was 4.6%/y in both groups. Sufficient diagnostic testing was unavailable in 39 patients. 156/270 (57.8%) of classifiable recurrent strokes were ESUS while 114/270 were non-ESUS recurrent strokes (table). Risk of recurrence by etiologies did not differ between treatments except for cardioembolism.

The majority of recurrent strokes after ESUS were deemed to be embolic but source remained undetermined in >80%. Recurrent strokes classified as cardioembolic were reduced by rivaroxaban.

Supported by Bayer

clinical trials.gov: 02313909

SUPRAVENTRICULAR EXTRASYSTOLES ON STANDARD 12-LEAD ELECTROCARDIOGRAM PREDICT ATRIAL FIBRILLATION AFTER EMBOLIC STROKE OF UNDETERMINED SOURCE

G. Ntaios¹, K. Perlepe², G. Sirimarco³, D. Strambo³, A. Eskandari³, E. Karagkiozi¹, A. Vemmou⁴, E. Koroboki⁵, E. Manios⁶, K. Makaritsis¹, P. Michel⁷ and K. Vemmos⁸

¹University of Thessaly, Department of Medicine, Larissa, Greece

²University of Thessaly- Larissa- Greece, Department of Medicine, Λάρισα, Greece

³Centre Hospitalier Universitaire Vaudois and University of Lausanne, Stroke Center and Neurology Service- Department of Clinical Neurosciences, Lausanne, Switzerland

⁴Medical School of Athens- Alexandra Hospital, Department of Clinical Therapeutics, Athens, Greece

⁵National and Kapodistrian University of Athens, Department of Neurology, Athens, Greece

⁶Medical School of Athens- Alexandra Hospital, Department of Clinical Therapeutics, Larissa, Greece

⁷Centre Hospitalier Universitaire Vaudois and University of Lausanne- Lausanne, Stroke Center and Neurology Service- Department of Clinical Neurosciences, Lausanne, Switzerland

⁸Metropolitan General Hospital, Department of Medicine, Athens, Greece

We analyzed consecutive patients with embolic stroke of undetermined source (ESUS) from three stroke registries to assess whether the presence of supraventricular extrasystoles (SVE) on standard 12-lead electrocardiogram (ECG) is associated with the detection of atrial fibrillation (AF), stroke recurrence and death during follow-up.

We measured the number of SVEs in all available ECGs of patients hospitalized for ESUS. Multivariate stepwise regression with forward selection of covariates assessed the association between SVE (classified in four groups according to their number per 10seconds of ECG: no SVE, >0-1SVEs, >1-2SVEs and >2SVEs) and outcomes during follow-up. The Kaplan–Meier product limit method estimated the 10-year cumulative probabilities of outcomes in each group. We calculated the negative prognostic value (NPV) of the presence of any SVE to predict AF.

Among 853 ESUS patients followed for 2857 patient-years, 226 (26.5%) patients had at least one SVE. AF was detected in 125 (14.7%) of patients: 8.9%, 22.5%, 28.1% and 48.3% in patients with no SVE, >0-1SVE, >1-2SVE and >2SVE respectively. In multivariate regression analysis, compared to patients with no SVEs, the corresponding hazard-ratios were 1.80 [95% confidence intervals (95% CI):1.06-3.05], 2.26 (95% CI:1.28-4.01) and 3.19 (95% CI:1.93-5.27). The NPV of the presence of any SVE for the prediction of new AF was 91.4%. There was no statistically significant association of SVE with the risk of ischemic stroke recurrence and death.

In ESUS patients with no SVEs during hospitalization, the probability that AF will not be detected during a follow-up of 3.4 years is > 91%.

N/A

YIELD AND PROGNOSTIC SIGNIFICANCE OF ROUTINE TRANSTHORACIC ECHOCARDIOGRAPHY IN TIA AND STROKE: A POPULATION-BASED COHORT STUDY

S. Lyons¹, J. Wall¹, G. Yiin¹, L. Nicola¹, M. Kubiak¹, P.M. Rothwell¹; on behalf of the Oxford Vascular Study

¹University of Oxford, Nuffield Department of Clinical Neurosciences, Oxford, United Kingdom

Echocardiography is recommended to identify a cardiac source of embolism causing TIA or stroke. However, the diagnostic yield of routine transthoracic echocardiography and any related prognostic value are uncertain.

Consecutive eligible consenting patients with TIA and stroke in a population-based study (the Oxford Vascular Study) underwent transthoracic echocardiography within 1 month of the event. High- and medium-risk echocardiographic findings were identified based on the TOAST classification.

Among 906 eligible patients a high-risk cardiac source of embolism was identified in 14 cases (1.5%), while at least one medium-risk source was identified in 202 (22.3%) cases, including moderate or severe left atrial dilatation in 99 (10.9%). High-risk cardiac source of embolism was more common if there was a history of atrial fibrillation (5.4%; p = 0.021), or myocardial infarction (6.8%; p = 0.022), but was found in only 0.6% of patients in the absence of symptomatic cardiac disease. During 4287 patient-years of follow-up 47 patients had a recurrent stroke. The presence of a high-risk cardiac source of embolism was associated with an increased risk of recurrent stroke (HR = 8.71, 95% CI 3.00-25.28), but the presence of a medium-risk source was not (HR = 0.89, 0.44-1.81), although moderate or severe left atrial dilatation was associated with an increased risk of recurrent ischaemic stroke after adjustment for age, sex and known atrial fibrillation (HR 2.63, 1.24-5.56).

In the absence of symptomatic cardiac disease, routine transthoracic echocardiography has a low yield of findings which might alter management, but it does provide some useful prognostic information.

N/A

PREVALENCE AND OVERLAP OF POTENTIAL EMBOLIC SOURCES IN PATIENTS WITH EMBOLIC STROKE OF UNDETERMINED SOURCE

G. Ntaios¹, K. Perlepe¹, D. Lambrou¹, G. Sirimarco², D. Strambo², A. Eskandari², E. Karagkiozi¹, A. Vemmou³, E. Koroboki³, E. Manios³, K. Makaritsis¹, K. Vemmos³ and P. Michel²

¹University of Thessaly- Larissa- Greece, Department of Medicine, Larissa, Greece

²Centre Hospitalier Universitaire Vaudois and University of Lausanne, Stroke Center and Neurology Service- Department of Clinical Neurosciences, Lausanne, Switzerland

³Medical School of Athens- Alexandra Hospital, Department of Clinical Therapeutics, Athens, Greece

We aimed to assess the prevalence and degree of overlap of potential embolic sources (PES) in patients with embolic stroke of undetermined source (ESUS).

In a pooled dataset derived from three prospective stroke registries, patients were categorized in one or more groups according to the PES that was/were identified. We categorized PES as follows: atrial cardiopathy, arterial disease, left ventricular (LV) disease, cardiac valvular disease, patent foramen ovale and cancer.

In 800 ESUS patients (43.1% women, median age 67.0 years), three most prevalent PES were LV disease, arterial disease and atrial cardiopathy, which were present in 54.4%, 48.5% and 48.1% of patients respectively. The majority of patients (60.6%) had more than one PES, whereas only 34.7% and 4.7% of patients had a single or no PES respectively. In 26.7% of patients, there were three or more PES present. On average, each patient had two PES (median of 2). During a mean follow-up of 3.7 years (2,934 patient-years), stroke recurrence occurred in 101 (12.6%) patients (23.3 recurrences per 100 patient-years). In multivariate analysis, the risk of stroke recurrence was not statistically different across PES groups or between patients with 0–1, 2 or ≥3 PES.

There is major overlap of PES in patients with ESUS. This may possibly explain the negative results of the NAVIGATE ESUS and RE-SPECT ESUS trials and offer a rationale for a randomized controlled trial of combination of anticoagulation and aspirin for the prevention of stroke recurrence in patients with ESUS.

N/A

DERIVATION AND EXTERNAL VALIDATION OF A SCORE TO IDENTIFY PATIENTS WITH EMBOLIC STROKE OF UNDETERMINED SOURCE AT HIGH RISK FOR STROKE RECURRENCE

G. Ntaios¹, G. Georgiopoulos², K. Perlepe¹, G. Sirimarco³, E. Cuadrado-Godia⁴, A. Rodríguez-Campello⁴, V. Arnao⁵, M. Paciaroni⁵, E. Diez-Tejedor⁶, B. Fuentes⁶, A. Arauz⁷, S.F. Ameriso⁸, M. Gómez-Schneider⁸, M.A. Barboza⁹, A.M. Iglesias Mohedano¹⁰, A. García Pastor¹⁰, J. Putaala¹¹, T. Tatlisumak^12,13, K. Vemmos² and P. Michel³

¹University of Thessaly, Department of Internal Medicine, Larissa, Greece

²Medical School of Athens- Alexandra Hospital- Athens- Greece, Department of Clinical Therapeutics, Athens, Greece

³Centre Hospitalier Universitaire Vaudois and University of Lausanne, Stroke Center and Neurology Service- Department of Clinical Neurosciences-, Lausanne, Switzerland

⁴Hospital del Mar. Neurovascular Research Group- IMIM-Hospital del Mar Institut Hospital del Mar d’ Investigacions Mèdiques- Universitat Autònoma de Barcelona, Stroke Unit- Department of Neurology, Barcelona, Spain

⁵University of Perugia, Stroke Unit, Perugia, Italy

⁶La Paz University Hospital – Autónoma University of Madrid- IdiPAZ Health Research Institute, Department of Neurology and Stroke Center-, Madrid, Spain

⁷Instituto Nacional de Neurología y Neurocirugía Manuel Velasco Suárez, Stroke Clinic, Mexico City, Mexico

⁸Institute for Neurological Research- FLENI, Department of Neurology, Buenos Aires, Argentina

⁹Hospital Dr. Rafael A. Calderón Guardia- CCSS. University of Costa Rica, Neurosciences Department, Costa Rica, Costa Rica

¹⁰Hospital General Universitario Gregorio Marañón- IiSGM Health Research Institute- Universidad Complutense de Madrid, Vascular Neurology Section- Stroke Center, Madrid, Spain

¹¹Helsinki University Hospital and University of Helsinki, Department of Neurology-, Helsinki, Finland

¹²Department of Clinical Neurosciences- Institute of Neuroscience and Physiology & Department of Neurology, Sahlgrenska Academy at University of Gothenburg, Gothenburg, Sweden

¹³Helsinki University Hospital and University of Helsinki, Department of Neurology, Helsinki, Finland

A tool to stratify the risk of stroke recurrence after Embolic Stroke of Undetermined Source (ESUS) could be useful for research and clinical practice. We aimed to develop and externally validate a score for the identification of ESUS patients at high risk for stroke recurrence.

We pooled the data of consecutive ESUS patients from 11 prospective stroke registries. We performed multivariable Cox regression analysis in 3 registries (n = 884) to identify predictors of ischemic stroke recurrence. Based on the coefficients of the fitted multivariable model, we generated an integer-based point system. We validated the score externally in 8 other registries (n = 820).

In the derivation cohort, age, leukoaraiosis and multiterritorial infarcts (defined as acute and/or chronic non-lacunar multiterritorial infarcts) emerged as independent predictors of stroke recurrence. The final score assigns 1 point per every decade after 35 years of age, 2 points for leukoaraiosis and 3 points for multiterritorial infarcts. The rate of recurrences was 2.1/100 patient-years (95% CI:1.44-3.06) in patients with a score of 0–4 (low risk), 3.74 (95% CI:2.77-5.04) with a score of 5–6 (intermediate risk) and 8.23 (95% CI:5.99-11.3) with a score of 7–12 (high risk). Compared to low-risk patients, recurrence risk was significantly higher in intermediate-risk (HR:1.78, 95% CI:1.1-2.88) and high-risk patients (HR:4.67, 95% CI:2.83-7.7). The score was well calibrated in both derivation and external validation cohorts [Hosmer-Lemeshow χ2:12.1 (p = 0.357) and χ2:21.7 (p = 0.753), respectively]. The AUC of the score were 0.63 (95% CI:0.58-0.68) and 0.60 (95% CI:0.54-0.66), respectively.

The proposed score can assist in the identification of ESUS patients at high recurrence risk.

N/A

PATENT FORAMEN OVALE CLOSURE IN STROKE PATIENTS WITH MIGRAINE IN THE CLOSE TRIAL

J.L. Mas¹, A. Charles-Nelson², G. Chatellier²; on behalf of the CLOSE trial investigators

¹Hopital Sainte-Anne- Paris- France & Université Paris Descartes & INSERM U1266, Neurology, Paris, France

²Georges Pompidou European Hospital- AP-HP- INSERM Centre d’investigation clinique 1418- Paris., Epidemiology and Clinical Research Unit, Paris, France

The association between migraine and PFO has led some to propose PFO closure to reduce the risk of migraine attacks, but the benefit of this strategy remains controversial.

The CLOSE trial compared transcatheter PFO closure plus antiplatelet therapy to antiplatelet therapy alone to prevent stroke recurrence in patients 16 to 60 years old with a PFO and an otherwise unexplained ischemic stroke (NCT00562289). Migraine with or without aura according to International Headache Society criteria was systematically recorded at inclusion. Number and type of migraine attacks, as well as preventive and acute treatments were recorded at each follow-up visit.

Among 473 patients randomized to PFO closure or antiplatelet therapy, 145 (30.7%; mean age, 41.9 +/- 10.0 years; women 58.6%) had migraine, including 75 with migraine with aura. 67 were randomized to PFO closure and 78 to antiplatelet therapy. The 2 groups had comparable baseline characteristics, except for a higher proportion of women and migraine with aura in the PFO closure group. During a median follow-up of 5.5 years, there were no differences between PFO closure and antiplatelet therapy in the mean annual number of migraine attacks (12.5 ± 20.0 vs 10.9 ± 14.5, p = 0.73) and the use of preventive treatments (21 [31.3%] vs 19 [24.4%)], p = 0.45). Likewise, no between-group differences in these criteria were found in patients with migraine with aura and in those (39.4%) with substantial or severe migraine-related disability at inclusion (HIT-6 score > = 56).

In young and middle-aged adults with PFO-associated stroke and migraine, PFO closure did not reduce migraine attacks.

NCT00562289

CLOSURE OR MEDICAL THERAPY IN PATIENTS WITH PATENT FORAMEN OVALE AND CRYPTOGENIC STROKE OR TRANSIENT ISCHEMIC ATTACK: A PROSPECTIVE INTERVENTIONAL CASE STUDY

S. Poli¹, E. Siebert², F. Haertig¹, K. Mueller², W. Bocksch², M. Gawaz², U. Ziemann¹ and C. Zuern³

¹University Hospital Tuebingen, Neurology, Tuebingen, Germany

²University Hospital Tuebingen, Cardiology, Tuebingen, Germany

³University Hospital Basel, Cardiology, Basel, Switzerland

Studies investigating closure of patent foramen ovale (PFO) in patients with cryptogenic ischemic stroke or transient ischemic attack (CIS/TIA) have shown inconclusive results; identification of patients who will benefit from closure remains challenging. We aim to introduce an SOP for interventional and medical management of PFO-CIS/TIA patients and compare outcomes of patients ≤60 years who underwent PFO closure with results of recent RCT, and patients >60 years who underwent closure with patients >60 years with high-risk PFO treated with medical therapy alone (MTA).

Prospective cohort study including consecutive PFO-CIS/TIA patients at the stroke unit of Tuebingen university hospital, Germany. According to SOP, patients underwent closure in case ≤70 years and high-risk PFO, i.e. PFO with either presence of atrial septal aneurysm, spontaneous, or high-grade right-to-left shunt during Valsalva. Follow-up was performed after one year for assessment of the primary endpoint (recurrent IS or intracranial hemorrhage) and secondary endpoints.

236 consecutive PFO-CIS/TIA patients were included. During follow-up of 2.6 ± 1.3 years, recurrent IS rate after closure was 2.9% in patients ≤60 years. In patients with high-risk PFO >60 years who underwent closure, stroke recurrence rate was 7% vs 3.8% in MTA patients. No intracranial hemorrhage was observed.

Results of our real-world study of PFO patients with CIS/TIA were comparable to recent RCT in patients ≤60 years. However, patients >60 years who underwent closure of high-risk PFO had higher rates of recurrent IS vs patients treated with MTA and might therefore not profit from the intervention.

Tuebingen University Ethics Committee protocol no. 522/2012BO2

INFLUENCE OF AGE ON PROGNOSIS OF CRYPTOGENIC STROKE IN PATIENTS WITH PATENT FORAMEN OVALE ON MEDICAL TREATMENT ALONE: SYSTEMATIC REVIEW AND META-REGRESSION

S. Mazzucco¹, R. Luengo-Fernandez¹ and P.M. Rothwell¹

¹University of Oxford, Nuffield Department of Clinical Neurosciences, Oxford, United Kingdom

Results of published trials on patent foramen ovale (PFO) closure versus medical therapy alone after cryptogenic stroke apply to patients aged ≤60 years. Since untreated prognosis in older patients is uncertain, we estimated the absolute risk of stroke recurrence after cryptogenic TIA/stroke in relation to age, and the relative risk of stroke in patients with PFO vs no-PFO stratified by age.

We did a systematic review of studies of cryptogenic TIA or stroke with >100 patients reporting ischaemic stroke recurrence on medical therapy alone in patients with PFO or in patients with PFO vs no-PFO. Absolute risk of stroke (/100-patient-years) was related to mean study cohort age by linear regression weighted by 1/variance of the risk estimate. Odds ratio for stroke recurrence in patients with PFO vs no PFO was determined by meta-analysis stratified by age (</≥ 65).

Among 22 eligible cohorts identified, mean (SD) stroke risk per 100-patient-year was 2.33 (1.53), but there was heterogeneity between studies (p < 0.001), which was explained by higher risk with increasing mean study age (meta-regression p < 0.001). In studies of patients with PFO vs no PFO, risk of stroke recurrence was only increased in patients with PFO at age ≥65 years (OR = 1.94, 1.15-3.29, p = 0.01, phet = 0.65).

The absolute risk of stroke recurrence after cryptogenic TIA/stroke in patients with PFO increases with age, as does the relative risk of stroke in patients with PFO vs no-PFO. Future trials on secondary prevention of stroke in PFO should include older patients.

N/A

Scientific Communications 04-Sex Differences in Stroke

SEX DIFFERENCES IN TREATMENT AND OUTCOMES AFTER STROKE:A POOLED ANALYSIS OF FIVE CLINICAL TRIALS INCLUDING TWENTY THOUSAND PARTICANTS

C. Carcel¹, X. Wang¹, E.C. Sandset², C. Delcourt¹, H. Arima¹, R. Lindley³, M.L. Hackett¹, P. Lavados⁴, T.G. Robinson⁵, P. Muñoz-Venturelli⁴, V.V. Olavarría⁴, A. Brunser⁴, E. Berge⁶, J. Chalmers¹, M. Woodward¹ and C.S. Anderson¹

¹The George Institute for Global Health, University of New South Wales, Sydney, Australia

²Oslo University Hospital, Department of Neurology, Oslo, Norway

³University of Sydney, The George Institute for Global Health and Westmead Clinical School, Sydney, Australia

⁴Universidad del Desarrollo, Unidad de Neurología Vascular, Santiago, Chile

⁵University of Leicester, Department of Cardiovascular Sciences, Leicester, United Kingdom

⁶Oslo University Hosptal, Department of Internal Medicine, Oslo, Norway

To explore sex differences in the management and outcome after stroke using a large sample with high-quality international trial data.

Individual participant data were obtained from five randomized controlled trials in patients with acute stroke with data on demographics, stroke-related factors, pre-stroke health, medication use, in-hospital treatment and outcome. Study-specific crude and adjusted models were used to estimate differences in management and outcomes for women compared with men, and then pooled using random-effects meta-analysis.

There were 19,652 participants included, of whom 7721 (40%) were women. After multivariable adjustments, women had a higher survival rate at 3–6 months for ischemic (odds ratio [OR] 0.83; 95% confidence interval [CI] 0.70-0.97) stroke. Women with ischemic stroke had a higher likelihood of being disabled (OR 1.19; 95% CI 1.03-1.36) at 3–6 months follow up. Whilst in hospital, women were more likely to be admitted to an acute stroke unit (OR 1.17, 95% CI 1.01-1.34). Men were more likely to be intubated (OR 0.72, 95% CI 0.53-0.99), treated for fever (OR 0.82, 95% CI 0.70-0.95) and admitted to an intensive care unit (OR 0.86, 95% CI 0.76-0.98). Women had higher odds of being prescribed antihypertensive agents (OR 1.22, 95% CI 1.13-1.31) and lower odds of being prescribed antiplatelets (OR 0.86; 95% CI 0.79-0.93), glucose lowering (OR 0.86; 95% CI 0.78-0.94) and lipid lowering agents (OR 0.85; 95% CI 0.77-0.94).

Women who had ischemic stroke had better survival but were also more disabled and had poorer quality of life. Variations in management may explain the disparities

N/A

SEX DIFFERENCES IN THE PRESENTATION AND OUTCOMES AFTER TRANSIENT ISCHEMIC ATTACKS AND MINOR STROKES

A.Y.X. Yu¹, A.M. Penn², M.L. Lesperance³, N.S. Croteau³, R.F. Balshaw⁴, K. Votova⁵, M.B. Bibok⁶, M. Penn², V. Saly², J. Hegedus², C. Zerna⁷, E. Klourfeld⁸, L. Bilston⁸, Z.M. Hong⁸, S.B. Coutts⁸; on behalf of the SpecTRA study group

¹University of Toronto-SHSC, Medicine, Toronto, Canada

²Island Health, Stroke Rapid Assessment Unit, Victoria, Canada

³University of Victoria, Department of Mathematics and Statistics, Victoria, Canada

⁴University of Manitoba, George & Fay Yee Centre for Healthcare Innovation, Winnipeg, Canada

⁵University of Victoria, Division of Medical Sciences, Victoria, Canada

⁶Island Health, Department of Research and Capacity Building, Victoria, Canada

⁷University of Calgary, Department of Community Health Sciences, Calgary, Canada

⁸University of Calgary, Department of Clinical Neurosciences, Calgary, Canada

Sex differences have been described in the presentation, care, and outcomes of patients with major strokes, but these differences are less understood after minor events. We examined sex differences in the symptoms (focal versus non-focal), diagnoses, and outcomes of patients with acute minor or transient neurological events.

In a sub-study of SpecTRA (Spectrometry for Transient Ischemic Attack Rapid Assessment), a prospective multi-center cohort study of patients with transient ischemic attacks, minor strokes, or stroke mimics, we described the presenting symptoms (focal versus non-focal) by sex. We examined the association between sex and outcome using logistic regression, adjusting for age, hypertension, diabetes, atrial fibrillation, coronary artery disease, prior stroke, and smoking. Our primary outcome was diagnosis (cerebral ischemia versus mimic). Secondary outcomes were recurrent stroke and composite outcome of stroke, myocardial infarction, or death at 90 days.

We included 1,648 patients (47% female) with median age 70 [IQR 59,80] years; 92% were investigated with MRI. A similar proportion of men and women experienced non-focal symptoms (chi-square p = 0.26). Compared to men, women were less likely diagnosed with cerebral ischemia (68% versus 77%; adjusted odds ratio, aOR = 0.64; 95% CI [0.51,0.81]), but the 90-day risks of stroke or composite outcome were similar (stroke aOR = 0.90, 95% CI [0.47,1.69]; composite aOR = 0.85, 95% CI [0.53,1.34]).

Women and men presented with similar symptoms, but were less often diagnosed with cerebral ischemia. Nevertheless, they had similar risks of subsequent vascular events. This suggests missed opportunities to prevent vascular events in women.

N/A

SEX DIFFERENCES IN STROKE FEATURES AND OUTCOMES. EVIDENCES FROM A LARGE SPANISH COHORT STUDY

M. Alonso de Leciñana¹, M.D.M. Castellanos², Ó. Ayo³, A. Morales⁴, J. Rodríguez-Antigüedad⁵, M. Millán⁶, J.J. Mengual⁷, J. Serena⁸, J.F. Arenillas⁹, B. Fuentes¹, F. Purroy¹⁰, J. Krupinski¹¹; The Investigators of the RENISEN Registry. Stroke Project. Spanish Society of Neurology Stroke Group.

¹University Hospital La Paz, Neurology. Stroke Center, Madrid, Spain

²Complejo Hospitalario Universitario A Coruña, Neurology, A Coruña, Spain

³Complejo Hospitalario Universitario de Albacete, Neurology, Albacete, Spain

⁴Hospital Clínico Universitario Virgen de la Arrixaca, Neurology, Murcia, Spain

⁵Hospital Universitario Donostia, Neurology, San Sebastian, Spain

⁶Hospital Germans Trias i Pujol, Neurology, Badalona, Spain

⁷Hospital Moises Broggi., Neurology, Sant Joan Despí, Spain

⁸Hospital Universitari DrJosep Trueta, Neurology, Girona, Spain

⁹Hospital Universitario de Valladolid, Neurology, Valladollid, Spain

¹⁰Hospital Universitari Arnau de Vilanova, Neurology, Lleida, Spain

¹¹Hospital Universitari Mutua de Terrassa, Neurology, Terrasa, Spain

To analyze differences in clinical characteristics and outcomes of stroke between men and women in the Spanish population.

Data from the Spanish Stroke National Registry RENISEN including patients hospitalized from 34 centers (January/2011-January/2018). Age, gender, risk factors, stroke etiology, severity (NIHSS), outcomes (mRS) and destination at discharge were recorded. The prognostic role of sex was explored in a multivariate logistic regression analysis.

14599 men and 10567 women. Median age (IQR): 71 (18) vs 78 (16), p < 0.01. 11% were cerebral hemorrhages (ICH), 73% ischemic strokes (IS), 10% TIA. Hypertension and atrial fibrillation were more frequent among women (73% vs 69%, p < 0.01 and 24% vs 16.5%, p < 0.01, respectively), whereas tobacco use (22% vs 8%, p < 0.01) and diabetes (31% vs 27%, p < 0.01) were more frequent among men. Cardiac embolism was the most frequent cause of IS/TIA in women (40% vs 29%, p < 0.01) and large vessel atherosclerosis in men (25% vs 14%, p < 0.01). Women suffered more severe strokes [median baseline NIHSS (IQR)]: [7 (14) vs 5(11), p < 0.01], worse outcome (mRS 0–2): 46% vs 57% (p < 0.01) and returned home less frequently: 52% vs 60% (p < 0.01). After adjustment for confounders, female sex was an independent predictor of death or dependency [OR (95% CI)]: 1.25 (1.18-1.33)

Differences in risk factors, severity and outcomes of stroke are confirmed between men and women in the Spanish population, female sex being and independent predictor of poor outcome. Raising awareness of these differences may help implement specific action plans for prevention and management.

N/A

GENDER DIFFERENCES IN THE USE OF HEALTH CARE RESOURCES FOR ACUTE ISCHEMIC STROKE- RESULTS FROM THE FEDERAL GERMAN DRG STATISTICS

C. Krogias¹, R. Weber², D. Bartig³, S. Meves⁴, V. Caso⁵, W. Hacke⁶ and J. Eyding⁷

¹Ruhr University Bochum, Neurology- St. Josef-Hospital, Bochum, Germany

²Alfried-Krupp-Krankenhaus Essen, Neurology, Essen, Germany

³drg market, drg market, Osnabrück, Germany

⁴Katholische Kliniken Ruhrhalbinsel, Neurology, Essen, Germany

⁵University of Perugia- Santa Maria della Misericordia Hospital, Stroke Unit, Perugia, Italy

⁶University of Heidelberg, Neurology, Heidelberg, Germany

⁷Kliniken Dortmund, Neurology, Dortmund, Germany

The federal database of DRG statistics contains over 1.25 million hospitalized ischemic stroke patients 2013–2017.

We assessed gender related differences in the access to stroke care in the 2017 database, regarding access to stroke unit (SU), intravenous thrombolysis (IVT), mechanical thrombectomy (MT), early neurological or geriatric rehabilitation, and in-hospital mortality. P-values of < 0.001 were considered as statistically significant.

In 2017, a total of 227,757 ischemic stroke patients were treated in German hospitals (118.896 male vs. 108.861 female). Men had strokes at younger ages while more women had strokes in the age groups over 80 years (p < .0001). Access to SU was significantly lower for women (74.5% male vs 71.9% female, p < .001). Transfer to early rehabilitation was significantly more frequent for women (13.6% male vs 17.8% female, p < .001). The use of IVT did not differ between men and women overall (IVT 15.8% male vs 16.0% female, n.s.) and in all age groups. Surprisingly there was a highly significant difference in the use of MT in favor of women overall (5.2% male vs 6.5% women; p < .001), and in all age groups. Women had an higher mortality rate (5.8% male, 8.8% women, p < 001) with the difference more pronounced with older age.

Overall, SU admissions and recanalizing treatment rates are high in Germany. Women received MT significantly more frequent in all age groups. Women showed a significantly higher in-hospital mortality. However, independent of age, access to SU treatment was significantly lower for women than for men.

n/a

SEX DIFFERENCES IN VULNERABLE PLAQUE COMPOSITION IN PATIENTS WITH A MILD-TO-MODERATE CAROTID ARTERY STENOSIS

D. van Dam-Nolen^1,2, N.C.M. van Egmond¹, K. Dilba¹, G. Crombag³, C. Lucci⁴, M. Liem⁵, M.E. Kooi³, J. Hendrikse⁴, P.J. Nederkoorn⁵, P.J. Koudstaal², D. Bos^1,6 and A. van der Lugt¹

¹Erasmus University Medical Center Rotterdam, Department of Radiology and Nuclear Medicine, Rotterdam, The Netherlands

²Erasmus University Medical Center Rotterdam, Department of Neurology, Rotterdam, The Netherlands

³CARIM School for Cardiovascular Diseases- Maastricht University Medical Center, Department of Radiology and Nuclear Medicine, Maastricht, The Netherlands

⁴University Medical Center Utrecht, Department of Radiology, Utrecht, The Netherlands

⁵Academic Medical Center, Department of Neurology, Amsterdam, The Netherlands

⁶Erasmus University Medical Center Rotterdam, Department of Epidemiology, Rotterdam, The Netherlands

For yet not fully understood reasons, stroke incidence rate is higher in men compared to women < 75 years of age. Carotid atherosclerosis is a major cause of ischemic strokes. Several plaque components are important determinants for plaque progression and rupture. Sex differences in these so-called vulnerable plaque characteristics, including intraplaque haemorrhage (IPH), lipid-rich necrotic core (LRNC), thin or ruptured fibrous cap (TRFC), and ulcerations, may help further explain sex differences in strokes. This study aims to analyse sex differences in plaque composition in symptomatic patients.

We selected 238 patients from the Plaque At RISK-study (PARISK), which included patients with recent ischemic symptoms and < 70% ipsilateral carotid artery stenosis. Plaque characteristics were assessed with MDCTA (n = 201) and MRI (n = 224). Multivariable logistic regression analyses were used to assess the effects of sex on presence and volume of plaque characteristics. Differences in volumes were analysed comparing quartile (Q) 4 with Q1-3.

Women had significant lower plaque burden (OR = 0.10, 95% CI:0.03-0.27). Women had also lower prevalence and volume of IPH (OR = 0.30, 95% CI:0.13-0.65; OR = 0.19, 95% CI:0.05-0.52) and LRNC (OR = 0.42, 95% CI:0.22-0.80; OR = 0.35, 95% CI:0.11-0.92) after adjusting for plaque burden. This association remained significant for IPH volume (OR = 0.26, 95% CI:0.07-0.78) and LRNC presence (OR = 0.41, 95% CI:0.19-0.88), but not for IPH presence (OR = 0.45, 95% CI:0.18-1.07) and LRNC volume (OR = 0.41, 95% CI:0.12-1.21) after adjusting for cardiovascular risk factors and medication. Sex was not associated with presence of TRFC, ulcerations and calcifications.

Vulnerable plaque components like IPH and LRNC are less common in women compared to men in symptomatic patients with mild-to-moderate carotid stenosis.

N/A

SEX DIFFERENCES IN CLOT EXTENT, ATHEROSCLEROSIS, COLLATERALS AND TISSUE DAMAGE IN ACUTE ISCHEMIC STROKE PATIENTS TREATED WITH ENDOVASCULAR THROMBECTOMY

A. van der Meij¹, M.L.E. Bernsen², G. Holswilder³, J. Hofmeijer⁴, F.H.M. Spaander⁵, P.J. Nederkoorn⁶, J.M.M. Martens², I.R. van den Wijngaard^1,7, H.F. Lingsma⁸, P.R. Konduri⁹, C.B.L.M. Majoie⁹, W.J. Schonewille¹⁰, N.D. Kruyt¹, M.A.A. van Walderveen³, M.J.H. Wermer¹; on behalf of the MR CLEAN Registry investigators

¹Leiden University Medical Center, Neurology, Leiden, The Netherlands

²Rijnstate Hospital, Radiology, Arnhem, The Netherlands

³Leiden University Medical Center, Radiology, Leiden, The Netherlands

⁴Rijnstate Hospital, Neurology, Arnhem, The Netherlands

⁵Zaans Medical Center, Neurology, Zaandam, The Netherlands

⁶Amsterdam University Medical Center- location Amsterdam Medical Center, Neurology, Amsterdam, The Netherlands

⁷Haaglanden Medical Center, Neurology, Den Haag, The Netherlands

⁸Erasmus MC University Medical Center, Public Health, Rotterdam, The Netherlands

⁹Amsterdam University Medical Center- location Amsterdam Medical Center, Radiology, Amsterdam, The Netherlands

¹⁰Sint Antonius Hospital, Neurology, Nieuwegein, The Netherlands

Previous studies suggest that women have poorer outcomes after endovascular treatment (EVT) than men, which might be explained by clot extent, intracranial atherosclerosis, collateral status or early ischemic tissue damage. We aimed to assess sex differences in these factors.

We included patients from the MR CLEAN Registry who underwent EVT between 2014–2016. We assessed the association between sex and clot location (intracranial internal carotid artery (ICA), ICA-T, M1, M2, other), clot burden score (CBS), intracranial atherosclerosis (absent or present), collateral status (poor (<50%) or good (≥50%) filling), early tissue damage (Alberta Stroke Program Early CT Score (ASPECTS) < 8) and good outcome (modified Rankin Scale ≤2) after 90 days with logistic regression with adjustments for relevant confounders.

1488 patients were included, median age was 71 years and 47% were women. Clot location was similar in men and women, except that the ICA was less common in women (Odds Ratio (OR) = 0.51,95% CI = 0.31–0.82) and the M1 segment more prevalent (OR = 1.22,95% CI = 0.99–1.50). CBS was similar in men and women (adjusted Odds Ratio (aOR) = 1.06,95% CI = 0.85–1.33). Intracranial atherosclerosis was less prevalent in women than in men (aOR = 0.61,95% CI = 0.48–0.78). Furthermore, poor collateral status was less common in women compared with men (aOR = 0.62,95% CI = 0.49–0.78). There was no difference in early tissue damage between women and men (aOR = 0.88,95% CI = 0.70–1.10). Outcome was worse in women than in man (33% vs. 40% good outcome).

Women seem to have comparable clot extent and early ischemic tissue damage, less atherosclerosis and better collaterals than men. Therefore, these factors are unlikely to explain the poorer outcome after EVT in women.

N/A

SEX RELATED RUPTURE RATE OF UNRUPTURED INTRACRANIAL ANEURYSMS: AN INDIVIDUAL PATIENT DATA META-ANALYSIS

C.C.M. Zuurbier¹, L.A. Mensing¹, M.J.H. Wermer², S. Juvela³, A.E. Lindgren⁴, J.E. Jääskeläinen⁴, T. Koivisto⁴, A. Morita⁵, H. Arai⁶, K. Nozaki⁷, Y. Murayama⁸, T. Ishibashi⁸, H. Takao⁸, G.J.E. Rinkel¹, Y.M. Ruigrok¹ and J.P. Greving⁹

¹University Medical Center Utrecht, Neurology, Utrecht, The Netherlands

²Leiden University Medical Center, Neurology, Leiden, The Netherlands

³University of Helsinki, Clinical Neurosciences, Helsinki, Finland

⁴Kuopio University Hospital- and Institute of Clinical Medicine- School of Medicine- Faculty of Health Sciences- University of Eastern Finland, Neurosurgery of NeuroCenter, Kuopio, Finland

⁵UCAS Japan Coordinating Office- University of Tokyo- Nippon Medical School, Neurological Surgery, Tokyo, Japan

⁶Juntendo University- Medical School, Neurosurgery, Tokyo, Japan

⁷Shiga University of Medical Science, Neurosurgery, Shiga, Japan

⁸Tokyo Jikei University School of Medicine, Endovascular Neurosurgery, Endovascular Neurosurgery, Japan

⁹University Medical Center Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands

Women have a higher risk of rupture of unruptured intracranial aneurysms (UIA), but female sex is not an established independent risk factor. This suggests that the higher rupture risk in women may be explained by a higher prevalence of patient- or aneurysm-related risk factors for rupture. We studied sex differences in rupture rate and in patient- and aneurysm-related risk factors for UIA rupture.

We pooled individual patient data for 9,195 patients (6,018 women, 65%) with 11,278 UIA and 22.657 person-years follow-up from 6 prospective cohort studies. We analyzed rupture rates per patient with a Cox proportional hazard regression model, stratified per cohort and adjusted for all items of the PHASES score (population, hypertension, age, size of aneurysm, earlier subarachnoid hemorrhage from another aneurysm, site of aneurysm) and smoking.

Rupture occurred in 159 women (1.1%/person-years) and 59 men (0.7%/person-years). Women were older (62.3 vs 59.6 years), more often had hypertension (44% vs 42%), were less often smokers (18% vs 43%), had more often internal carotid artery aneurysms (23% vs 16%) and had larger sized aneurysms (≥ 7 mm, 26% vs 24%) than men. The hazard ratio (HR) of rupture for women compared to men was 1.49 (95% CI: 1.11–2.02) and the adjusted HR 1.44 (95% CI 1.04–1.99).

The risk of UIA rupture in women is higher than in men; this sex difference is not explained by differences in the patient- and aneurysm-related risk factors that constitute the phases score.

N/A

STROKE RISK IN WOMEN WITH MIGRAINE USING HORMONAL CONTRACEPTION: A SYSTEMATIC REVIEW AND META-ANALYSIS

H. van Os¹, K.M. Linstra¹, M.D. Ferrari¹, O.M. Dekkers², F.M. Helmerhorst³, K.A. Ryan⁴, W.M. Lijfering², S.J. Kittner⁵ and M.J.H. Wermer¹

¹Leiden University Medical Center, Neurology, Leiden, The Netherlands

²Leiden University Medical Center, Epidemiology, Leiden, The Netherlands

³Leiden University Medical Center, Gynaecology, Leiden, The Netherlands

⁴University of Maryland School of Medicine, Department of Medicine, Baltimore, USA

⁵Baltimore Veterans Administration Medical Center and the University of Maryland School of Medicine, Neurology, Baltimore, The Netherlands

Migraine increases the risk of ischemic stroke approximately two-fold, especially in young women. Presence of both migraine and combined oral contraceptive (COC) use has been proposed to result in a strongly supra-additive increase in ischemic stroke risk with reported odds ratios (ORs) up to 17. We performed a meta-analysis to investigate this possible synergy between migraine and COC use.

This systematic review and meta-analysis was executed according to the PRISMA statement. Inclusion criteria were cohort or case-control design, women of reproductive age, information available on both migraine and hormonal contraception, and first ever ischemic stroke. We extracted adjusted odds ratios (aORs) from observational studies and performed a subanalysis based on estrogen dose.

We identified 782 eligible articles of which five studies (including 13170 women) met the inclusion criteria. Ischemic stroke risk in COC users versus women who did not use COC ranged from 1.22–4.40 (pooled aOR: 2.26;95% CI:1.28–3.98). Ischemic stroke risk in women with versus without migraine ranged from 1.03–3.45 (pooled aOR:1.65;95% CI:0.99–2.75). Ischemic stroke risk for women with migraine using COC versus women with neither risk factors present ranged from 2.01–16.90 (pooled aOR:4.95;95% CI:2.13–11.48). Ischemic stroke risk in women with both migraine and COC containing < 50 μg estrogen ranged from 1.80–6.59 (aOR:2.05;95% CI:1.48–2.83).

Migraine patients who use COC are at increased risk of ischemic stroke, but the combined risk is not higher than the additive risk of the separate risk factors.

N/A

ESTIMATION OF LESION AGE IN WAKE-UP STROKE PATIENTS – COMPARISON OF CT BASED QUANTITATIVE LESION WATER UPTAKE TO DWI-FLAIR MISMATCH ASSESSMENT

H. Leischner¹, F. Flottmann¹, U. Hanning¹, P. Sporns², A. Kemmling², T. Faizy¹, G. Thomalla³, J. Fiehler¹ and G. Broocks¹

¹University Medical Center Hamburg-Eppendorf – UKE, Department for Neuroradiological Diagnosis and Intervention, Hamburg, Germany

²University Hospital Münster, Department of Radiology, Münster, Germany

³University Medical Center Hamburg-Eppendorf – UKE, Department of Neurology, Hamburg, Germany

Magnetic resonance imaging (MRI) is frequently used to estimate lesion age in patients with wake-up stroke. However, MRI has a limited availability and feasibility, and is more time-consuming in the acute situation. Therefore, in this study we used computed tomography (CT) based quantification of lesion water uptake (NWU) to estimate the lesion age of the patients hypothesizing that NWU is not inferior to MRI in identifying patients with lesion age < 4.5 hours from symptom onset.

50 patients with acute anterior circulation stroke and known symptom onset were analyzed. Mismatch between diffusion-weighted MRI (DWI) and fluid-attenuated inversion recovery (FLAIR) was diagnosed by consensus reading. NWU was calculated in admission CT. An established cut-off for NWU was applied to distinguish patients within and beyond 4.5 hours in a blinded approach.

In 32 patients, the time from symptom onset to CT imaging was < 4.5h and 18 patients presented >4.5h (range: 0.5-7.8h). The median time from CT to MRI was 36 minutes (IQR: 24–55). DWI-FLAIR mismatch correctly assigned 34/50 patients (68%) to the corresponding time-window (i.e., ≤4.5 or >4.5 hours) with a sensitivity of 70% (95% CI: 51–85%) and specificity of 72% (95% CI: 47–90%). CT-based NWU correctly assigned 43/50 (86%) with a sensitivity of 91% (95% CI: 75–98%) and specificity of 78% (95% CI: 75–94%) using 11.5% as established threshold.

CT-based quantitative assessment of NWU was comparable to DWI-FLAIR mismatch in identifying patients within the thrombolysis time window. Future trials could use quantitative NWU as imaging biomarker to stratify wake-up stroke instead of MRI.

N/A

MIND THE HEART: ECG-GATED CARDIAC CT-ANGIOGRAPHY IN THE ACUTE PHASE OF ISCHEMIC STROKE

V. Guglielmi¹, N.S. Groeneveld¹, N.H.J. Lobé², S.M. Boekholdt³, A. van Randen², L.F.M. Beenen², H.A. Marquering^2,4, J. Schaafsma⁵, K. Hanneman⁶, F.L. Silver⁵, D.M. Mandell⁷, C.B.L.M. Majoie², Y.B.W.E.M. Roos¹, R.N. Planken² and J.M. Coutinho¹

¹Amsterdam UMC- University of Amsterdam, Neurology, Amsterdam, The Netherlands

²Amsterdam UMC- University of Amsterdam, Radiology and Nuclear Medicine, Amsterdam, The Netherlands

³Amsterdam UMC- University of Amsterdam, Cardiology, Amsterdam, The Netherlands

⁴Amsterdam UMC- University of Amsterdam, Biomechanical Engineering and Physics, Amsterdam, The Netherlands

⁵University Health Network UHN, Neurology, Toronto, Canada

⁶University Health Network UHN, Joint Department of Medical Imaging- Cardiothoracic Division, Toronto, Canada

⁷University Health Network UHN, Joint Department of Medical Imaging- Neuroradiology Division, Toronto, Canada

We hypothesize that ECG-gated CT-angiography (CTA) of the heart and complete aortic arch in patients with acute ischemic stroke (AIS) is both practically feasible in the acute phase (eligible for reperfusion therapy) and superior to echocardiography as a first-line screening method for detection of structural cardio-aortic sources of ischemic stroke.

We performed a single-center prospective pilot study in 50 patients with AIS who were potentially eligible for reperfusion therapy. Patients underwent CTA from aortic arch to intracranial arteries, followed by prospective sequential ECG-gated CTA of the heart. Routine work-up for cardioembolism (ECG, Holter, echocardiography) was also performed. We predefined potential sources of embolism from the heart and aortic arch, and imaging was systematically scored by cardio-radiologists (CTA) and cardiologist (echocardiography).

Median age was 73 years (IQR 64–83), 31/50 (62%) were male and median NIHSS was 8 (IQR 4–21). Median duration of cardiac scan protocol was 81 seconds (IQR 77–99). Forty patients underwent transthoracic echocardiography and median time between CTA and echocardiography was 1 day (IQR 1–10). CTA revealed 13 high-risk cardio-aortic sources of emboli in 12 patients (24%), namely intracardiac thrombus (N = 4), dilated cardiomyopathy (N = 4), signs of endocarditis (N = 1), acute aortic dissection (N = 2), and ulcerated noncalcified aortic arch atheroma (N = 2). Echocardiography revealed high-risk cardio-aortic sources in 5/40 (13%) patients (two-sided McNemar test, p < 0.008).

Preliminary data suggest incorporating ECG-gated cardiac CTA in the AIS scan-protocol is feasible and promising as a screening method for detection of structural cardio-aortic sources of AIS. A larger prospective multicenter study is currently underway.

N/A

QUANTIFICATION OF HEMORRHAGIC VOLUME IN PATIENTS WITH HEMORRHAGIC TRANSFORMATION AFTER AN ACUTE ISCHEMIC STROKE

K. Van Kranendonk¹, K.M. Treurniet¹, A.M.M. Boers², O.A. Berkhemer¹, L.A. van den Berg³, V. Chalos – Andreou⁴, H.F. Lingsma⁵, W.H. van Zwam⁶, A. van der Lugt⁷, R.J. van Oostenbrugge⁸, D.W.J. Dippel⁴, Y.B.W.E.M. Roos³, H.A. Marquering², C.B.L.M. Majoie¹; The MR CLEAN investigators

¹Amsterdam UMC- location AMC, Department of Radiology and Nuclear Medicine, Amsterdam, The Netherlands

²Amsterdam UMC- location AMC, Department of Biomedical Engineering and Physics, Amsterdam, The Netherlands

³Amsterdam UMC- location AMC, Department of Neurology, Amsterdam, The Netherlands

⁴Erasmus MC-University Medical Center Rotterdam, Department of Neurology, Rotterdam, The Netherlands

⁵Center for Medical Decision Making- Erasmus MC-University Medical Center Rotterdam, Department of Public Health, Rotterdam, The Netherlands

⁶Cardiovascular Research Institute Maastricht CARIM- Maastricht University Medical Center, Department of Radiology, Maastricht, The Netherlands

⁷Erasmus MC-University Medical Center Rotterdam, Department of Radiology & Nuclear Medicine, Rotterdam, The Netherlands

⁸Cardiovascular Research Institute Maastricht CARIM- Maastricht University Medical Center, Department of Neurology, Maastricht, The Netherlands

Hemorrhagic transformation (HT) is a frequent complication of acute ischemic stroke and impairs functional outcome. Usually, HT is classified based on its radiological appearance according to the ECASS classification, which divides HT in four subtypes; hemorrhagic infarction type 1(HI1) and 2(HI2) and parenchymal hematoma type 1(PH1) and 2(PH2). Accurate discrimination of these subtypes can be challenging. Therefore, we aimed to quantify hemorrhagic volume and assess its prognostic value by determining the association of hemorrhagic volume with functional outcome.

We included patients from the MR CLEAN. HT was classified according to the ECASS II classification and total hemorrhagic volume was quantified by its delineation on follow-up (24h-5day) CT. We assessed functional outcome using the modified Rankin Scale 90 days after stroke onset. Ordinal logistic regression with adjustment for potential confounders was used to determine the association of hemorrhagic volume and ECASS classification with functional outcome.

All patients with follow-up imaging were included (n = 478). In total, 222 had HT of which 76 had HI1, 71 had HI2, 36 had PH1 and 39 had PH2. Hemorrhagic volume was significantly associated with poor functional outcome. However, this association was no longer statistically significant in the adjusted analysis (Table 1.). Additionally, HI2 and PH2 were significantly associated with poor functional outcome in the multivariable analysis.

In our population, hemorrhagic volume quantification might add prognostic value beside the ECASS classification. However, in our pre-defined adjusted analysis, the statistically signficant effect was lost and only HI2 and PH2 were signifcantly associated with poor functional outcome.

N/A

BLOOD-ARTERY DISRUPTION AFTER MECHANICAL THROMBECTOMY FOR ACUTE ISCHEMIC STROKE: A DUAL-ENERGY CT STUDY

A. Renú¹, S. Amaro¹, A. Ló pez-Rueda², C. Laredo ¹, X. Urra¹, N. Macias², L. Llull¹, S. Rudilosso¹, V. Obach¹ and A. Chamorro¹

¹Hospital Clinic, Institut Clinic de Neurociències, Barcelona, Spain

²Hospital Clinic, Radiology, Barcelona, Spain

The presence of post-interventional hyperdensities in the subarachnoid space as a consequence of blood-artery disruption (BAD) is not uncommon after mechanical thrombectomy (MT), although its clinical relevance is controversial. We aimed to assess the predictors and clinical relevance of different patterns of BAD after MT.

A consecutive series of acute stroke patients treated with MT from a prospective single-center registry were retrospectively reviewed. BAD was defined as the presence of new hyperdensities in the subarachnoid or intraventricular compartments in a post-treatment Dual Energy-CT (DE-CT), and was further classified according to their composition (isolated contrast staining versus extravasated blood) and extension (diffuse = hyperdensities in more than one extraparenchymal compartments). Logistic regression models adjusted by relevant confounders assessed the association of BAD with pretreatment, procedural variables and clinical outcome (ordinal distribution of the Rankin Scale at 90 days).

Overall, 424 patients fulfilled inclusion criteria (median-NIHSS = 16). BAD was observed in 120 (28%) patients (isolated contrast staining n = 22, extravasated blood n = 98), and was diffuse in 72 (17%). In multivariate models, BAD was independently associated with more device passes (OR = 1.50, 95% CI = 1.17–1.92, p = 0.001), stentriever in comparison with aspiration technique (OR = 5.50, 95% CI = 1.28–23.69, p = 0.022), M2-occlusions (OR = 2.14, 95% CI = 1.19–3.84, p = 0.011) and coexistent parenchymal hematomas (OR = 2.96, 95% CI = 1.26–6.94, p = 0.013). In adjusted ordinal regression models the association of BAD with worse clinical outcome was only significant if it was diffuse (OR = 3.0, 95% CI = 1.77–5.08, p < 0.001) or blood-related (OR = 1.61, 95% CI = 1.03–2.52, p = 0.037).

These findings support the clinical relevance of diffuse blood-artery disruption after MT and reinforce the need for improvements in reperfusion strategies.

N/A

NONCONTRAST CT SIGNS AS PREDICTORS OF HAEMATOMA EXPANSION, FUNCTIONAL OUTCOME AND RESPONSE TO TRANEXAMIC ACID IN ACUTE SPONTANEOUS INTRACEREBRAL HAEMORRHAGE

Z.K. Law^1,2, A. Ali¹, K. Krishnan³, A. Bischoff¹, J. Appleton¹, P. Scutt¹, L. Woodhouse¹, S. Pszczolkowski⁴, R. Dineen⁴, P.M. Bath¹, N. Sprigg¹

¹University of Nottingham, Stroke Trials Unit, Nottingham, United Kingdom

²UKM Medical Centre, Department of Medicine, Kuala Lumpur, Malaysia

³Nottingham University Hospitals NHS Trust, Stroke Services, Nottingham, United Kingdom

⁴University of Nottingham, Division of Clinical Neuroscience, Nottingham, United Kingdom

Blend, black hole and island signs, heterogeneous density and irregular shape were reported to predict haematoma expansion. We explored the value of these noncontrast CT (NCCT) signs in predicting haematoma expansion and functional outcome in intracerebral haemorrhage.

Baseline and 24-hour CT of the Tranexamic acid for IntraCerebral Haemorrhage-2 (TICH-2) trial participants were analysed. Independent blinded assessors measured haematoma volumes and assessed the presence of the NCCT signs. Haematoma expansion was defined as an increase in haematoma volume of >33% or >6 mL on 24-hours CT. Poor functional outcome was defined as modified Rankin Scale of >3 at day 90.

Of 2325 patients recruited (mean age 68.9 years; 1184 male, 56%), 2077 (89.3%) had valid baseline and 24-hour scans. 568 (27.0%) had haematoma expansion while 1259 (54.6%) had poor functional outcome. Blend sign (adjusted odds ratio [aOR] 1.58, 95% CI 1.19–2.09; p = 0.001) and black hole sign (aOR 1.54, 1.07–2.23; p = 0.022) were independent predictors of haematoma expansion on multivariate logistic regression. Black hole sign (aOR 1.54, 1.12–2.13; p = 0.009), island sign (aOR 2.14, 1.20–3.82; p = 0.010) and irregular shape (aOR 2.38, 1.84–3.08; p < 0.001) were significant predictors of poor functional outcome. There was no significant interaction between the presence of NCCT signs and treatment effect of tranexamic acid on haematoma expansion or functional outcome (P interaction all >0.05).

Blend and black hole signs predict haematoma expansion while black hole, island signs and irregular shape predicts poor functional outcome after intracerebral haemorrhage. NCCT signs did not predict a better response to tranexamic acid.

EudraCT Number 2012–004108-37

STROKE ETIOLOGY AND THROMBUS IMAGING CHARACTERISTICS: A MR CLEAN REGISTRY SUBSTUDY

N. Boodt¹, K. Compagne², B. Dutra³, N. Samuels¹, N. Arrarte³, P. Konduri³, M. Kappelhof³, G. Lycklama⁴, H. Marquering³, M. Charles³, H. Lingsma⁵, D. Dippel⁶, A. van der Lugt²; on behalf of the MR CLEAN Registry Investigators

¹Erasmus Medical Center, Neurology-Radiology-Public Health, Rotterdam, The Netherlands

²Erasmus Medical Center, Radiology, Rotterdam, The Netherlands

³Amsterdam UMC-location Academic Medical Center, Radiology and Nuclear Medicine, Amsterdam, The Netherlands

⁴Haaglanden Medical Center, Neuroradiology, The Hague, The Netherlands

⁵Erasmus Medical Center, Public Health, Rotterdam, The Netherlands

⁶Erasmus Medical Center, Neurology, Rotterdam, The Netherlands

Thrombus imaging characteristics are associated with thrombus histology and might be useful to help identify stroke subtype in patients with cryptogenic stroke. We aimed to study the association between stroke etiology and thrombus CT characteristics in patients with acute ischemic stroke due to a large vessel occlusion.

For 1488 consecutive patients enrolled in the MR CLEAN Registry, stroke etiology was determined using adapted TOAST criteria. The association of stroke etiology with the hyperdense artery sign on non-contrast CT (NCCT) was estimated with univariable and multivariable logistic regression. Additionally, for 408 patients with available thin-section NCCT and CTA imaging, we assessed the association of stroke etiology with the following thrombus characteristics: Absolute and relative attenuation, location, distance from the ICA-terminus, length and perviousness.

A non-cardioembolic etiology was associated with the hyperdense artery sign, higher absolute thrombus attenuation, more proximal occlusions and increasing thrombus length (table 1). Thrombus characteristics for cardioembolic and cryptogenic strokes were similar.

The hyperdense artery sign, absolute thrombus attenuation, location, distance from the ICA-T and length were associated with stroke etiology. Additionally, our study supports the general hypothesis that many cryptogenic strokes have a cardioembolic cause.

(This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 777072).

N/A

FINDINGS FROM THE CÚRAM REGISTRY OF CLOT HISTOLOGY AND STROKE PATHOLOGY: ‘RESTORE REGISTRY’

S. Fitzgerald^1,2, A. Rentzos³, J. Thornton^4,5, D. Dunker³, J.E. Karlsson³, T. Tatlisumak⁶, P. Brennan⁷, A. O’Hare⁷, S. Power⁷, I. Szikora⁸, G. Tsivgoulis⁹, K. Psychogios¹⁰, A. Pandit¹, A. Douglas¹, R. Rossi¹, O.M. Mereuta¹, M. Gilvarry¹¹, D.F. Kallmes², W. Brinjikji² and K.M. Doyle¹

¹National University of Ireland Galway, CÚRAM–Centre for Research in Medical Devices, Galway, Ireland

²Mayo Clinic, Department of Radiology, Rochester MN, USA

³Sahlgrenska University Hospital, Department of Interventional and Diagnostic Neuroradiology, Gothenburg, Sweden

⁴Royal College of Surgeons in Ireland, Beaumont Hospital, Dublin, Ireland

⁵Beaumont Hospital Dublin, Department of Radiology, Dublin, Ireland

⁶Sahlgrenska University Hospital, Department of Neurology, Gothenburg, Sweden

⁷Beaumont Hospital Dublin, Department of Radiology, Dubln, Ireland

⁸National Institute of Neurosciences, Department of Neurointerventions, Budapest, Hungary

⁹National & Kapodistrian University of Athens, Second Department of Neurology, Athens, Greece

¹⁰Metropolitan Hospital, Acute Stroke Unit, Piraeus, Greece

¹¹Cerenovus, Galway Neuro Technology Centre, Galway, Ireland

The association between poor neurological outcomes and clot fragmentation or failure to remove clot suggests that understanding the science of clot may be beneficial in further advancing the field of acute ischemic stroke (AIS) intervention.

This large multi-centre international registry aims to investigate the relationships between histological compositions of occlusive clots, non-contrast computed tomography (NCCT) imaging and clinical and procedural information in AIS patients. Histopathological analysis was performed using the Martius Scarlett Blue stain (MSB) which identifies platelet-rich regions, in addition to red blood cells (RBCs), white blood cells (WBCs) and fibrin, allowing for more accurate clot characterization than previous studies. Histological composition was quantified using Orbit Image Analysis (www.orbit.bio) and per-pass analysis was performed.

Ninety-two patients were included in the study and a total of 142 clot samples (passes) were collected. RBC-rich clots are Hyperdense; there was a significant positive correlation between the proportion of RBCs and mean Hounsfield Unit (HU) density on NCCT (p < 0.001*, n = 57). Platelet-rich clots are Isodense on NCCT; there was a significant negative correlation between platelets and mean HU density on NCCT (p = −0.015*, n = 57). Cardiac clots are older, more organised clots than large-artery arteriosclerotic (LAA) clots. Cardiac clots had significantly more fibrin (p = 0.024*, n = 142) and WBCs (p = 0.045*, n = 142) and significantly less RBCs (p = 0.013*, n = 142) than LAA clots. No significant difference in histological composition was noted between procedural passes.

This suggests that suspected aetiology and diagnostic imaging may help to identify the composition of the occlusive thrombus.

Acknowledgements: Science Foundation Ireland (13/RC/2073) and Cerenovus.

N/A

PREVALENCE AND EVOLUTION OF SUSCEPTIBILITY WEIGHTED IMAGING LESIONS IN PATIENTS WITH ARTIFICIAL HEART VALVES (PRESERVE)

T. Meinel¹, B. Philipe², J. Dürrenmatt³, F. Meinel⁴, Z. Felix² and A. Marcel³

¹Inselspital Bern, Neurology, Bern, Switzerland

²Inselspital, Neuroradiology, Bern, Switzerland

³Inselspital, Neurology, Bern, Switzerland

⁴University Hospital Rostock, Radiology, Rostock, Germany

In patients with mechanical heart valves (HV), SWI lesions suspected to be caused by metallic degenerative metallic abrasion on cerebral MRI has been described years ago. However, the scientific basis for this hypothesis is very poor.

In this singlecenter, retrospective registry analysis we provide a temporal and topical description of SWI lesions in 165 patients with HV. SWI lesions were rated according to established criteria by two experienced readers blinded to the valve status.

In patients with mechanical HV 81% had three or more SWI lesions, in patients with biological HV 76% had three or more SWI lesions. There was a median increase in SWI lesions of 4 (IQR 1–7) from the preoperative to the first postoperative MRI scan. 22.5% of patients fulfilled the criteria of possible or probable CAA according to the modified Boston criteria, although only 16% of those patients had superficial siderosis. On multivariable logistic regression adjusting for pre-specified confounders, duration of extracorporal circulation (β 0.738, 95%-CI 0.366–1.111, P < 0.001) significantly increased the odds of having SWI lesions on postoperative MRI.

SWI lesions are almost universally found in patients with neurologic symptoms irrespective of valve type with the vast majority of patients having multiple lesions. SWI lesions evolve around the valve implantation and remain stable thereafter in most patients. SWI lesions associated with heart valves can mimic CAA, but it is questionable whether they share the same pathophysiological features and long-term bleeding risk as CMB seen with CAA or classical vascular risk factors.

N/A

REFINING PRECLINICAL RESEARCH: THE ANIMAL STUDY REGISTRY

K. Diederich¹, B. Bert², C. Heinl², J. Chmielewska², B. Grune², A. Bruns³, A. Riesel³, F. Buchert³, B. Orywahl-Wild³, L. Lewejohann¹, M. Greiner³ and G. Schönfelder²

¹German Federal Institute for Risk Assessment, Laboratory Animal Science, Berlin, Germany

²German Federal Institute for Risk Assessment, Department Experimental Toxicology and ZEBET, Berlin, Germany

³German Federal Institute for Risk Assessment, Department Exposure, Berlin, Germany

The demand for greater transparency and reproducibility of animal studies is rising as the repercussions of publication bias and selective reporting and its potential countermeasures are increasingly discussed. The Animal Study Registry (www.animalstudyregistry.org) is an online registry for animal research and was launched on January 7th 2019 as a reaction to the reproducibility crisis. It is operated by the German Centre for the Protection of Laboratory Animals (Bf3R) at the German Federal Institute for Risk Assessment (BfR). The registry provides a platform to register an exact study plan prior to the start of the experiment including a detailed description of the methods and statistical planning.

The Animal Study Registry is designed for both confirmatory and exploratory studies. Once submitted, the study can still be changed or retracted within two weeks. Subsequently, the registration becomes binding and the study receives a Digital Object Identifier number which marks the study as intellectual property. The visibility of the registered study can be restricted for up to five years. At the end of the embargo period the study will automatically become publicly accessible.

The Animal Study Registry fosters transparency and reproducibility of animal studies and improves the quality of biomedical research by preventing selective reporting and dissemination bias, by encouraging the publication of negative or neutral results, and by supporting thorough planning including statistical analysis.

Consequently, registering a study in Animal Study Registry proves commitment to transparency and data quality to reviewers and editors, to third party donors and to the general public.

N/A

RETROSPECTIVE LARGE-SCALE ANALYSIS OF ASSOCIATIONS BETWEEN SELECTED OUTCOME MEASURES FOLLOWING MIDDLE CEREBRAL ARTERY OCCLUSION IN MICE

C. Sobey¹, S. Zhang¹, H.A. Kim¹, X. Li², G. Drummond¹ and T. Phan³

¹La Trobe University, School of Life Sciences, Bundoora, Australia

²La Trobe University, Dept of Mathematics and Statistics, Bundoora, Australia

³Monash University, Department of Medicine- School of Clinical Sciences at Monash Health, Clayton, Australia

Pre-clinical stroke studies model the pathophysiology of clinical stroke where a range of parameters are measured to assess the severity of outcome. However, post-stroke pathology is complex and variable, and associations between parameters are difficult to identify.

We performed a retrospective large-scale analysis on 716 control C57Bl6 mice subjected to either transient (1 h) middle cerebral artery occlusion (tMCAO) or permanent MCAO (pMCAO). Structural equation modelling (SEM) models were constructed to identify cause-effect relationships among numerous variables recorded at 24, 48, or 72 h.

Our collective data demonstrate that following tMCAO infarct volume is fully developed within 24 h, whereas brain edema continues to evolve. pMCAO resulted in larger infarct (70%) and edema volume (>2-fold) than tMCAO. 85% of lung infections resolved within 48 h of tMCAO, accompanied by improved functional outcome. There was greater leukocyte infiltration in brains of mice receiving pMCAO, and more severe leukopenia. Multivariable analyses revealed that edema is positively correlated with infarct volume (β = 0.778) and clinical score (β = 0.365) in tMCAO, but not pMCAO. Age is correlated positively with lung infection (β = 0.540), and negatively with mobility (β = −0.322) following tMCAO. However, after pMCAO, age does not appear to affect infarct volume, functional outcome or lung infection. Univariable analysis (Spearman Rank correlation) showed clinical score to be negatively correlated with circulating leukocytes (ρ = −0.286), but positively correlated with brain leukocytes (ρ = 0.332) following tMCAO, but not pMCAO.

Large-scale analysis of animal experiments provides insight into relationships between variables not available when studies are analysed in isolation.

N/A

LOOKING FOR A MINIMALLY INVASIVE FOCAL STROKE MODEL IN THE PIG: A JOURNEY FULL OF PITFALLS

L. Rodriguez Esparragoza¹, J. Fernandez-Ferro², R.A. Benn³, M. Gomez Bravo⁴, C. Galan Arriola³, J. Sanchez⁵ and B. Ibañez³

¹Instituto Germans Trias i Pujol, Neurovascular research, Barcelona, Spain

²Hospital Universitario Rey Juan Carlos de Mostoles, Neurology, Madrid, Spain

³Centro Nacional de Investigaciones Cardiovasculares CNIC Carlos III, Laboratorio traslacional para la imagen y terapia cardiovascular, Madrid, Spain

⁴Beth Israel Deaconess Medical Center/Harvard Medical School, Plastic Surgery, Boston, USA

⁵Philips beria, Healthcare, Madrid, Spain

Many pitfalls are present in the development of a pig stroke model, and deserve careful considerations. In this work we describe our experience.

Four different trials were developed to achieve a brain infarct in swine: a classical endovascular approach; an endovascular injection of microspheres just before rete mirable; a direct mechanical occlusion of a superficial M3 middle cerebral artery branch through craniectomy; and adirect parenchymal injection of endothelin at two different depths (2 mm and 4–5 mm).

Brain infarction was impossible to achieve following the classical rodent endovascular model, due to the impossibility to get through the rete mirabile. The microsphere trial resulted in uncontrollable, unreproducible, multisite (frequently bilateral and brainstem) infarctions that give way to severe brain lesions manifesting with sudden coma. Direct occlusion of a superficial MCA branch required highly-demanding surgical skills and has three major problems: only permanent occlusion was feasible (through electrocoagulation), frequent trauma to the brain was inflicted during surgery, and only very small infarcts were obtained giving a highly effective collateral network in the breed. Direct endothelin injection appeared to be a reproductible, easy and minimally invasive method for obtaining both permanent and reperfusion brain injuries; with different volumes depending on the depth of injection.

Achieving a minimally invasive and reproducible stroke model in swine could be of utmost interest for the stroke field. Difficulties on the classical endovascular approach forced us to look for new approaches among whom the direct injection of the strong vasopressor endothelin in the cerebral parenchyma seems the most promising.

N/A

A MOUSE MODEL OF FOCAL CEREBELLAR INFARCT AND THE EFFECT OF ROCK INHIBITOR FASUDIL

S. Aykan¹, M. Anzabi¹, A. Lopes¹, D. Chung¹ and C. Ayata¹

¹Massachusetts General Hospital Harvard University, Radiology, Charlestown, USA

Cerebellar infarcts account for ∼3% of all strokes and are often found incidentally on neuroimaging. Experimental models representing this unique infarct location are scarce. We aimed to develop a small cerebellar stroke model, characterize the lesion and neurological deficits, and test the effects of Rho-kinase inhibitor fasudil in mice.

The potent vasoconstrictor L-NIO (100 mM) with 2.4 uL (n = 17) or vehicle (n = 21) was stereotaxically injected into the cerebellum in C57bl/6J mice. Fasudil (10 mg/kg, IP) or vehicle (saline 0.1 ml, IP) treatment started 24 hours after injection and continued twice a day for 2 weeks. We assessed behavior with sensory, motor and cognitive tests at multiple time points over 4 weeks.

L-NIO reproducibly induced small and highly focal cerebellar infarcts (CI) (0.089 ± 0.013 mm3) mainly involving white matter of cerebellar lobules IV-V, when examined on day 28. Compared with sham, CI (n = 17) showed increase in right (p = 0.035) and left stride length (p = 0.004), decrease in the distance between the center of the plantar of the front paw and the corresponding hind foot (CR) (p = 0.024). In static rod test, 70% of CImice fell from the 6mm rod, compared with 20% in the sham group (p = 0.001). CI also increased forepaw foot faults compared to sham (p = 0.020). Fasudil (n = 10) reduced infarct volumes by a third and improved the performance in the foot fault test compared with vehicle (n = 9).

This is the first description of a focal cerebellar infarct model in mice that induces reproducible lesions and neurological deficits. Fasudil reduced lesion volumes and some deficits.

N/A

RAPID ONSET OF NEURONAL CYTOTOXIC EDEMA DURING STROKE-INDUCED SPREADING DEPOLARIZATION REVEALED BY INTRAVITAL 2-PHOTON IMAGING AND QUANTITATIVE SERIAL SECTION ELECTRON MICROSCOPY

S. Kirov¹, J. Sword¹ and I. Fomitcheva²

¹Medical College of Georgia at Augusta University, Neuroscience and Regenerative Medicine, Augusta, USA

²Medical College of Georgia at Augusta University, Neurosurgery, Augusta, USA

Spreading depolarizations (SDs) are waves of sustained neuronal and glial depolarization that propagate a massive disruption of ion gradients through the brain. SD is associated with migraine aura, brain trauma, and stroke. SD causes rapid neuronal swelling and dendritic beading with spine loss as assessed in real time with 2-photon laser scanning microscopy (2PLSM). Very little is known about the immediate onset of SD-induced cytotoxic edema at the ultrastructural level.

In vivo 2PLSM followed by quantitative serial section electron microscopy (ssEM) were used to assess dendritic integrity in the sensorimotor cortex of urethane-anesthetized mice expressing EGFP during SD evoked by transient bilateral common carotid artery occlusion.

SD-disrupted dendrites contained a watery cytoplasm with disordered microtubules reflecting water influx. SD triggered the collapse of dendritic spines to dendritic shafts as evidenced by an increase in the frequency of shaft synapses. The overall density of synapses was unchanged, and the postsynaptic dendritic membranes remained attached to the presynaptic axonal boutons. Mitochondrial swelling and fragmentation were detected during SD, and many mitochondria appeared in the form of elongated interconnected organelles. The cytoplasm of recuperated dendrites after reperfusion contained arrays of microtubules, tubular mitochondria, and recovered cisterns of SER and all spines had synapses.

SD is the principal mechanism of the neuronal cytotoxic edema implicated in the rapid disruption of dendrites, but synaptic connections remain preserved. Even in tissue with severe energy deficits as during global ischemia, the SD-inflicted dendritic injury is reversible if blood flow is rapidly restored after SD onset.

N/A

M. Bacigaluppi¹, H. Descamps¹, D. De Feo¹, G. Giorgia Serena¹, I. Lelios², A. Bergamaschi¹, M. Greter², B. Becher² and G. Martino¹

¹San Raffaele Scientific Institute, Neurology and Neuroimmunology Unit, Milano, Italy

²University of Zurich, Institute of Experimental Immunology, Zurich, Switzerland

Microglial cells represent parenchymal phagocytes of the central nervous system (CNS). Although microglial cells are known to participate to brain homeostasis and to support neuronal function, their role in pathological conditions such as cerebral ischemia remains elusive.

We here studied the dynamics and role of microglia proliferation in a clinical relevant model of acute ischemic stroke, namely transient middle cerebral artery occlusion, induced without altering the integrity of the skull. Over time we followed the expansion of microglia both by histology as well as by cytofluorimetry. By use of transgenic models devoid of CSF1-R ligands, namely CSF-1 or IL-34, we studied the role of CSF-1R signaling in acute ischemic stroke.

We observed that microglia cell number increases steadily after stroke, reaching a peak at day 7, together with an increased up-regulation of CSF-1R-dependent signaling. In acute stroke microglia proliferation relies on the CSF-1R ligand CSF-1, rather than on IL-34. Interfering with CSF-1R signalling, either with the specific CSF1R inhibitors or by neutralizing CSF-1, reduces microglia and contributes to an amelioration of stroke functional outcome.

The targeted and timely inhibition of microglia proliferation, ideally guided by the future development of neuroimaging biomarkers, might be beneficial in cerebral ischemia.

N/A

ROLE OF TLR4 IN N1/N2 NEUTROPHIL PROGRAMMING AFTER STROKE

A. García-Culebras¹, V. Durán Laforet¹, C. Peña-Martínez¹, A. Moraga¹, I. Ballesteros¹, M.I. Cuartero¹, J. De la Parra¹, S. Palma-Tortosa¹, A. Hidalgo², Á.L. Corbi³, M.A. Moro¹ and I. Lizasoain¹

¹Universidad Complutense Madrid, Unidad de Investigación Neurovascular- Departamento de Farmacología y Toxicología- Facultad de Medicina and Instituto Universitario de Investigación en Neuroquímica, Madrid, Spain

²Fundación CNIC, Area of Cell and Developmental Biology, Madrid, Spain

³Centro de Investigaciones Biológicas CSIC, Departamento de Biología Celular, Madrid, Spain

After stroke, the population of infiltrated neutrophils in the brain is heterogeneous, including a population of alternative neutrophils (N2) that express M2 phenotype markers. We aimed to explore the role of TLR4 on neutrophil infiltration and polarization in this setting.

Focal cerebral ischemia was induced by occlusion of the middle cerebral artery (MCAO) in TLR4KO and WT mice. Cerebral infarct size was measured by Nissl staining and by MRI after MCAO. Leukocyte infiltration was quantified 48h after MCAO by double immunofluorescence staining and flow cytometry. To elucidate the molecular mechanisms underlying TLR4-mediated N2 phenotype, a cDNA microarray analysis was performed in neutrophils isolated from blood 48h after stroke in WT and TLR4KO mice.

TLR4-deficient mice showed an increase in neutrophil infiltration 48h after stroke compared to WT mice, concomitantly to neuroprotection. Cytometric and stereological analyses revealed an increased infiltration of N2 neutrophils (Ym1+ cells) into the ischemic core 48h after MCAO in TLR4-deficient mice. These effects were further confirmed by depleting peripheral neutrophils or using mice with TLR4 genetically ablated in the myeloid lineage. cDNA microarray analysis in neutrophils showed that TLR4 modulates the induction of several pathways previously associated with ischemia-induced inflammation, migration of neutrophils into the parenchyma and their functional priming.

TLR4 deficiency increases the levels of alternative neutrophils (N2), an effect associated with neuroprotection after stroke, supporting that modulation of neutrophil polarization is a major target of TLR4 and highlighting the crucial role of TLR4 at the peripheral level after stroke.

N/A

OSTEOPROTEGERIN PREVENTS THE GROWTH OF INTRACRANIAL ANEURYSMS PROMOTING COLLAGEN BIOSYNTHESIS AND VASCULAR SMOOTH MUSCLE CELL PROLIFERATION VIA TGFΒ1

T. Miyata¹, M. Minami², H. Kataoka¹, K. Hayashi¹, T. Yang¹, Y. Yamamoto¹, K. Shimizu¹, M. Yokode² and S. Miyamoto¹

¹Kyoto University Graduate School of Medicine, Department of Neurosurgery, Kyoto, Japan

²Kyoto University Graduate School of Medicine, Department of Clinical Innovative Medicine Institute, Kyoto, Japan

The pathogenesis of intracranial aneurysms (IAs) are characterized by reduced extracellular matrix and decreased number of vascular smooth muscle cells (VSMCs) in cerebral vascular walls. It is previously reported that osteoprotegerin (OPG) promotes collagen biosynthesis and VSMCs proliferation via TGFβ1. Here we investigated whether to prevent IAs development through the effect of OPG on collagen expression and VSMCs proliferation in experimentally induced IAs in rats.

IAs were surgically induced in 7-week-old male Sprague Dawley rats. 1 week after the operation, mouse recombinant OPG at 0.125 g/L or vehicle was continuously infused into the lateral ventricle via an osmotic pump. 5 weeks after the 1st operation, aneurysmal size, media thickness of IAs were measured in the both groups. The expression of collagen (Col1a2 and Col3a1), TGFβ1 and pSmad 2/3 in aneurysmal walls were examined by immunohistochemistry and RT-PCR. Primary cell culture of mouse VSMCs were analyzed by MTS assay with or without SB431542.

In the OPG treatment group, size and wall thickness of IAs were significantly smaller (p < 0.05) and thicker (p < 0.05) than in the control group. In the immunohistochemistry, collagen and pSmad 2/3 were upregulated in the aneurysmal wall in the OPG group. In the RT-PCR study, OPG treatment significantly upregulated expression of collagen and TGFβ1 genes (p < 0.05). MTS assay revealed the OPG treatment significantly promoted VSMCs proliferation and SB431542 canceled this effect (p < 0.05).

Our results indicate that OPG has a suppressive effect on IAs development through the activation of collagen biosynthesis and VSMC proliferation via TGFβ1 in aneurysmal walls.

N/A

Scientific Communications 07-Nursing and Allied Health Professionals

LONG-TERM TRENDS IN CARE NEEDS OF STROKE SURVIVORS IN CARE HOMES AND PREDICTORS FOR DISCHARGE TO CARE HOMES

A. Clery¹, A. Bisquera¹, A. Bhalla², C. Sackley¹, C. Wolfe¹ and Y. Wang¹

¹King’s College London, School of Population Health and Environmental Sciences, London, United Kingdom

²Guy’s and St Thomas’ NHS Foundation Trust, Stroke Unit, London, United Kingdom

The care needs of stroke survivors discharged to care homes are increasing over time. We aim to assess the needs of and identify predictors for discharge to care homes.

Using data from the South London Stroke Register between 1995 and 2014, first-ever stroke survivors were categorised into four cohorts: 1995–1999, 2000–2004, 2005–2009, 2010–2014. The baseline characteristics and 3-month follow-up care of each cohort were summarised and tested for trends over time. Multivariable logistic regression models determined associations between these factors and discharge to care homes.

Of 4,879 patients, 452 (10.1%) were discharged to care homes. This proportion decreased from 17.1% in 1995–1999 to 4.2% in 2010–2014. Over time, hospital stays shortened (median: 89–50 days, p < 0.001) and stroke unit admissions increased (31.0-92.5%, p < 0.001). The total number of 3-months post-stroke prescriptions also increased over time, including for clopidogrel (0.0-25.0%, p < 0.001) and statins (1.7-45.0%, p < 0.001), as well as physiotherapy (14.8-38.9%, p < 0.001). However, over time the proportion of patients with a GCS score < 13 has increased (29.1-48.6%, p = 0.005). Compared to patients discharged to their own homes, predictors of discharge to care homes include (odds ratio, 95% confidence interval): hospital stay of >60 days (15.9, 10.2-25.7), old age (5.8, 3.5-9.7), White ethnicity (1.7, 1.2-2.5), GCS < 13 (2.0, 1.4-2.8), incontinence (1.8, 1.2-2.5), and a post-stroke BI < 15 (3.7, 2.1-6.8).

Over time, fewer but more severe stroke survivors are discharged to care homes with improved access to stroke interventions. Further research should investigate whether care homes are well-equipped to meet the increasing needs of incoming patients.

N/A

UNDERSTANDING STROKE CAREGIVER READINESS: CONTENT VALIDATION OF THE PREPAREDNESS ASSESSMENT FOR THE TRANSITION HOME AFTER STROKE (PATH-S)

B. Lutz¹, T. Harvath², J. Joseph³ and M. Camicia⁴

¹University of North Carolina-Wilmington, School of Nursing, Wilmington- NC, USA

²University California Davis, Family Caregiving Institute, Davis- CA, USA

³University of California Davis, Betty Irene Moore School of Nursing, Davis- CA, USA

⁴Kaiser Foundation Rehabilitation Center, Rehabilitation, Vallejo- CA, USA

Stroke family caregivers are often unprepared to provide post-discharge care after transitioning home from inpatient rehabilitation. Based on previous qualitative research, the Preparedness Assessment for the Transition Home after Stroke (PATH-s), a pre-discharge caregiver readiness assessment, was developed. We conducted cognitive interviews to understand how stroke family caregivers interpreted PATH-s items and their thoughts about completing the instrument.

Following completion of the PATH-s, a trained researcher conducted item-by-item semi-structured interviews in 2 phases with 20 family caregivers of stroke survivors admitted to a U.S. inpatient rehabilitation facility. Items were revised in phase 2 (n = 9) based on analyses of phase 1 interviews (n = 11). To refine the instrument, participants were asked about item clarity and their thoughts about completing the PATH-s. Using thematic analyses, interviews were coded independently by two experienced research nurses to establish inter-rater reliability.

Upon PATH-s completion, participant interviews revealed 3 main themes: uncertainty, anticipation, and cues to action. Uncertainty about prognosis and recovery, long-term implications of stroke, resource needs, and physical ability to sustain the caregiver role were reported. Participants also described anticipating stroke survivor care management strategies, future decisions they would need to make, and managing self-care post-discharge. Caregivers reported that the PATH-s increased self-awareness about the future and provided cues to action to begin to address their needs during inpatient rehabilitation.

Assessment tools, e.g. the PATH-s, may activate family caregivers to engage in care discussions and offer opportunities for rehabilitation providers to individualize care plans to better prepare family members for the caregiving role.

N/A

POST-STROKE COMPLICATIONS AND CAREGIVERS’ STRESS AMONG STROKE PATIENTS IN A DEVELOPING COUNTRY

A. Kumar¹, D. Khurana², S. Kaur¹, S. Ghai¹, M. Modi², S. Pattanaik³, P. Sharma², M. Kumar², A. Dhaliwal², D. Kumar², M. Nagi¹ and C. Narayanan⁴

¹Post Graduate Institute of Medical Education and Research, National Institute of Nursing Education, Chandigarh, India

²post graduate institute of medical education and research, Neurology, Chandigarh, India

³Post graduate institute of medical education and research, Pharmacology, Chandigarh, India

⁴IIT Ropar, Computer science and technology, Punjab, India

In India, most stroke patients get discharged to home care. Caregivers face a challenging task to give the proper careresulting instress and physical burden. Inadequate home care may lead to complications and increased morbidity and mortality

To assess health problems of stroke-survivors and their caregivers.

170 bedridden stroke-survivors within 1 month of stroke, having significant disability and their caregiver were recruited. Follow-up was done at3 and 6months. Observation technique, interview schedule and Modified Caregiver Strain Index were used to collect data. Data was analyzed applying descriptive and inferential statistics.

Mean age of patients and caregivers was 58.64 (SD:13.92) years and 35.94 (SD:9.56) years. 98 patients and 105 caregivers were males. Allpatients were discharged on Ryle’s tube and Foley’s catheter in-situ, and 23 patients were discharged on tracheostomy-tube in-situ. At 3 and 6 months follow up, bedsores, aspiration pneumonia, UTI, frozen shoulder and deep venous thrombosis (DVT) was seen in 20.75%and 27.66%, 6.72% and 9.0%, 8.0% and 11%, 8.5% and 24.4%, and 2.0% and 2.66% respectively. Mortality at 6months was strongly association with post-stroke-complications (p = 0.001; Chi-square test). 23.6% and 30.5% caregivers had severe care-giver stress at 3 and 6 months, respectively.

Post stroke complications are frequent in patients on home care and cause a significant burden on the caregivers. We proposed to develop a Stroke Home Care mobile application for post-stroke home care to tackle post-stroke complications and improve quality of life of stroke survivors.

N/A

DYSPHAGIA SCREENING AND RISKS OF PNEUMONIA AND ADVERSE OUTCOMES IN ACUTE STROKE: AN INTERNATIONAL MULTICENTRE STUDY

M. Ouyang¹, E. Boaden², H. Arima³, P. Lavados⁴, L. Billot⁵, M.L. Hackett⁶, V.V. Olavarría⁷, P. Muñoz-Venturelli⁸, L. Song¹, K. Rogers⁵, S. Middleton⁹, O.M. Pontes-Neto¹⁰, T.H. Lee¹¹, C. Watkins², T. Robinson¹² and C.S. Anderson¹

¹The George Institute for Global Health, Stroke Division, Beijing, China

²University of Central Lancashire, Faculty of Health and Wellbeing, Preston, United Kingdom

³Fukuoka University, Department of Preventive Medicine and Public Health, Fukuoka, Japan

⁴Universidad de Chile, Departamento de Ciencias Neurológicas-, Santiago, Chile

⁵The George Institute for Global Health, Statistic Division, Sydney, Australia

⁶The George Institute for Global Health, Stroke and Neurology Division, Sydney, Australia

⁷Universidad del Desarrollo, Departmento de Neurología y Psiquiatría- Clínica Alemana de Santiago- Facultad de Medicina, Santiago, Chile

⁸Clínica Alemana Universidad del Desarrollo, Centro de Estudios Clínicos- Instituto de Ciencias e Innovación en Medicina- Facultad de Medicina, Santiago, Chile

⁹Australian Catholic University, Nursing Research Institute- St Vincent’s Health Sydney Australia, Sydney, Australia

¹⁰University of Sao Paulo, Department of Neuroscience and Behavioral Sciences- Ribeirão Preto Medical School, Sao Paulo, Brazil

¹¹Chang Gung University, Stroke Center and Department of Neurology, Taoyuan, Taiwan R.O.C

¹²University of Leicester, Department of Cardiovascular Sciences and NIHR Leicester Biomedical Research Centre, Leicester, United Kingdom

Guidelines recommend dysphagia screening/assessment before any oral intake in stroke patients, but data are limited on how effective such recommendations operate in practice. We aimed to determine associations of a ‘brief’ screen (a simple drinking test) and ‘detailed’ assessment (a more systematic examination) of dysphagia on adverse clinical outcomes in acute stroke.

Post-hoc analysis was undertaken of the multi-centre, cluster crossover, randomised controlled trial of lying-flat vs. sitting-up head positioning in 11,093 acute stroke patients from nine countries. The outcomes of pneumonia and death or disability (mRS 3–6) at 90 days were assessed in multivariate logistic regression models.

Overall, 8784 (79.2%) and 3917 (35.3%) patients had a dysphagia screen and assessment, respectively, but their frequency and timing varied widely across regions. Compared to ‘screen-pass’ patients, those who failed had higher risks of pneumonia (adjusted odds ratio [aOR] 2.99, 95% CI 2.19-4.09) and death or disability (aOR 1.69, 95% CI 1.44-1.99) after adjusted for baseline and hospital characteristics. Similar results were obtained for those who received a dysphagia assessment on pneumonia (aOR 3.04, 95% CI 2.10-4.38) and poor outcome (aOR 2.26, 95% CI 1.80-2.85). Subsequent use of feeding restrictions was also related to higher risks of pneumonia and poor outcome (aOR 5.66, 95% CI 3.18-10.11; and aOR 2.01, 95% CI 1.60-2.53, respectively).

Failing a dysphagia screen/assessment is associated with increased risks of pneumonia and poor clinical outcome after acute stroke. Further studies exploring the effectiveness of alternative approaches to management of feeding are required to understand how patient outcomes can be improved.

N/A

URINARY CATHETERISATION AND CLINICAL OUTCOMES IN ACUTE STROKE PATIENTS: AN INTERNATIONAL MULTICENTRE STUDY

M. Ouyang¹, L. Billot², L. Song¹, M.L. Hackett³, P.M. Lavados⁴, V.V. Olavarría⁴, P. Muñoz-Venturelli⁵, S. Middleton⁶, O.M. Pontes-Neto⁷, T.H. Lee⁸, C. Watkins⁶, T. Robinson⁹ and C.S. Anderson¹

¹The George Institute for Global Health at Peking University Health Science Center, Stroke Division, Beijing, China

²The George Institute for Global Health, Statistic Division, Sydney, Australia

³The George Institute for Global Health, Neurology & Mental Health Division, Sydney, Australia

⁴Clinica Alemana Universidad del Desarrollo, Department of Neurology and Psychiatry, Santiago, Chile

⁵Instituto de Ciencias e Innovación en Medicina, Centro de Estudios Clínicos- Facultad de Medicina-Clínica Alemana Universidad del Desarrollo, Santiago, Chile

⁶Australian Catholic University, Nursing Research Institute- St Vincent’s Health Sydney Australia, Sydney, Australia

⁷University of Sao Paulo, Department of Neuroscience and Behavioral Sciences- Ribeirão Preto Medical School, Sao Paulo, Brazil

⁸Chang Gung University, Stroke Center and Department of Neurology- Linkou Chang Gung Memorial Hospital and College of Medicine, Taoyuan, Taiwan R.O.C

⁹University of Leicester, Department of Cardiovascular Sciences and NIHR Leicester Biomedical Research Centre, Leicester, United Kingdom

Indwelling urine catheters (IUC) are required to manage bladder dysfunction in patients disabled after acute stroke, such as urinary incontinence or urinary retention. There is limited data about the use of catheter and the dependency outcome in patients at a long-term of follow-up, especially in large sample size of acute stroke patients across different regions. We aimed to determine the predictors of IUC and their association with clinical outcomes in acute stroke.

A retrospective analysis of the data on 11,093 acute stroke patients randomised to lying-flat vs sitting-up head positioning in the international multi-centre, cluster cross-over, randomised controlled trial (Head Positioning in Acute Stroke Trial [HeadPoST]). Multilevel binomial logistic regression analysis was used with study outcomes of death or disability (modified Rankin Scale 3–6) and urinary tract infection (UTI) within 90-days.

Overall, 1167 (12%) patients had an IUC (median duration 5 days), but the frequency and duration of use varied widely across regions. IUC use was most frequent in older patients with high comorbidity, more severe neurological impairment and an intracerebral haemorrhage (ICH). IUC was associated with death or dependency (adjusted odds ratio [aOR] 1.56, 95% CI 1.23-1.98), but not UTI after adjustment for antibiotic treatment and stroke severity at discharge (aOR 1.07, 95% CI 0.56-2.04). There was no heterogeneity of associations across major patient characteristics and in sensitivity analysis.

IUC use predicts poor outcome after acute stroke. Further studies are needed to better understand mechanisms relationship to outcome and to improve appropriate IUC use.

N/A

FEWER BUT LONGER TREATMENT SESSIONS FOR APHASIA ARE ASSOCIATED WITH BETTER RECOVERY IN THE VERY EARLY REHABILITATION IN SPEECH (VERSE) CLINICAL TRIAL COHORT

T. Rai^1,2, E. Godecke^3,4,5, E. Armstrong³, D. Cadilhac^4,5,6, N. Ciccone³, M. Rose^5,7, A. Holland⁸, A. Whitworth⁹, F. Ellery¹⁰, G. Hankey¹¹, S. Middleton¹², J. Bernhardt^4,5,10; The VERSE Collaborators

¹University of Technology Sydney, School of Mathematical and Physical Sciences, Sydney, Australia

²University of Technology Sydney, Graduate Research School, Sydney, Australia

³Edith Cowan University, School of Medical and Health Sciences, Perth, Australia

⁴Centre of Research Excellence in Stroke Rehabilitation and Brain Recovery, Florey Institute of Neuroscience and Mental Health, Melbourne, Australia

⁵Centre of Research Excellence in Aphasia Rehabilitation and Recovery, La Trobe University, Melbourne, Australia

⁶Monash University, School of Clinical Sciences at Monash Health, Melbourne, Australia

⁷La Trobe University, School of Allied Health, Melbourne, Australia

⁸University of Arizona, Speech- Language and Hearing Sciences, Arizona, USA

⁹Curtin University, School of Occupational Therapy- Social Work and Speech Pathology, Perth, Australia

¹⁰University of Melbourne, Florey Institute of Neuroscience and Mental Health, Melbourne, Australia

¹¹University of Western Australia, School of Medicine and Pharmacology, Perth, Australia

¹²Nursing Research Institute, St Vincent’s Health Australia Sydney and Australian Catholic University, Sydney, Australia

The Very Early Rehabilitation of SpEech after stroke (VERSE) trial recently showed no benefit of more therapy (26 hours) compared to usual care (13 hours) delivered in the first 12 weeks of stroke recovery. This paper examines the question: What therapy frequency and session length is associated with recovery in individual aphasia therapy regimens in the first 12 weeks after stroke?

Participants with aphasia, recruited from 17 Australia/New Zealand acute-stroke units were randomised to one of three treatment arms. This secondary analysis used linear mixed models to investigate the effects of therapy frequency and session length on the Western Aphasia Battery-Aphasia Quotient (AQ) at 12 weeks, after controlling for therapy amount.

Participants (n = 214) received aphasia therapy (mean = 1377 minutes; sd = 735) in 36 sessions (sd = 19) with mean session length: 39 minutes (sd = 11) in the first 12 weeks. The mean AQ was 41.2 (sd = 28.0) at baseline and 67.7 (sd = 29.5) at 12 weeks. After controlling for baseline stroke and aphasia severity, site and therapy amount, the session length had a significant effect (b = 0.33, p = 0.024) of increasing AQ at 12 weeks by 0.33 points per minute increase in session length. On average, 1350 minutes of therapy provided in 30 sessions of 45 minutes each, can be expected to achieve 5 points more on AQ at 12 weeks, than if 45 sessions of 30 minutes each was provided.

Sessions of longer duration were associated with better recovery than a larger number of shorter sessions which achieved the same total amount of therapy.

ANZ Trials Registry: ACTRN12613000776707

EXPERIENCES OF PATIENTS WITH STROKE AND FAMILY MEMBERS OF POST-ACUTE REHABILITATION DECISION-MAKING AND COMMUNITY-BASED REHABILITATION

T. Green¹, K. Marnane², L. Gustafsson³, I. Rosbergen⁴, R. Grimley⁵, E. Horton⁶ and G. Cadigan⁷

¹Queensland University of Technology, Nursing, Brisbane, Australia

²University of Queensland, Rehabilitation, Brisbane, Australia

³Griffith University, Rehabilitation, Gold Coast, Australia

⁴Sunshine Coast University Hospital, Rehabilitation, Nambour, Australia

⁵SSCUH, Medicine, Nambour, Australia

⁶University of the SUnshine Coast, Nursing, Sippy Downs, Australia

⁷Queensland Health, Statewide Stroke Network, Brisbane, Australia

Decisions for ongoing inpatient or outpatient rehabilitation are typically made by the acute medical and rehabilitation team; the perspectives and preferences of patients and family members may not be specifically elicited. In this study we explored the experiences of patients with stroke and their families in post-acute rehabilitation service decision-making.

N/A

A qualitative descriptive approach was used to document the perceptions and lived experiences of patients and families receiving post-acute community rehabilitation care and services, and their involvement in rehabilitation decision-making.

Data Collection: Semi-structured interviews were conducted with 15 patients and family members from 7 sites across the state, between November 2017 and March 2018. Demographic data including age, sex, relationship to patient, employment status, other caregiving responsibilities, and post-stroke functions outcomes were collected.

Data Analysis: Thematic analysis was used to analyze the qualitative data.

Three major themes:

1. Communication delays hindered access to community rehabilitation services, GP support.

2. Transitional support gaps: patients receiving in-home rehabilitation services were impressed with the service but did find the service tailed off quickly with little follow up over time; lack of support for patients with stroke who were also caregivers and for partners/family members providing the support to the person with stroke; limited pharmacy support

3. Service delivery gaps: overwhelming amount of ’ paper’; lack of choice in where community outpatient rehabilitation was provided; waits contributed to significant fatigue.

Patients and families felt included in some aspects of post-acute rehabilitation care planning, but felt that input was not always attended to/heard.

N/A

INTERPRETING NEUTRAL RESULTS IN AN INTERVENTION AIMED AT PROMOTING PSYCHOSOCIAL WELL-BEING FOLLOWING STROKE: A PROCESS EVALUATION

L. Bragstad^1,2,3, E.G. Hjelle^1,2, U. Sveen^3,4,5, B.A. Bronken⁶, R. Martinsen⁶, K. Kvigne⁷, G. Kitzmüller⁸, M. Mangset³ and M. Kirkevold^1,2

¹University of Oslo, Department of Nursing Science, Oslo, Norway

²University of Oslo, Research Center for habilitation and rehabilitation services and models CHARM, Oslo, Norway

³Oslo University Hospital, Department of Geriatric Medicine, Oslo, Norway

⁴Oslo Metropolitan University, Faculty of Health Sciences, Oslo, Norway

⁵Oslo University Hospital, Department of Physical Medicine and Rehabilitation, Oslo, Norway

⁶Inland Norway University of Applied Sciences, Department of Health Studies, Elverum, Norway

⁷Inland Norway University of Applied Sciences, Department of Nursing, Elverum, Norway

⁸UIT- The Arctic University of Norway, Department of Health and Care Sciences, Narvik, Norway

An intervention promoting psychosocial well-being following stroke was tested in a multicenter randomized controlled trial (RCT). A process evaluation was conducted to strengthen the quality of the trial, and support interpretation of trial outcomes. The aim was to combine results from the RCT and the process evaluation to evaluate how the implementation of the intervention may have influenced the results of the trial.

The RCT enrolled patients from eleven hospitals in South-Eastern Norway. In-person assessment interviews were conducted with all participants at baseline, 6-, and 12-months post-stroke. An explanatory sequential mixed method process evaluation was conducted alongside the clinical trial (Figure 1). The primary outcome was psychosocial well-being at 12 months post-stroke, measured by the General Health Questionnaire-28 (GHQ-28). Qualitative and quantitative process data was collected from the trial coordinators and trial participants for the process evaluation.

In total, 322 patients were randomized into the intervention (n = 166) and the control group (n = 156). Relative to baseline, both groups improved significantly on the primary outcome at 12 months, however, there were no between-group differences in psychosocial well-being. The intervention was delivered with high implementation fidelity, but the implementation was influenced by several moderating factors. Participant responsiveness, quality of delivery, context, and recruitment were factors that may have influenced the results of the RCT.

This process evaluation sheds light on factors that may have influenced the neutral results of the RCT. These factors include assessment of what constitutes usual care and how the intervention was implemented.

ClinicalTrials.gov, NCT02338869.

COMPUTER GAME BASED THERAPY FOR POST-STROKE UPPER LIMB IMPAIRMENTS – A META ANALYSIS.

D. Gandhi¹, A. Sterba², M. Kate³, H. Khatter³ and J. Pandian⁴

¹Christian Medical College and Hospital, Department of Neurology & College of Physiotherapy, Ludhiana, India

²Masaryk University, Faculty of Medicine, Brno, Czech Republic

³Christian Medical College and Hospital, Department of Neurology, Ludhiana, India

⁴Christian Medical College and Hospital, Department of Neurology & College of Medical Sciences, Ludhiana, India

The need for simplified repetitive task oriented rehabilitation program for upper limb motor recovery after stroke has been established. This aids changes in representational organization in motor cortex. Computer Game Based Therapy (CGBT) is one such mode being researched widely. We aim to determine if CGBT (using handheld remote/objects) with conventional therapy for post-stroke upper limb impairments is better than conventional therapy alone.

A literature review based on pre-set inclusion-exclusion criteria yielded 763 studies. Ten studies were included for analysis. International Classification of Impairment Disability and Handicap based outcomes were: (1) Impairment- Fugl Meyer scale (2) activity limitation- Action research arm test-ARAT, Box-block test-BBT, Wolf motor function test-WMFT and (3) Participation restriction- Functional Independence Measure-FIM. RevMan 5.0 software was used for analysis.

Figure 1: WMFT.

FMA reported (pvalue: 0.26, 95% CI, -1.44 to 5.31, I2 = 70%). Studies reporting WMFT (figure 1) (pvalue: 0.002, 95% CI, -6.46 to -1.43, I2 = 0%) and BBT (pvalue: 0.04, 95% CI, 0.14 to 6.07, I2 = 65%) showed differences with adjunctive CGBT for 30–60 minutes sessions continued over 10 days to 6 weeks. ARAT (pvalue: 0.63, 95% CI, -2.47 to 4.08, I2 = 0%) and FIM showed (pvalue: 0.52, 95% CI, -2.6 to 5.23, I2 = 0%).

Short-term adjunctive CGBT for upper limb stroke rehabilitation may improve fine and gross motor movements. With growing digitalization CGBT seems promising especially when delivered as a community based rehabilitation. Identification of long term effects of CGBT aid to the comprehensiveness of CGBT for post-stroke upper limb impairments

NA

Scientific Communications 08-Reperfusion Treatment (IVT)

EFFECT OF REPERFUSION THERAPIES IN STROKE PATIENTS WITH PRE-EXISTING DISABILITY

M. Millan¹, M. Gomis¹, N. Pérez de la Ossa¹, A. Ramos-Pachón¹, M. Hernández-Pérez¹, L. Dorado¹, L. Muñoz¹, M. Paré¹, N. Zhu¹, A. Sorrentino¹, E. Palomeras², C. Castaño¹, S. Remollo¹, M. Terceño³, R. García-Sort¹, O. Fagúndez¹, R. Ló pez-Castilla¹, I. Nuño¹ and A. Dávalos¹

¹Germans Trias i Pujol Hospital, Neuroscience, Badalona, Spain

²Hospital de Mataró, Neurology, Mataró, Spain

³Hospital Dr. J. Trueta, Neurology, Girona, Spain

Stroke patients with premorbid disability were excluded from trials of intravenous thrombolysis (iv tPA) or endovascular treatment (EVT), so their use in clinical practice is controversial arguing futility reasons. We aimed to analyze the prevalence of previously disabled patients treated with reperfusion therapies over the years and the clinical response to treatment.

Retrospective study of consecutive patients admitted in a single comprehensive stroke center from 2003 to 2018 treated with iv tPA or EVT(+/-iv tPA). We assess in three periods (2003-2007(n = 212), 2008–2012(n = 584), 2013–2018(n = 963)) the proportion of pre disabled treated patients. Symptomatic intracranial hemorrhage (sICH), mortality and good outcome (modified Rankin Scale (mRS) 0 or 1 or to return to prestroke mRS) at 90 days were compared between previously disabled patients (mRS 2–4) and not disabled patients (mRS 0–1).

Among 1759 patients, 1506(85.6%)/253(14.4%) had prestroke mRS 0–1/mRS 2–4, respectively. Rate of iv tPA/EVT therapy in disabled patients increased over years (0.4%/0%, 9.5%/2.3%, 12.8%/5.8% for each period). Rate of sICH was similar in both groups (3.7%/2.9%). At 90 days, mortality was significantly higher in predisabled patients (14% versus 28%, adjusted OR 1.5, 95% CI 1.1-2.2) whereas no differences were found in the proportion of patients who achieved good outcome (38.8% versus 34.6%, adjusted OR 0.7, 95% CI 0.5-1.0). Similar results were found when consider premorbid disability mRS 3–4 (n = 125).

Despite a higher mortality, previously disabled patients return as often as independent patients to their prestroke level of function supporting not being routinely excluded from reperfusion therapies.

N/A

TENECTEPLASE 0.4 MG/KG IN MODERATE AND SEVERE ISCHEMIC STROKE – RESULTS FROM THE NORWEGIAN TENECTEPLASE STROKE TRIAL (NOR-TEST)

C.E. Kvistad¹, V. Novotny¹, L. Thomassen¹ and N. Logallo²

¹Haukeland University Hospital, Department of Neurology, Bergen, Norway

²Haukeland University Hospital, Department of Neurosurgery, Bergen, Norway

Tenecteplase may represent an alternative to alteplase as thrombolytic treatment in acute ischemic stroke. There is limited data on tenecteplase 0.4 mg/kg in patients with increased stroke severity. We aimed to assess safety and efficacy of tenecteplase 0.4 mg/kg in patients with moderate and severe ischemic stroke included in the Norwegian Tenecteplase Stroke Trial (NOR-TEST).

NOR-TEST was a phase III trial designed to investigate safety and efficacy of tenecteplase 0.4 mg/kg versus alteplase 0.9 mg/kg in ischemic stroke. Moderate stroke was defined as admission NIHSS 6–14 and severe stroke as NIHSS ≥15. Stroke mimics were excluded. Rates of favorable outcome at 90 days, symptomatic intracerebral hemorrhage (sICH) and mortality after 7 and 90 days were assessed.

In patients with moderate stroke (n = 261) and severe stroke (n = 87), there were no differences in rates of favorable outcome, sICH or mortality at 7 days between those receiving tenecteplase as compared to alteplase. All cause mortality at 90 days was increased in patients with severe stroke treated with tenecteplase (10 [26.3%] vs. 4 [9.1%], p = 0.045). One patient died of sICH in the tenecteplase group and two patients died of sICH in the alteplase group.

Tenecteplase 0.4 mg/kg showed similar rates of favourable outcome and sICH in patients with moderate and severe stroke as compared to alteplase. However, mortality after 90 days was increased in patients with severe stroke receiving tenecteplase. Future studies assessing tenecteplase 0.4 mg/kg should closely monitor safety parameters in patients with severe stroke.

NOR-TEST was registered with ClinicalTrials.gov, number NCT01949948.

SAFETY AND EFFICACY OF DUAL ANTIPLATELET PRETREATMENT IN ACUTE ISCHEMIC STROKE PATIENTS TREATED WITH INTRAVENOUS THROMBOLYSIS: A SYSTEMATIC REVIEW AND META-ANALYSIS

A. Katsanos¹, K. Malhotra², N. Goyal³, N. Ahmed⁴, D. Strbian⁵, L. Palaiodimou¹, T. Karapanayiotides⁶, A. Alexandrov³, J. Grotta⁷, A. Alexanrov³ and G. Tsivgoulis^1,3

¹National and Kapodistrian University of Athens, Second Department of Neurology, Athens, Greece

²Charleston Area Medical Center- West Virginia University, Department of Neurology, Charleston, USA

³University of Tennessee Health Science Center, Department of Neurology, Memphis, USA

⁴Karolinska University Hospital, Department of Neurology, Stockholm, Sweden

⁵Helsinki University Central Hospital, Department of Neurology, Helsinki, Finland

⁶AHEPA University Hospital- Aristotle University of Thessaloniki, Second Department of Neurology, Thessaloniki, Greece

⁷Memorial Hermann Hospital -Texas Medical Center, Clinical Innovation and Research Institute, Houston, USA

Conflicting data exist regarding the safety and efficacy of IV thrombolysis (IVT) in acute ischemic stroke (AIS) patients receiving dual antiplatelet pretreatment (DAPP). The aim of the present systematic review and meta-analysis is to assess the safety and outcome of DAPP history among AIS patients treated with IVT.

We performed a comprehensive literature search in Ovid MEDLINE, Ovid Embase and Scopus to identify studies that investigated the safety and efficacy of DAPP among AIS patients treated with IVT. We defined DAPP as any antiplatelet combination of ASA, clopidogrel, dipyridamole, and any other antiplatelet agent. Odds ratios (ORs) were calculated using random-effects model.

We identified 10 studies comprising 68,037 patients. In unadjusted analyses, DAPP was associated with higher likelihood of symptomatic intracranial hemorrhage (sICH, OR = 2.26, 95% CI 1.39–3.67) and 3-month mortality (OR = 1.47, 95% CI 1.27–1.73). There was no association between DAPP and 3-month favorable functional outcome (FFO, mRS 0–1), and 3-month functional independence (FI, mRS 0–2). In adjusted analyses, history of DAPP was not associated with pooled sICH (OR = 2.03, 95% CI 0.75–5.25), 3-month mortality (OR = 1.11, 95% CI 0.87–1.40), 3-month FFO (OR = 0.92, 95% CI 0.77–1.09), and 3-month FI (OR = 1.01, 95% CI 0.89–1.15).

After adjustment for potential confounders DAPP was not associated with higher risk of adverse outcomes in AIS patients treated with IVT. History of DAPP should not be used as a reason to withhold IVT in otherwise eligible AIS patients.

N/A

THE ENHANCED CONTROL OF HYPERTENSION AND THROMBOLYSIS STROKE STUDY (ENCHANTED) BLOOD PRESSURE (BP) ARM: COMPARATIVE EFFECTS OF INTENSIVE- VERSUS STANDARD-BP LOWERING TREATMENT IN LARGE-VESSEL OCCLUSION

J. Minhas¹, X. Wang², J. Chalmers², C. Anderson^3,4, T. Robinson^1,5; ENCHANTED Investigators

¹Department of Cardiovascular Sciences, University of Leicester, Leicester, United Kingdom

²The George Institute for Global Health, University of New South Wales, Sydney, Australia

³The George Institute China, Peking University Health Science Centre, Beijing, China

⁴Neurology Department, Royal Alfred Hospital, Sydney, Australia

⁵National Institute for Health Research Leicester Biomedical Research Centre, University of Leicester, Leicester, United Kingdom

Current international guidelines mandate blood pressure (BP) treatment for values above 180/105 mmHg for acute ischaemic stroke (AIS) patients irrespective of recanalisation therapy. However, there are scarce data specifically evaluating optimal BP targets for large vessel occlusion (LVO) patients, the stroke sub-type accounting for the largest proportion of stroke-related death and disability. For LVO patients undergoing mechanical thrombectomy (MT), higher pre-treatment BP is purportedly associated with adverse outcomes. Despite confidence in the benefits of early recanalisation in LVO, BP management is an area of contention. This post-hoc analysis will assess the effect of stroke subtype, including LVO, on modifying the comparative effects of intensive- versus standard-BP lowering treatment effects.

The BP arm of the ENCHANTED trial compared a strategy of intensive (systolic BP 130–140 mmHg) versus guideline-recommended (SBP < 180 mmHg) BP lowering in 2227 thrombolysis-treated AIS patients for superiority with respect to shift (‘improvement’) in 90-day modified Rankin score and the key safety outcome of any intracranial haemorrhage. The main results will be presented at the International Stroke Conference in February 2019. This post-hoc analysis will use logistic regression analysis of patient subgroups including CT or MR angiogram-confirmed LVO and those with an NIHSS >10 likely to have LVO, as well as patients undergoing mechanical thrombectomy.

Data will be presented on patients with CT or MR angiogram-confirmed LVO, on patients undergoing mechanical thrombectomy, and on patients with NIHSS >10 likely to have LVO.

The authors will present the conclusion of the analyses described above at ESOC 2019.

ClinicalTrials.gov Identifier: NCT01422616

THE ENHANCED CONTROL OF HYPERTENSION AND THROMBOLYSIS STROKE STUDY (ENCHANTED) BLOOD PRESSURE (BP) ARM: EFFECTS OF TIMING, TARGET AND INTENSITY OF SYSTOLIC-BP CONTROL ON OUTCOME

J. Minhas¹, X. Wang², J. Chalmers², C. Anderson^3,4, T. Robinson^1,5; ENCHANTED Investigators

¹Department of Cardiovascular Sciences, University of Leicester, Leicester, United Kingdom

²The George Institute for Global Health, University of New South Wales, Sydney, Australia

³The George Institute China, Peking University Health Science Centre, Beijing, China

⁴Neurology Department, Royal Alfred Hospital, Sydney, Australia

⁵National Institute for Health Research Leicester Biomedical Research Centre, University of Leicester, Leicester, United Kingdom

Secondary stroke prevention randomised controlled trials (RCTs) have confirmed that blood pressure (BP) lowering reduces the risk of recurrent stroke. The International Stroke Trial (IST) showed a systolic BP (SBP) level around 150 mmHg was associated with lowest risks of mortality in acute ischaemic stroke (AIS). Similarly, the Safe Implementation of Thrombolysis in Stroke (SITS) data supports a potential optimum SBP of 141–150 mmHg after pharmacological reperfusion therapy. However, no randomised study to date has determined whether time-to or duration at 140 mmHg SBP target affects key safety outcomes including symptomatic intracranial haemorrhage (sICH) or haemorrhagic transformation. This post-hoc analysis will explore the effects of BP lowering, in particular, the timing, target and intensity of SBP control for optimum balance between functional recovery and risk of sICH.

The BP arm of the ENCHANTED trial compared a strategy of intensive (systolic BP 130–140 mmHg) versus guideline-recommended (SBP < 180 mmHg) BP lowering in 2227 thrombolysis-treated AIS patients for superiority with respect to shift (‘improvement’) in 90-day modified Rankin score and the key safety outcome of any intracranial haemorrhage. The main results will be presented at the International Stroke Conference in February 2019. This post-hoc analysis will use an analysis model for the primary efficacy outcome and pre-specified baseline characteristics will be explored to determine whether they modify the association of SBP parameters and outcome.

Data will be presented for all patients and outcome based on timing, target and intensity of SBP control.

The authors will present the conclusion of the analyses described above at ESOC 2019.

ClinicalTrials.gov Identifier:NCT01422616

THE INCIDENCE AND ASSOCIATIONS OF EARLY NEUROLOGICAL DETERIORATION IN POST THROMBOLYSIS PATIENTS WITH ACUTE ISCHAEMIC STROKE: RESULTS FROM THE SITS REGISTRY

W.M. Yu¹, A.C. Cameron¹, A.H. Abdul-Rahim², J. Kõrv³, P. Sevcik⁴, D. Toni⁵ and K.R. Lees⁶

¹Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, United Kingdom

²Institute of Neuroscience and Psychology, University of Glasgow, Glasgow, United Kingdom

³Department of Neurology and Neurosurgery, University of Tartu, Tartu, Estonia

⁴Department of Neurology, Faculty of Medicine in Pilsen- Charles University and Department of Neurology- University Hospital Pilsen, Plzen, Czech Republic

⁵Department of Human Neurosciences, University La Sapienza, Rome, Italy

⁶School of Medicine- Dentistry and Nursing, University Of Glasgow, Glasgow, United Kingdom

Early neurological deterioration (END) after stroke onset may indicate serious outcomes. Estimated rates of END after intravenous thrombolysis among small patient samples have varied widely, up to 29.8%. We studied the incidence and factors associated with END among patients following intravenous thrombolysis.

We analysed SITS-International registry patients with known outcomes enrolled in 2010- 2017. END was defined as an increase in NIH Stroke Scale (NIHSS) ≥4 at 24 hours from baseline NIHSS.

We determined the incidence of END, and used logistic regression models to inspect associations along with outcomes. We adjusted for variables found significant in univariate analyses (p < 0.05). Principal outcomes were incidence of END, associated predictors of END, ordinal day-90 mRS and day-90 mortality.

We excluded 53,539 patients and included 50,726 patients. The incidence of END was 3415/50726 (6.7%; 95% CI 6.5-7.0%).

Among several significant variables, factors independently associated with END on multivariate analysis were large vessel disease (LVD) with carotid stenosis (OR 3.0, 95% CI 2.47-3.63, p < 0.001), other LVD (OR 2.63, 95% CI 2.21-3.13, p < 0.001), and ischaemic stroke versus TIA/other (OR 3.47, 95% CI 2.56-4.70, p < 0.001).

END associated with worsening outcome on ordinal mRS with adjusted OR 3.36 (95% CI 3.27-2.44, p < 0.001) by day-90 compared to no END.

The adjusted OR for day-90 mortality was 10.40 (95% CI 9.02-12.00, p < 0.001).

The routinely observed rate of END reflected by real-world data is low, but END greatly increases risk of disability and mortality. Readily identifiable factors predict END, and may help with understanding causal mechanisms to assist prevention of END.

N/A

METANALYSIS OF GENOME-WIDE ASSOCIATION STUDIES IDENTIFY A NEW SUGGESTIVE LOCUS ASSOCIATED WITH RTPA-INDUCED PARENCHYMAL HEMATOMA

C. Carrera Vasconez¹, N. Cullell², N. Torres-Águila³, E. Muiño⁴, J. Castillo⁵, A. Dávalos⁶, A. Bustamante⁷, C. Vives-Bauza⁸, V. Obach⁹, T. Segura¹⁰, J. Martí-Fàbregas¹¹, J. Jiménez-Conde¹², J. Roquer¹², A. Slowik¹³, T. Tatlisumak¹⁴, S. Daniel¹⁵, C. Cruchaga¹⁶, J.M. Lee¹⁷, J. Montaner¹⁸, I. Fernández-Cadenas⁴; on behalf of the International Stroke Genetics Consortium (ISGC) and the GENESTROKE Consortium

¹Vall d’Hebron Research Institute VHIR- Universitat Autònoma de Barcelona – IIB- Sant Pau, Neurovascular Research Laboratory – Stroke Pharmacogenomics and Genetics, Barcelona, Spain

²Fundació Docència i Recerca Mútua Terrassa – IIB- Sant Pau, Stroke Pharmacogenomics and Genetics, Barcelona, Spain

³Vall d’Hebron Research Institute VHIR – IIB- Sant Pau, Neurovascular Research Laboratory – Stroke Pharmacogenomics and Genetics, Barcelona, Spain

⁴IIB- Sant Pau, Stroke Pharmacogenomics and Genetics, Barcelona, Spain

⁵Health Research Institute of Santiago de Compostela IDIS, Clinical Neurosciences Research Laboratory, Santiago de Compostela, Spain

⁶Hospital Germans Trias i Pujol, Department of Neuroscience, Badalona, Spain

⁷Vall d’Hebron Research Institute VHIR, Neurovascular Research Laboratory, Barcelona, Spain

⁸Institut d’Investigacio Sanitaria de Palma IdISPa, Neurobiology Laboratory, Palma de Mallorca, Spain

⁹Hospital Clínic i Provincial de Barcelona, Department of Neurology, Barcelona, Spain

¹⁰Hospital Universitario de Albacete, Department of Neurology, Albacete, Spain

¹¹Hospital de la Santa Creu i Sant Pau- IIB- Sant Pau, Department of Neurology, Barcelona, Spain

¹²IMIM-Hospital del Mar, Department of Neurology- Neurovascular Research Group, Barcelona, Spain

¹³Jagiellonian University Medical College, Deartment of Neurology, Krakow, Poland

¹⁴Sahlgrenska Academy at University of Gothenburg and Sahlgrenska University Hospital, Department of Clinical Neuroscience/Neurology, Gothenburg, Sweden

¹⁵Department of Neurology, Helsinki University Hospital, Helsinki, Finland

¹⁶Washington University School of Medicine, Department of Psychiatry, Saint Louis, USA

¹⁷Washington University School of Medicine, Department of Neurology, Saint Louis, USA

¹⁸Virgen del Rocío & Macarena Hospitals- IBIS – Vall d’Hebron Research Institute VHIR, Department of Neurology – Neurovascular Research Laboratory, Sevilla, Spain

Hemorrhagic transformation is a serious complication of tissue-plasminogen activator (rt-PA) treatment. Its most severe form, parenchymal hematoma (PH), can result in neurological deterioration and death. We aimed to identify genetic loci associated with rt-PA-induced PH in stroke patients.

We performed a genome-wide metanalysis combining two GWAS strata in 1,870 ischemic stroke patients treated with rt-PA. Case-control strata was constructed identifying samples genotyped on similar arrays and ancestral background. Cases were defined as the presence of PH-1 or PH-2 by ECASS criteria on the 24-hours follow up CT. In addition, hemorrhagic infarcts and remote parenchymal hematoma were excluded. Quality controls, 1000G imputation and association analysis were testing separately. Subsequently, the data were combined in a fixed-effect metanalysis. We included as covariates 3 principal components, sex, age and baseline NIHSS. Expression quantitative trait loci (eQTLs) were evaluated by Genotype-Tissue Expression (GTEx) portal.

Analysis after quality controls comprised 6,324,092 single nucleotide polymorphisms (minor allele frequency >1%).127 patients (6.7%) presented PH. The strongest association signal was reached on locus at 7q.21 (p-value: 6.10x10-08; OR:4.18 95% CI: 2.49-7.03). Local eQTL effects were not detected by GTEx for the top SNP.

The candidate region is located in an intronic region of a gene expressed in neuronal and glial cells. Previously, an association of this gene with blood-brain-barrier permeability and brain damage after cerebral ischemia was reported in murine models.

Our results identify a new locus at 7q21 associated with PH. Future studies should elucidate the role of these genetic polymorphisms in the development of hemorrhagic events.

N/A

PREDICTION OF SYMPTOMATIC INTRACRANIAL HEMORRHAGE AFTER ENDOVASCULAR TREATMENT FOR ISCHEMIC STROKE: PERFORMANCE OF 6 PREDICTION MODELS BASED ON STUDIES WITH IV THROMBOLYTICS ALONE

N. van der Ende^1,2, E. Venema^1,3, B. Roozenbeek^1,2, H. Lingsma³, M. Kappelhof⁴, C. Majoie⁴, J. Boiten⁵, A. van der Lugt², D. Dippel¹; on behalf of the MR CLEAN Registry investigators

¹Erasmus MC University Medical Center, Neurology, Rotterdam, The Netherlands

²Erasmus MC University Medical Center, Radiology and Nuclear Medicine, Rotterdam, The Netherlands

³Erasmus MC University Medical Center, Public Health, Rotterdam, The Netherlands

⁴Academic Medical Center, Radiology and Nuclear Medicine, Amsterdam, The Netherlands

⁵Haaglanden Medical Center, Neurology, The Hague, The Netherlands

Several prediction models for symptomatic intracranial hemorrhage (sICH) after intravenous thrombolytics (IVT) for ischemic stroke exist and most have shown reasonable performance (C-statistics for external validation in patients treated with IVT varied between 0.56–0.70). Prediction models to identify patients with an increased risk of sICH after treatment with endovascular thrombectomy (EVT) are not available. We aimed to assess and compare the performance of prediction models for sICH after IVT in patients treated with IVT and EVT.

We used data from the MR CLEAN Registry (n = 1488) and included patients treated with IVT and EVT (n = 1161). Six existing prediction models for sICH were evaluated: MSS, HAT, SEDAN, GRASPS, SITS and SPAN-100. Performance of the models was expressed as discrimination (C-statistic) and calibration (slope, ideally 1 and intercept, ideally 0).

SICH occurred in 69 (5.9%) patients. The C-statistic varied between 0.52 and 0.64 (Table 1). SITS had the highest discriminative ability compared with other models. Intercepts varied from 0.00 to 0.27, slopes varied from 0.79 to 1.70 (Table 1).

Prediction models for sICH performed at most moderate in patients treated with IVT and EVT, and discriminative performance was worse than previously reported in patients treated with IVT alone. Therefore, a new prediction model for sICH in patients treated with EVT is needed.

N/A

SIMULATION-BASED TRAINING OF THE RAPID EVALUATION AND MANAGEMENT OF ACUTE STROKE (STREAM) – A MULTICENTRIC PROSPECTIVE INTERVENTIONAL TRIAL

F. Bohmann¹, N. Kurka¹, K. Gruber¹, J. Guenther¹, H. Rai¹, P. Rostek¹, R. Du Mesnil², P. Zickler³, M. Ertl³, A. Berlis⁴, S. Poli⁵, A. Mengel⁵, P. Ringleb⁶, S. Nagel⁶, J. Pfaff⁷, F. Wollenweber⁸, L. Kellert⁸, M. Herzberg⁹, L. Koehler¹⁰, K.G. Haeusler¹⁰, A. Alegiani¹¹, C. Schubert¹¹, C. Brekenfeld¹², C. Doppler¹³, Ö. Onur¹³, C. Kabbasch¹⁴, W. Pfeilschifter¹; STREAM Trial Investigators

¹University Hospital Frankfurt, Neurology, Frankfurt, Germany

²University Hospital Frankfurt, Neuroradiology, Frankfurt, Germany

³Klinikum Augsburg, Neurology, Augsburg, Germany

⁴Klinikum Augsburg, Neuroradiology, Augsburg, Germany

⁵University Hospital Tuebingen, Neurology, Tuebingen, Germany

⁶University Hospital Heidelberg, Neurology, Heidelberg, Germany

⁷University Hospital Heidelberg, Neuroradiology, Heidelberg, Germany

⁸University Hospital Munich, Neurology, Munich, Germany

⁹University Hospital Munich, Neuroradiology, Munich, Germany

¹⁰University Hospital Berlin – Charite, Neurology, Berlin, Germany

¹¹University Medical Centre Hamburg Eppendorf, Neurology, Hamburg, Germany

¹²University Medical Centre Hamburg Eppendorf, Neuroradiology, Hamburg, Germany

¹³University Hospital Cologne, Neurology, Cologne, Germany

¹⁴University Hospital Cologne, Neuroradiology, Cologne, Germany

Acute stroke care is an interdisciplinary multiprofessional task. The aim of this trial is to investigate the effect of a multimodal intervention comprising principles of crew resource management (CRM) and stroke team simulation training in a controlled prospective pretest-posttest design.

We are conducting a five-step intervention at 7 tertiary care neurocentres in Germany. The intervention consists of: (1) analysis of the current acute stroke workflow, (2) presentation and critical peer-to-peer review during a central meeting leading to a revision of the algorithm procedures where needed. (3) A train-the-trainer seminar on medical simulation and debriefing followed by (4) in situ simulation-based stroke team training with comprehensive debriefing at all sites implementing the revised algorithm and (5) the provision of stroke team training materials to each centre.

Acute stroke process times and staff surveys will be collected during two observation phases of three months each before and after the intervention. Primary outcome measure is the median door-to-needle time. Secondary outcome measures include the median thrombectomy process times of patients receiving thrombectomy and hemorrhagic transformation in follow-up imaging as a safety endpoint. Staff satisfaction, perceived patient safety and the staff’s stroke readiness are assessed by means of a pre-designed questionnaire. https://clinicaltrials.gov/ct2/show/NCT03228251.

188 datasets were collected in the preinterventional observation phase with a mean door-to-needle time of 34 minutes (IQR 25–43 minutes). Data collection of the second observation phase will be completed in March 2019. We aim to present the final results at ESOC 2019.

N/A

N/A

REGULAR SIMULATION-TRAINING IN STROKE THROMBOLYSIS; ARE SHORT DOOR-TO-NEEDLE TIMES SUSTAINABLE DESPITE A HIGH TURNOVER OF NEUROLOGY REGISTRARS?

S.C. Ajmi¹, K.O. Sandve², R.K. Advani³, S.A. Qvindesland⁴, L. Fjetland², K.D. Kurz², H.L. Ersdal⁵ and F.M.W. Kurz¹

¹Stavanger University Hospital, Neurology, Stavanger, Norway

²Stavanger University Hospital, Radiology, Stavanger, Norway

³Ostfold Hospital Kalnes, Neurology, Fredrikstad, Norway

⁴Stavanger University Hospital, Education and Research, Stavanger, Norway

⁵Stavanger University Hospital, Anesthesiology and Intensive Care, Stavanger, Norway

In acute stroke, rapid revascularization is crucial for good patient outcome. We reduced median door-to-needle time from 27 to 13 minutes in our center 6 months after implementing simulation-training as part of a quality improvement project. Studies show that experienced neurology registrars achieve shorter treatment times than inexperienced. Maintaining short treatment times in a hospital with high staff turnover can thus be challenging. The aim was to assess whether short door-to-needle times can be sustained through regular simulation-training, despite high staff turnover.

During a 22-month period starting February 2017 we completed three clusters of weekly simulation-based team-training sessions. In this period a total of 11 inexperienced registrars were employed, 10 of these between 12–22 months after quality improvement leaving 4 experienced registrars. We therefore compared median door-to-needle time 0–11 and 12–22 months after quality improvement. We also compared treatment times for inexperienced neurology registrars before and within 6 months of attending a simulation-training session.

In the 22-month period, 298 patients with suspected acute ischemic stroke were treated with intravenous thrombolysis (Table 1). Median door-to-needle time was 14 minutes (IQR 8–25) 0–11 months after quality improvement compared to 15 minutes (IQR 10–26, p = 0.731) after 12–22 months (Figure1). Median door-to-needle time for inexperienced registrars was 21 minutes (IQR 15–45) compared to 15 minutes (IQR 10–22, p = 0.001) for inexperienced registrars that had attended a simulation session.

Clusters of weekly in-situ simulation-based team-training sessions can sustain short door-to-needle times even with high staff turnover. Simulation-training sessions significantly reduce treatment times for inexperienced registrars.

N/A

ACCESS TO MECHANICAL THROMBECTOMY FOR CEREBRAL ISCHAEMIA: A POPULATION-BASED STUDY IN THE NORTH-OF-FRANCE

D. Leys¹, N. Dequatre-Ponchelle², M. Ferrigno¹, H.C. Henon¹, F. Mounier-Vehier³, S. Moulin¹, B. Casolla¹, R. Tortuyaux², M. Chochoi², C. Moreau¹, I. Girard-Buttaz⁴, J.P. Pruvo⁵, P. Goldstein⁶, C. Cordonnier¹; the Stroke 59–62 network.

¹University of Lille. Inserm U1171. CHU Lille., Neurology., Lille, France

²Lille University Hospital, Neurology, Lille, France

³Lens Hospital, Neurology, Lens, France

⁴Valenciennes, Neurology, Valenciennes, France

⁵University of Lille. Inserm U1171. CHU Lille., Neuroradiology., Lille, France

⁶Lille University Hospital, Emergency department. SAMU 59, Lille, France

Local organisations for access to mechanical thrombectomy (MT) are based on facilities that were already available when MT was not yet proven effective. We aimed at evaluating MT rates and outcomes at the population level.

We evaluated in 2016–2017 rates of MT and outcomes in inhabitants from the North-of-France (NoF) and its 12 districts. The NoF has 12 stroke units for 4 million inhabitants, including one with MT (Lille). Nowhere in the region can patients go faster to a MT centre other than Lille, the closest being Kortrijk and Amiens. We defined good outcomes at 3 months as modified Rankin scales (mRS) 0–2 or similar the pre-stroke mRS.

Six-hundred and sixty-six patients underwent MT (with intravenous thrombolysis [IVT] in 454, 68.1%). Besides, 1,595 received IVT alone. The rate of MT was 81 (95% confidence interval [CI] 72–90) per million inhabitants-year, ranging from 36 to 108 between districts. The rate of IVT was 249 (95% CI 234–264), ranging from 155 to 268. After 3 months, 279 (41.9%) patients who underwent MT had good outcomes, and 167 (25.1%) had died. Patients living outside the district of Lille were less likely to have good outcomes after adjustment on age, sex, baseline severity, and delay (odds ratio between 0.55 and 0.58 for good outcomes, and between 1.2 and 2.1 for death, Lille being the reference).

Although MT rates were high, outcomes were significantly worse in patients living outside the district of Lille. This is not explained by differences in delays.

N/A

FIVE-YEAR EVOLUTION OF REPERFUSION THERAPY AND EARLY FUNCTIONAL OUTCOME AND MORTALITY WITH ACUTE ISCHEMIC STROKE IN FRANCE

C. El Khoury¹, C. Aboa-Eboulé², L. Fraticelli¹, O. Guerrier³, M. Bischoff¹, M. Buisson⁴, S. Cakmak⁵, F. Philippeau⁶, P. Serre¹, K. Blanc-Lasserre⁷, A.E. Vallet⁸, N. Nighoghossian³, T.H. Cho³, L. Mechtouff³, E. Ong³, L. Derex³; the RESUVal Group

¹RESUVal, Hospital Center Lucien Hussel, Vienne, France

²Department of ambulatory cardiology, Louis Pradel Hospital- Hospices Civils de Lyon HCL, Bron, France

³Stroke Center- Department of Neurology, Hô pital Neurologique- Hospices Civils de Lyon-, Bron, France

⁴Clinical Investigation Center- Inserm 1407, Louis Pradel hospital- Hospices Civils de Lyon, Bron, France

⁵Hospital Center Villefranche-sur-Saône, Department of Stroke Medicine, Villefranche-sur-Saône, France

⁶Hospital Center Bourg-en-Bresse, Department of Stroke Medicine, Bourg-en-Bresse, France

⁷Hospital Center Valence, Department of Stroke Medicine, Valence, France

⁸Hospital Center Lucien Hussel, Department of Stroke Medicine, Vienne, France

Our study aimed to analyze the temporal trend of thrombolysis rate after the implementation of a regional emergency network (RESUVal network) for acute ischemic stroke (AIS) through a real-life registry.

In 2010, a regional emergency network was developed based on cooperation between emergency physicians and neurologists. The network includes 5 Primary (PSC), one Comprehensive Stroke Center (CSC) and 37 emergency departments covering a population of 3 million of inhabitants.

From 2010 to 2015, 18 986 AIS were recorded in the regional Hospitalization Database, among them 2 215 thrombolyzed AIS enrolled in the RESUVal registry, with an overall incidence of 11.7%. Median age remained stable around 74 years old [63; 82] but elderly (>80) increased over time from 23,65% in 2010 to 32,81% in 2015 (p = 0.0003). The annual incidence rate of thrombolysis increased modestly around 1% increment from the first year (10.20%) to the second (11.05%) and third year (12.67%) and remained stable in the fourth (12.19%), and the fifth (12.09%) until it increased again in the sixth (14.5%). Patients treated within 3 hours (60.6%), between 3 and 4.5 hours (30.8%), and beyond 4.5 hours (4.7%) of onset symptoms remained stable. About 40% of the patients had a good early recovery after thrombolysis and the 3O-day all-cause mortality rate remained stable at 12.12%. At 3 months, 44.33% of patients presented a minor disability (mRS < 2).

The network has enabled emergency physicians to be more responsive, improving the thrombolysis rate with a 3-month evolution of functional outcome benefitted for patients.

N/A

ESTIMATING THE EFFECTIVENESS AND COST-EFFECTIVENESS OF ESTABLISHING ADDITIONAL ENDOVASCULAR THERAPY STROKE CENTRES IN ENGLAND: A DISCRETE EVENT SIMULATION

P. White¹, P. McMeekin², D. Coughlan³, M. Allen⁴, G. Ford⁵; The PEARS Collaborators

¹Newcastle University, Institute of Neuroscence, Newcastle, United Kingdom

²Northumbria University, Health and Life Sciecnes, Newcastle, United Kingdom

³Newcastle University, Institute of Health and Scoiety, Newcastle, United Kingdom

⁴University of Exeter, Medical School, Exeter, United Kingdom

⁵Oxford University, Radcliffe Department of Medicine, Oxford, United Kingdom

In acute stroke treatment time is brain. We have previously modelled that the optimal number of thrombectomy centres in England would be 30 (net 6 new centres). We now estimate the relative effectiveness and cost-effectiveness over 12-months associated with increasing the number of centres from 24 to 30

We constructed a discrete event situation (DES) to estimate the effectiveness and lifetime cost-effectiveness using 1 year’s incidence of stroke. 2,000 iterations of the simulation were performed comparing baseline 24 centres with reconfiguration to 30.

21,740 stroke patients would potentially be affected by the change in service configuration. Of these the mean number of patients treated with EVT would be 2,538 (SD 6.9). The median time to treatment for eligible early presenters (< 270 minutes since onset of symptoms) would reduce from 195 (IQR 155–249) to 165 (IQR 105–224) minutes. The extra 6 centres overall generate an additional 190 QALYS (95% CI -6, 399 and resulted in savings to the healthcare system of £1,864,156 (95% CI -1,203,690, £5,017,356). Our model predicts reconfiguration would lead to an additional 16 mRS 0, 10 mRS 1, 4 mRS 2 patients and 5, 8, 9 & 6 fewer patients in mRS categories 3 to 6 respectively.

Changes in acute stroke service configuration will produce clinical and cost benefits when the time taken for patients to receive treatment is reduced. Benefits are likely to be cost saving before any capital investment is considered. Results are particularly sensitive to how resulting QALYs are valued and future care costs.

N/A

THE STOCKHOLM STROKE TRIAGE STUDY – RAPID THROMBECTOMY USING A THREE STEP PREHOSPITAL TRIAGE SYSTEM

M. Mazya¹, A. Berglund¹, N. Ahmed¹, S. Holmin², A.C. Laska³, J. Mathé⁴, C. Sjöstrand¹ and E. Eriksson¹

¹Karolinska University Hospital, Department of Neurology, Stockholm, Sweden

²Karolinska University Hospital, Department of Neuroradiology, Stockholm, Sweden

³Karolinska Institutet, Department of Clinical Sciences, Stockholm, Sweden

⁴Capio S:t Göran Hospital, Department of Neurology, Stockholm, Sweden

In a 2.3M population region (6519 km2), with one comprehensive and six primary stroke centres (CSC, PSC), we aimed to accelerate delivery of endovascular thrombectomy (EVT) by implementing a prehospital triage system for PSC bypass based on symptom assessment and ambulance-hospital teleconsultation.

Patients with suspected stroke assessed by ambulance nurses, were evaluated for moderate-to-severe hemiparesis (≥2 NIHSS points for ipsilateral arm and leg: the A2L2 test). If present, ambulances teleconsulted the CSC for EVT eligibility and direction to CSC or nearest PSC. If absent, ambulances prenotified the nearest hospital. Data were prospectively collected for all stroke-alert ambulance transports, October 2017-October 2018. We evaluated the system’s predictive performance for large artery occlusion (LAO) acute ischemic stroke (AIS), and EVT initiation. Onset-to-treatment times were compared with the previous year, when patients were taken to the nearest hospital without teleconsultation.

We recorded 2905 stroke-alert ambulance transports, with 323 (11%) A2L2+teleconsult positive cases triaged to CSC bypassing nearest PSC. System accuracy for AIS+LAO was 87% (PPV 41%, NPV 93%), and EVT initiation 91% (PPV 26%, NPV 99%). EVT was performed in 84/323 (26%) triage-positive, and 35/2582 (1.4%) triage-negative patients. In EVT cases with known onset time (n = 77), median onset-to-groin was 137 min (previous year: 203 min, n = 76), p < 0.001. Median onset-to-needle time (IV thrombolysis) was unchanged: 115 min (n = 337) versus 116 min (n = 351) the previous year, p = 0.49.

The Stockholm Stroke Triage system, combining symptoms and teleconsultation, results in markedly reduced time from stroke onset to thrombectomy, without delaying intravenous thrombolysis treatment.

N/A

EARLY DETECTION OF LARGE VESSEL OCCLUSIONS: A COMPARISON OF PREHOSPITAL STROKE SCALES FOR DIFFERENT OCCLUSION LOCATIONS IN THE MR CLEAN REGISTRY

M. Duvekot^1,2, E. Venema^2,3, H.F. Lingsma³, J.M. Coutinho⁴, H.B. van der Worp⁵, J. Hofmeijer⁶, A. van der Lugt⁷, H. Kerkhoff¹, D.W.J. Dippel², B. Roozenbeek^2,7; the MR CLEAN Registry investigators

¹Albert Schweitzer Hospital, Neurology, Dordrecht, The Netherlands

²Erasmus MC University Medical Center, Neurology, Rotterdam, The Netherlands

³Erasmus MC University Medical Center, Public Health, Rotterdam, The Netherlands

⁴Amsterdam University Medical Centers, Neurology, Amsterdam, The Netherlands

⁵University Medical Center Utrecht, Neurology and Neurosurgery, Utrecht, The Netherlands

⁶Rijnstate Hospital, Neurology, Arnhem, The Netherlands

⁷Erasmus MC University Medical Center, Radiology and Nuclear Medicine, Rotterdam, The Netherlands

Several prehospital stroke scales exist to identify stroke patients with a large vessel occlusion (LVO) to allow transport directly to an endovascular-capable center. Our aim was to compare the sensitivities of these prehospital stroke scales, specifically for different occlusion locations.

We used data of patients included in the MR CLEAN Registry of endovascular treatment for anterior circulation ischemic stroke (n = 1488). Four existing scales, RACE, C-STAT, PASS and G-FAST, were reconstructed based on NIHSS scores on arrival at the emergency department. We used cut points as proposed in the original studies. Primary outcome was the presence of a core-lab assessed intracranial LVO. Patients with A1, A2, M3 occlusions and patients without an occlusion were excluded. Sensitivity of each scale was calculated in all patients, and for specific occlusion locations.

1395 patients were available for analysis. Sensitivities varied between the scales and occlusion locations (Table 1). G-FAST had the highest sensitivity for all types of LVO, ranging from 0.89-0.99. All scales had lower sensitivities for M2 occlusions.

Table 1. Sensitivity per occlusion location (95% confidence interval)

Although prehospital stroke scales are generally sensitive for large vessel occlusions, they are less effective in detecting more distal occlusions. Prospective validation of these scales in the prehospital setting is currently carried out in the PRESTO study in the Netherlands. (https://www.presto-studie.nl)

N/A

IMPROVING TREATMENT TIMES FOR ACUTE STROKE WITH A SINGLE COMMUNICATION SYSTEM FOR INTERDISCIPLINARY CARE: FINAL RESULTS FROM IMPLEMENTING A SMARTPHONE COMMUNICATION APP

K. Bagot^1,2, D. Cadilhac^1,2, K. Smith³, S. Bernard³, J. Kim^1,2, T. Coupland⁴, D. Pearce⁵, D. Badcock⁵, M. Budge⁴, V. Nadurata⁴, W. Pearce³, H. Hall⁶, B. Kelly⁵, A. Spencer⁵, P. Chapman⁵, E. Oqueli⁵, R. Sahathevan^1,5, T. Kraemer⁵, G. Hocking⁶, C. Bladin^1,2,3

¹The Florey Institute, Public Health, Melbourne, Australia

²Monash University, Department of Medicine- Southern Clinical School, Melbourne, Australia

³Ambulance Victoria, Emergency medicine, Melbourne, Australia

⁴Bendigo Health, Medicine, Bendigo, Australia

⁵Ballarat Health Service, Medicine, Ballarat, Australia

⁶Ambulance Victoria, Medicine, Melbourne, Australia

Delivery of acute stroke care relies on clinicians communicating clinical information efficiently. Varied communication methods are used between paramedics and hospital clinicians (emergency, medical, neurology, radiology). Details are often repeated, contributing to delays. The aim is to determine if a digital communication app can improve care timelines for patients with suspected acute stroke.

A 12-month pre-post design was used. The Pulsara™ Stop Stroke/STEMI communication app was implemented pre-hospital (25 Ambulance Victoria branches) and within-hospital (2 rural hospitals) in Victoria, Australia. Pulsara provides secure, simultaneous, two-way, real-time communication. Eligible patients with suspected acute stroke assessed by paramedics or ED clinicians. Pre-hospital and hospital assessment/treatment times were recorded between August 2016 and August 2018. Median minutes (IQRS) were compared between two groups: Pulsara initiated (Pulsara), or not initiated (no Pulsara).

The no Pulsara (n = 213) and Pulsara (n = 391) groups were of similar age and sex (76 vs 75 years; 50% male). Pulsara was initiated by paramedics (n = 223) and ED (n = 168) clinicians. Patients were off ambulance stretchers faster by 8 minutes (no Pulsara: 19 minutes [11–29]; Pulsara 11 minutes [7–17]; p = .0001). Emergency department door-to-first medical review was faster by 17 minutes (no Pulsara: 23 minutes [8–67]; Pulsara 6 minutes [2–14]; p = .0004). Door-to-CT times were 44 minutes faster (no Pulsara: 71 minutes [43–147]; Pulsara: 27 minutes [18–44]; p = .0001). Pulsara was used in 96% (52/54) of tPA cases. The proportion of patients with door-to-needle times < 60 minutes was 12% pre-Pulsara, compared to 26% with-Pulsara.

Using a digital communication app improves acute stroke treatment times.

N/A

STROKE CARE AND TELEMEDICINE: FULL ONSITE EXPERTISE BETTER THAN TELEMEDICINE ASSISTED STROKE UNITS?

V. Zietemann¹, J. Held², F. Kraus¹, S. Platen³, R. Haberl¹, G.J. Hubert⁴

¹Munich Clinic – affiliated to University of Munich, TEMPiS Telemedical Stroke Center – Department of Neurology, Munich, Germany

²BAQ, Bayerische Arbeitsgemeinschaft für Qualitätssicherung in der stationären Versorgung, Munich, Germany

³University of Regensburg – Bezirksklinikum- Regensburg, Department of Neurology, Regensburg, Germany

⁴Munich Clinic, TEMPiS Telemedical Stroke Center – Department of Neurology, Munich, Germany

The Telestroke network TEMPiS in Germany comprises telemedicine assisted stroke units (TSU) with variable onsite expertise. Aim was to compare quality of TSU with only low-level (part-time) onsite neurology expertise (n = 13) to all nationally certified regional stroke units (cSU) with full time neurology expertise in the state of Bavaria (n = 22).

Quality data of all patients diagnosed with stroke in Bavaria is collected in compulsory Bavarian Stroke Registry. The Committee of German Stroke Registries predefined quality indicators (QI) of acute in-hospital stroke care. We compared 13 QI of these two groups taking into account national target values (available for 11 QI). We statistically compared both groups using percentages per hospital as unity and on patient data level.

12.605 and 5.296 patients were part of dataset of cSU and TSU, respectively. In 7 QI more than 90% of hospitals in both groups reached target values and in 3 QI more than 50%. In 1 QI (door-to-needle time ≤60 min in >90% of patients) value was only reached in less than 50% of hospitals of either group. In the remaining 2 QI without target values, significant differences were seen in rate of thrombolysis (of all ischemic stroke patients arriving ≤4h after onset) (mean 53% vs. 35%, p < 0.0001), favoring cSU and rate of pneumonia as complication (mean 5.1% vs. 3.1%, p = 0.019) favoring TSU.

There is high quality in acute stroke care in rural TSU despite low level onsite neurology expertise. However, our data revealed that rate of thrombolysis is lower than in cSU.

n/a

Scientific Communications 10- Intracerebral Haemorrhage

RISKS OF RECURRENT INTRACEREBRAL HAEMORRHAGE AND ISCHAEMIC STROKE AFTER SPONTANEOUS INTRACEREBRAL HAEMORRHAGE: A POPULATION-BASED COHORT STUDY

M. Poon¹, M. Rodrigues¹, C. Lerpiniere¹, L. Perry¹, K. McGoohan¹, J. Loan¹, R. al-Shahi Salman¹, N. Samarasekera¹

¹University of Edinburgh, Centre for Clinical Brain Sciences, Edinburgh, United Kingdom

The risks of recurrent intracerebral haemorrhage (ICH) and ischaemic stroke after ICH may differ by ICH location. We investigated the association between ICH location and risks of recurrent stroke.

We used a prospective population-based inception cohort in the Lothian region of Scotland comprising adults diagnosed with spontaneous ICH not known to be due to an underlying macrovascular, neoplastic or infectious cause between 1st June 2010- 31st May 2013. We performed Cox regression and Fine-Gray subdistribution hazard regression for cause-specific and competing risk analyses, respectively, adjusted for age, history of hypertension, and CT cerebral small vessel disease score.

248 adults with first-ever ICH survived at least 30 days. Adults with lobar ICH (n = 127) were more likely to be older and have dementia, and less likely to have prior hypertension and ischaemic stroke, than adults with non-lobar ICH (n = 121). During a median follow-up of 4 (IQR 1.1–5.0) years, the crude incidence rates of recurrent ICH and ischaemic stroke were 4.2 (95% CI 2.9–6.1) and 2.2 (95% CI 1.3–3.7) per 100 person-years, respectively. Lobar ICH location seemed to be associated with an increased risk of recurrent ICH (adjusted hazard ratio [aHR] 2.46, 95% CI 1.01–5.99, p = 0.047; and adjusted sub-hazard ratio [aSHR] 2.37, 95% CI 1.00–5.66, p = 0.051), but not risk of ischaemic stroke (aHR 1.17, 95% CI 0.39–3.47, p = 0.780; aSHR 0.97, 95% CI 0.35–2.68, p = 0.950).

Survivors of lobar ICH have a higher risk of recurrent ICH than survivors of non-lobar ICH, but their risks of ischaemic stroke are similar.

N/A

LONG-TERM SURVIVAL RATES, CAUSES OF DEATH AND TEMPORAL TRENDS IN FIVE-YEAR SURVIVAL RATES AFTER INTRACEREBRAL HEMORRHAGE. THE TROMSØ STUDY 1994–2016.

M. Carlsson¹, T. Wilsgaard², S.H. Johnsen¹, M.L. Løchen², I. Njølstad² and E.B. Mathiesen¹

¹UiT The Arctic University of Tromsø, Department of Clinical Medicine, Tromsø, Norway

²UiT The Arctic University of Tromsø, Department of Community Medicine, Tromsø, Norway

Data on long-term survival after intracerebral hemorrhage (ICH) are scarce. We compared long-term survival rates and causes of death within five years in 30 days survivors of first-ever ICH and the general population, and analysed temporal trends in five-year survival rates in ICH patients.

We included 187 participants of the population-based Tromsø Study who had a first-ever ICH between 1994 and 2011 and 935 participants without ICH, matched for birth year and sex. End of follow-up was Dec 31, 2016. Long-term survival rates were assessed by Kaplan Meier estimates. Hazard ratio (HR) for risk of death was estimated by Cox proportional hazards analysis. Comparison of causes of death was assessed by Fisheŕs exact test and trends in five-year survival by logistic regression analysis.

Cumulative five, 10, 15 and 20 year survival rates were 61%, 33%, 19% and 8% in 30 days survivors of ICH, and 69%, 50%, 32% and 21% in controls. The risk of death in ICH patients was significantly increased (HR 1.51, 95% Confidence Interval (CI) 1.22–1.86) compared to controls. Cardiovascular disease was the most common cause of death, and the proportion was significantly higher in ICH patients; 23.6% vs 9.6% (p < 0.001). Five-year survival rates did not change over time.

Long-term survival rates were significantly lower in ICH patients compared with controls. The most common cause of death within five years was cardiovascular disease, with a significantly higher proportion in ICH patients. Five-year survival rates in ICH patients remained stable throughout the observation period.

N/A

THREE YEAR STROKE RISK FOLLOWING SPONTANEOUS INTRACEREBRAL HAEMORRHAGE

G. Banerjee¹, D. Wilson¹, G. Ambler², I. Hostettler¹, C. Shakeshaft¹, H. Cohen³, T. Yousry⁴, R. Al-Shahi Salman⁵, G. Lip⁶, H. Houlden⁷, K. Muir⁸, M. Brown¹, H.R. Jäger⁴, D. Werring¹; on behalf of the CROMIS-2 collaborators

¹UCL Queen Square Institute of Neurology, Stroke Research Centre- Department of Brain Repair and Rehabilitation, London, United Kingdom

²University College London, Department of Statistical Science, London, United Kingdom

³University College London, Haemostasis Research Unit- Department of Haematology, London, United Kingdom

⁴UCL Queen Square Institute of Neurology, Lysholm Department of Neuroradiology and the Neuroradiological Academic Unit- Department of Brain Repair and Rehabilitation, London, United Kingdom

⁵University of Edinburgh, Centre for Clinical Brain Sciences, Edinburgh, United Kingdom

⁶University of Liverpool, Liverpool Centre for Cardiovascular Science, Liverpool, United Kingdom

⁷UCL Queen Square Institute of Neurology, Department of Molecular Neuroscience, London, United Kingdom

⁸University of Glasgow, Institute of Neuroscience & Psychology, Glasgow, United Kingdom

Intracerebral haemorrhage (ICH) is associated with high rates of mortality, and consequently data on stroke events in ICH survivors are limited. Our aim was to evaluate the association of ICH location on three-year stroke outcomes.

We included patients from the multicentre prospective observational CROMIS-2 ICH study. Outcome data were collected from patient and general practitioner questionnaires, and centrally collected hospital episode statistics. The outcomes of interest were cerebral ischaemic events (either stroke or TIA) or a further ICH following study entry.

There were 45 recurrent ICH events (absolute event rate, AER, 1.88 per 100 patient-years, 95% CI 1.41 to 2.52); 35 were in patients whose index event was lobar (n = 447, AER 3.77 per 100 patient-years, 95% CI 2.70 to 5.24), and 9 in patients presenting with non-lobar ICH (n = 580, AER 0.69 per 100 patient-years, 95% CI 0.38 to 1.41; Figure). In multivariable Cox regression, lobar ICH was associated with ICH recurrence (HR 8.70, 95% CI 3.29 to 23.03, p < 0.0001); similar results were found in competing risk analyses (Table 1). There were 70 cerebral ischaemic events (AER 2.93 per 100 patient-years, 95% CI 2.32 to 3.70); 29 in patients with lobar ICH (3.12 per 100 patient-years, 95% CI 2.17 to 4.49) and 39 in patients with non-lobar ICH (AER 2.97 per 100 patient-years, 95% CI 2.18 to 4.16). There was no association with ICH location; similar results were seen in competing risk analyses (Table 2).

Lobar ICH location is associated with recurrent ICH events, but not subsequent cerebral ischaemic events.

https://clinicaltrials.gov;NCT02513316

MIXED INTRACEREBRAL HAEMORRHAGE IN THE ETHNIC/RACIAL VARIATIONS OF INTRACEREBRAL HAEMORRHAGE (ERICH) STUDY: RISK FACTORS AND FUNCTIONAL OUTCOME

M. Pasi¹, S. Marini², A. Charidimou³, L. Xiong³, C. Kidwell⁴, A. Biffi³, J. Rosand², S.M. Greenberg³, M.E. Gurol³, D. Woo⁵, C.D. Anderson² and A. Viswanathan³

¹Centre Hospitalier Universitaire de Lille, Department of Neurology, Lille, France

²Center for Genomic Medicine- Massachusetts General Hospital, Department of Neurology, Boston, USA

³Hemorrhagic Stroke Research Program- Massachusetts General Hospital, Department of Neurology, Boston, USA

⁴University of Arizona, Departments of Neurology and Medical Imaging, Tucson, USA

⁵University of Cincinnati, Department of Neurology and Rehabilitation Medicine, Cincinnati, USA

We aimed: 1) To evaluate racial differences, Apolipoprotein E (APOE) epsilon allele frequency, vascular risk factors and end-organ damage of mixed-intracerebral haemorrhage patients ([ICH] and cerebral microbleeds [CMBs] in both lobar and deep areas) compared to strictly lobar-ICH (ICH and/or CMBs restricted to lobar areas) and strictly deep-ICH patients (ICH and/or CMBs restricted to deep areas); 2) To assess 90-day functional outcome of mixed-ICH patients.

Out of 3000 ICH patients recruited within the Ethnic/Racial Variations of ICH (ERICH) study, 934 patients who performed an MRI were used for analysis. Risk factors for mixed-ICH and 90-day modified Rankin scale (mRS>2) were evaluated in univariate and multivariate analyses.

Two-hundred and sixty-seven (28.6%) patients with a mixed-ICH pattern were compared to 264 (28.3%) strictly lobar-ICH, and 403 (43.1%) strictly deep-ICH patients. In comparison with strictly lobar-ICH and strictly deep-ICH, mixed-ICH patients were more likely to be black (vs lobar-ICH: odd ratio [OR] 1.6, 95% confidence interval [CI] 1.1–2.2; vs deep-ICH: OR 1.3, CI 1.01–1.7) to have hypertension (vs lobar-ICH: OR 1.6, CI 1.2–2.1; vs deep-ICH: OR 1.3, CI 1.04-1.6) and renal impairment (vs lobar-ICH: OR 1.8, CI 1.4–2.4; vs deep-ICH: OR 1.4, CI: 1.1–1.6). Mixed-ICH patients had a lower prevalence of APOE ε2 compared to strictly lobar-ICH patients (OR: 0.6, CI 0.5-0.9). Mixed-ICH had the worse functional outcome at 90 days after adjustment for relevant confounders (OR: 1.3, CI 1.1–1.5).

Mixed-ICH is more common in blacks and in individuals with hypertension-related vascular disease, and is associated with a poor functional outcome.

N/A

AN EARLY MORTALITY PROGNOSTIC SCORE IN INTRACEREBRAL HEMORRHAGE ESTABLISHED ON THE POPULATION-BASED BREST STROKE REGISTRY

F. Goupil¹

¹F. Goupil1 M.D- A. Dion2 M. Sci.- A. Jourdain1 M.D.- F.M. Merrien1 M.D.- F. Rouhart1 M.D.- I. Viakhireva-Dovganyuk1 M.D.- A. Leblanc1 M.D.- J.B. Noury1- P. Goas1 M.D.- and S. Timsit1- M.D.- Ph.D., 1. Neurology and stroke unit- CHRU- Brest- France – 2. Centre d’Investigation Clinique – INSERM CIC 0502- CHRU Brest- France, Brest, France

Intracerebral hemorrhages (ICH) account for 20% of strokes. The prognosis is severe, with early mortality of 40%. The aim was to establish a new 30-day prognostic mortality score.

Between January 1, 2008 and December 31, 2014, all non-traumatic and non-aneurysmal ICH were included on the population-based, prospective, multicenter, exhaustive stroke registry of Brest county (350,000 inhabitants). To allow internal validation of the score, the population was divided by randomization into a learning cohort (2/3) and a validation cohort (1/3). Univariate and multivariate odds ratios with 95% confidence interval were calculated by logistic regression. ROC curves for our score and Hemphill score were analyzed.

764 ICH were included (159 with vitamin K anticoagulant (VKA)). Median age was 79 years. 46.7% had deep and, 38.7% lobar ICH. Independent 30-day mortality risk factors, as well as the number of points allocated to each of them for the calculation of our score were: ICH with VKA (OR = 2.10, p = 0.0352), 1 point; age ≥ 80 years (OR = 3.21, p = 0.0002), 2 points; GCS < 5 (OR = 14.04, p < 0.0001), 4 points; 5 ≤ GCS ≤ 12 (OR = 4.82, p < 0.0001), 2 points; hematoma volume ≥ 30 ml (OR = 4.79, p < 0.0001), 2 points; no admission in intensive stroke unit (ISU)(OR = 2.28, p = 0.0381), 1 point; intra-ventricular hemorrhage (OR = 3.45, p < 0.0001), 2 points. On the validation cohort, the AUC of our score = 0.84 vs AUC of Hemphill’s score = 0.79.

We have developed a new 30-day prognostic mortality score for ICH, with high statistical performance. The use of these 2 independent poor prognostic factors: VKA, no admission in ISU may be useful for optimal care management.

N/A

MORTALITY AND FUNCTIONAL OUTCOME AFTER SURGICAL HAEMATOMA EVACUATION IN CEREBELLAR INTRACEREBRAL HAEMORRHAGE: AN INTERNATIONAL MULTICENTRE INDIVIDUAL PATIENT DATA META-ANALYSIS

J. Kuramatsu¹, A. Biffi², S.T. Gerner¹, J.A. Sembill¹, M.I. Sprügel¹, A. Leasure³, L. Sansing³, C. Matouk⁴, G.J. Falcone³, F.D. Testai⁵, D. Woo⁶, S. Schwab¹, K.N. Sheth³, H.B. Huttner¹; RETRACE Investigators

¹University Hospital Erlangen, Neurology, Erlangen, Germany

²Massachusetts General Hospital, Neurology, Boston, USA

³Yale School of Medicine & Yale New Haven Hospital, Neurology, New Haven, USA

⁴Yale School of Medicine & Yale New Haven Hospital, Neurosurgery, New Haven, USA

⁵University of Illinois at Chicago, Neurology, Chicago, USA

⁶University of Cincinnati, Neurology, Cincinnati, USA

The influence of surgical haematoma evacuation (SHE) on outcome in patients with cerebellar intracerebral haemorrhage (cICH) is unestablished. We aimed to determine the influence of SHE on functional outcome and mortality in cICH.

This multicentre study pooled individual-level data contributed by 64 hospitals across the USA and Germany (2006-2015). We addressed inter-study-variance, confounding, and bias using random-effect-models and propensity-score (1:1) matching to compare SHE versus conservative treatment. The primary outcome was the dichotomised modified Rankin-scale (mRS) 0–3 at three months. Secondary outcomes included mortality at three months, 12 months, and mRS 0–3 at 12 months. We further evaluated treatment related cut-off-values for ICH-volumes with outcomes.

We screened 6954 ICH-patients and identified 613 cICH. Patients treated with SHE(n = 184) compared to conservative (n = 429) showed lower GCS, larger ICH-volumes, and more frequent intraventricular haemorrhage (all, p < 0.001). After matching and adjustments, SHE was not significantly associated with the primary outcome (commonOR[cOR]:0.767,95% CI 0.341-1.726; p = 0.521). Patients treated with SHE showed reduced mortality rates at three (cOR:0.330,95% CI 0.115-0.946; p = 0.039) and 12 months (cOR:0.301,95% CI 0.132-0.685; p = 0.004), but functional outcome at 12 months was reduced (SHE:47/162(29.0%)versus64/162(39.5%); p = 0.047). We identified significant cut-off-values for SHE with ICH-volumes below 12 ml being associated with a reduced primary outcome (cOR:0.42,95% CI 0.19-0.93; p = 0.03) and with ICH-volumes between 15–30 ml (cOR: 0.23, 95% CI 0.10-0.52; p < 0.001) being associated with a reduced mortality.

SHE was not associated with improved functional outcome three months after cICH. Although mortality was reduced at three and 12 months, SHE was associated with larger proportions of patients being dependent. Based on our data we conclude that SHE should be avoided in patients with ICH-volumes < 12 ml, and should be considered for mortality reduction in patients with ICH-volumes above 15 ml.

NCT01829581, NCT03093233, NCT03183167;

LONG-TERM RISK AND TIME-COURSE OF RECURRENT INTRACEREBRAL HAEMORRHAGE AND ISCHAEMIC EVENTS FOLLOWING PRIMARY INTRACEREBRAL HAEMORRHAGE

L. Li¹, P. Lavallee¹, Y. Zuurbier¹, L. Silver¹, P.M. Rothwell; On behalf of the Oxford Vascular Study¹

¹Centre for Prevention of Stroke and Dementia, Nuffield Department of Clinical Neurosciences, Oxford, United Kingdom

Patients with primary intracerebral haemorrhage (PICH) are at risk of both recurrent ICH and ischaemic events. However, current risk estimates come predominantly from hospital-based studies performed in the 1990s. Moreover, there are uncertainties about the influence of ICH location on prognosis.

In a population-based study (OXVASC) of patients with a first PICH from 2002–2018 with face-to-face follow-up to 10 years, we determined the time-course and risks of recurrent ICH, recurrent ischaemic stroke or myocardial infarction (MI) stratified by location (lobar vs. deep) adjusting for age, and sex.

Of 255 patients with PICH (mean/SD age 75.5/13.1), 109 (42.7%) had lobar PICH, 144 (56.5%) deep, and 2 (0.8%) had fatal PICH out of the area with uncertain haematoma location. During 687 patient-years of follow-up, there were 15 recurrent PICH (annual rate 2.4%, 95% CI 1.3%-4.0%) and 9 ischaemic stroke/MI (1.4%, 0.7%-2.7%). Patients with lobar PICH had significantly higher 10-year risks of recurrent PICH than those with deep PICH (n/% lobar vs. deep – 11/32.5% vs. 4/13.3%, adjusted HR = 3.42, 95% CI 1.08-10.82, p = 0.04), particularly in the first year (adjusted HR = 4.14, 0.84-20.56, p = 0.08). In patients with lobar PICH, this 10-year risk of recurrent PICH exceeded the risk of recurrent ischaemic stroke/MI (11/32.5% vs. 4/15.9%), whereas the risks were more equivalent after deep PICH (4/13.3% vs. 5/14.0%).

The particularly high early risk of recurrent ICH after lobar PICH highlights the need for more effective treatments for this group.

N/A

HEMOSTATIC EFFICACY AND ANTI-FXA REVERSAL WITH ANDEXANET ALFA IN SPONTANEOUS INTRACRANIAL HEMORRHAGE

A. Demchuk¹, E. Zotova², J. Nakamya², P. Yue³, T. Milling Jr. ⁴, J. Curnutte³, J. Eikelboom², M. Crowther², S. Connolly²; ANNEXA-4 Steering Committee and Investigators

¹University of Calgary, Calgary Stroke Program- Department of Clinical Neurosciences- Hotchkiss Brain Institute, Calgary, Canada

²Population Health Research Institute, Population Health Research Institute, Hamilton, Canada

³Portola Pharmaceuticals, Portola Pharmaceuticals, South San Francisco, USA

⁴Seton/UT Southwestern Clinical Research Institute, Neurology, Austin, USA

Andexanet alfa is a recombinant modified human Factor Xa (FXa) decoy protein developed to reverse FXa inhibition from anticoagulants such as apixaban or rivaroxaban.

Patients with nontraumatic intracranial bleeding (ICrH) enrolled in ANNEXA-4 (clinicaltrials.gov NCT02329327) were analyzed if within 18h of FXa inhibitor use, and received andexanet bolus and 2-hour infusion. Baseline; 2h; and 12h post andexanet bolus brain imaging performed. Hemorrhage thickness and volumetric analysis (Quantomo) performed. Co-primary outcomes were change in anti-factor Xa activity (anti-fXa) and rate of good/excellent hemostatic efficacy at 12 hours (good (20-35% increase) excellent (<20% increase in hemorrhage thickness/volume)).

Nontraumatic ICrH present in 99 enrolled patients. Primary intracranial compartment of bleeding was ICH and/or IVH in 83 pts; SDH in 13 and SAH in 3. Mean age 77.5 yrs; median onset to andexanet bolus 4.7h (n = 39); hospital arrival to bolus 3.2 h (n = 99); baseline brain imaging to bolus 1.8 h (n = 99). Mean total (ICH+IVH) volume 15.1+/7−15.3 mls. After andexanet bolus/influsion apixaban patients (n = 50), median anti-fXa decreased from baseline 139.3 ng/ml to nadir 9.55 ng/ml; and for rivaroxaban patients (n = 40) from baseline 218.55 ng/ml to nadir 10.15 ng/ml. Good or excellent hemostasis occurred in 77 (78.6%) 12 hour post-bolus overall. Earlier andexanet bolus (less than < 1.725h from baseline imaging) hemostatic efficacy was 83.7% versus 73.5% if andexanat bolus administered greater than 1.725h after baseline imaging.

Andexanet substantially reduced anti-fXa levels in FXa inhibitor associated nontraumatic intracranial hemorrhage and was associated with a high rate of hemostatic efficacy at 12 hours after bolus.

clinicaltrials.gov: NCT02329327

Scientific Communications 11-Rare Causes and Stroke in the Young

GEOGRAPHIC DIFFERENCES IN THE ETIOLOGY OF STROKE AT YOUNG AGE: THE GLOBAL OUTCOME AFTER STROKE AT YOUNG AGE (GOAL)-INITIATIVE

M. Ekker¹, M. Dongen van¹, M. Jacob¹, K. Aarnio², A. Annamalai³, A. Arauz⁴, M. Arnold⁵, M. Barboza⁶, M. Bolognese⁷, R. Brouns⁸, B. Chuluun⁹, E. Chuluunbaatar¹⁰, B. Dagvajantsan⁹, S. Debette¹¹, A. Don¹², C. Enzinger¹³, E. Ekizoglu¹⁴, S. Fandler¹⁵, F. Fazekas¹⁵, A. Fromm¹⁶, T. Gattringer¹⁵, G. Gulli¹⁷, M. Hoffmann¹⁸, T. Hora¹⁹, C. Jern²⁰, K. Jood²¹, M. Kamouchi²², Y. Kim²³, T. Kitazono²⁴, S. Kittner²⁵, T. Kleinig²⁶, K. Klijn¹, J. Körv²⁷, T. Lee²⁸, D. Leys²⁹, N. Maaijwee⁷, N. Martinez-Majander², J. Marto³⁰, M. Mehndiratta³¹, V. Mifsud³², V. Montanaro¹⁹, M. Owolabi³³, V. Patel³⁴, M. Phillips³⁵, B. Piechowski-Iozwiak³², A. Pikula³⁶, J. Ruíz-Sandoval³⁷, B. von Sarnowski³⁸, F. Schreuder¹, R.H. Swartz⁴⁰, K.S. Tan⁴¹, D. Tanne⁴², T. Tatlisumak², V. Thijs²⁶, A.M. Tuladhar¹, M. Viana-Baptista³⁰, R. Vibo²⁷, T.Y. Wu⁴³, N. Yeşilot¹⁴, U. Waje-Andreassen¹⁶, A. Pezzini⁴⁴, J. Putaala² and F.E. de Leeuw¹

¹Radboud University Medical Centre- Donders Institute for Brain- Cognition and Behaviour, Department of Neurology, Nijmegen, The Netherlands

²Helsinki University Hospital, Department of Neurology, Helsinki, Finland

³North Tees and Hartlepool, NHS Foundation trust, Stockton on Tees, United Kingdom

⁴National Institute of Neurology and Neurosurgery of Mexico, Manuel Velasco Suárez, Mexico city, Mexico

⁵Inselspital University Hospital- University of Bern, Department of Neurology, Bern, Switzerland

⁶Hospital Dr. Rafael A. Calderon Guardia, Neurosciences Department, San Jose, Costa Rica

⁷Lucerne Cantonal Hospital, Neurocenter- Department of Neurology and Neurorehabilitation, Luzern, Switzerland

⁸Vrije Universiteit Brussel, The Faculty of Medicine and Pharmacy, Jette, Belgium

⁹Mongolian National Univeristy of Medical Sciences, International School of Traditional Medicine, Uluunbaatar, Mongolia

¹⁰Taipei Medical University, Depatrment of Global Health and Development, Taipei, Taiwan R.O.C

¹¹Inserm, Bordeaux Population Health Research Centre, Bordeaux, France

¹²Tel Aviv University, School of Medicine- Sackler Faculty of Medicine, Tel Aviv, Israel

¹³Medical University of Graz, Department of Neurology- Department of Radiology- division of neuroradiology, Graz, Austria

¹⁴Istanbul University- Faculty of Medicine, Department of Neurology, Istanbul, Turkey

¹⁵Medical University of Graz, Department of Neurology, Graz, Austria

¹⁶Haukeland University Hospital, Centre for Neurovascular Diseases- Department of Neurology, Bergen, Norway

¹⁷Ashford and St Peter’s Hospitals, Department of Stroke Medicine, Chertsey, United Kingdom

¹⁸University of Central Florida, Stroke Center- Orlando VA Medical Center, Orlando, USA

¹⁹SARAH Hospital of Rehabilitation, Department of Neurology, Brasilia, Brazil

²⁰University of Gothenburg, Insitute of Biomedicine- the Sahlgrenska Academy, Gothenburg, Sweden

²¹University of Gothenburg, Institute of Neuroscience and physiology- the Sahlgrenska Academy, Gothenburg, Sweden

²²Kyushu University, Department of Health Care Administration and Management- Center for Cohort Studies- Graduate School of Medical Sciences, Fukuola, Japan

²³Hanyang University, Department of Neurology- College of Medicine, Seoul, Republic of Korea

²⁴Kyushu University, Department of Medicine and Clinical Science- Graduate School of Medical Sciences- Center for Cohort Studies, Fukuola, Japan

²⁵University of Maryland School of Medicine, Department of Neurology- Veterans Affairs Maryland Health Care System, Baltimore, USA

²⁶Royal Adelaide Hospital, Department of Neurology, Adelaide, Australia

²⁷University of Tartu, Department of Neurology and Neurosurgery, Tartu, Estonia

²⁸Chang Gung Memorial Hospital- Linkou Medical Center- and College of Medicine- Chang Gung University, Department of Neurology, Taoyuan, Taiwan R.O.C

²⁹University of Lille, Department of Neurology, Lille, France

³⁰Universidade Nova de Lisboa, Department of Neurology- Hospital Egas Moniz- Centro Hospitalar Lisboa Ocidental and CEDOC- Nova Medical School, Lisbon, Portugal

³¹G.B. Pant Hospital, Department of Neurology, New Delhi, India

³²Cleveland Clinic Abu Dhabi, Neurological Institute, Abu Dhabi, United Arab Emirates

³³University of Ibadan, Department of Medicine- College of Medicine, Ibadan, Nigeria

³⁴Inkosi Albert Luthuli Central Hospital, Department of Neurology- Nelson R. Mandela School of Medicine, Mayville, South Africa

³⁵Waikato Hospital, Department of Neurology, Hamilton, New Zealand

³⁶University of Toronto, Department of Medicine Neurology- University Health Network, Toronto, Canada

³⁷Hospital Civil de Guadalajara Fray Antonio Alcalde, Department of Neurology, Guadalajara, Mexico

³⁸University Medicine, Department of Neurology, Greifswald, Germany

³⁹Hospital Civil de Guadalajara Fray Antonio Alcalde, Department of Neurology, Guadalajara, Mexico

⁴⁰Sunnybrook Health Sciences Centre, Department of Medicine Neurology, Toronto, Canada

⁴¹University of Malaya, Department of Medicine, Kuala Lumpur, Malaysia

⁴²Tel Aviv University, Department of Epidemiology and Preventive Medicine, School of Public Health, Sackler Faculty of Medicine, Tel-Hashomer, Israel

⁴³Christchurch Hospital, Department of Neurology, Christchurch, New-Zealand

⁴⁴University of Brescia, Department of Clinical and Experimental Sciences, Neurology Clinic, Brescia, Italy

Two million patients suffer a stroke at age 18 to 50 years each year, with huge geographical variation in etiology. Etiologic diagnosis is essential in the assessment of accurate treatment and prognosis.

Patients were included from the GOAL study, a worldwide multicenter individual patient meta-analysis based on pre-existing cohorts, of patients aged 18–50 years old with first-ever ischemic stroke or intracerebral hemorrhage. Ischemic stroke cause was defined by TOAST criteria and stratified for age, gender, ethnicity and geographic location. Furthermore, we compared TOAST classification with patients classified based on the determined stroke mechanism by the original investigators.

7700 ischemic stroke patients (mean age of 38.8 years) were available (Table 1). Most patients had undetermined etiology (35%), followed by ‘other determined etiology’ (22.6%), cardio embolism (CE) (21.9%), small vessel disease (SVD) (11%), and large vessel atherosclerosis (LVA) (9.6%). Black-African people more often had LVA, (29.9%, p < 0.001), and Caucasian more often had a CE cause (23.5%, p < 0.001) (Table 1). When comparing TOAST-classification with patients that had an exact classification of their stroke (Table 2), less patients were diagnosed with undetermined stroke (24.7% vs. 35%) and more with ‘other determined’ category (then 29.6% vs. 22.6%).

Considerable differences in etiology in young stroke patients are seen regarding age, gender, ethnicity and geographic location, providing further opportunities to find genetic and ethnic variation and personalize secondary preventive strategies. Break up of exact causes potentially leads to smaller proportion of strokes classified as undetermined, however causality of many of such conditions remain poorly proven.

N/A

MALIGNANCY IS THE MOST IMPORTANT CAUSE OF DEATH FOLLOWING FIRST-EVER STROKE IN YOUNG ADULTS

J. Verhoeven¹, M. Ekker¹, I. Vaartjes², C. Klijn¹ and F.E. de Leeuw¹

¹Radboud University Medical Center- Donders Institute for Brain- Cognition and Behaviour, Department of Neurology, Nijmegen, The Netherlands

²University Medical Center Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands

Long-term mortality in young stroke patients is high. It is important to better understand its causes to possibly unravel unknown causes of strokes and to provide more accurate information on its prognosis. We investigated causes of death in the young stroke population in the Netherlands.

We linked national registries (Hospital Discharge-, Cause of Death- and Population Register) using ICD-9/-10 codes for ischemic stroke, intracerebral hemorrhage and unspecified stroke. We included patients aged 18–49 years old with first-ever stroke between 1998–2010, with follow-up until 2017. Outcomes were causes of death categorized in stroke-related, cardiac or other vascular events, malignancies, infection, trauma and miscellaneous, stratified for stroke-subtypes and sex.

We identified 13481 patients (mean age 41.76 ± 6.8 years, 53.3% women); 7815 (58.0%) ischemic strokes, 2086 (15.4%) intracerebral hemorrhages and 3580 (26.6%) unspecified strokes. During a median follow-up period of 10.19 (IQR 7.1-13.6) years 1763 (13.1%) patients died. 32.7% of malignancies, 17.1% cardiac or other vascular events, 15.1% stroke-related, 4.9% infection, 4.4% trauma and 25.6% miscellaneous. Intracerebral hemorrhage patients were significantly more likely to die from stroke (χ2 = 15.4; p = 0.0001) or malignancies (χ2 = 18.9; p < 0.0001) than ischemic stroke patients. Men were more likely to die of cardiac or other vascular events (χ2 = 10.1; p = 0.0015) and less likely of malignancies (χ2 = 26.2; p < 0.0001) than women.

The major cause of death after young stroke is malignancy, which could implicate a common underlying disease-mechanism. Future research should investigate possible benefits of screening for malignancies in young stroke patients. Additionally, we demonstrate that young adults remain at increased risk of cardiovascular death.

N/A

FACTORS ASSOCIATED WITH POOR TREATMENT OUTCOMES OF STROKE IN YOUNG ADULTS IN THAILAND

K. Kongbunkiat¹, N. Kasemsap² and S. Tiamkao²

¹North-Eastern Stroke Research Group- Khon Kaen University, Medicine, Khon Kaen, Thailand

²Khon Kaen University, Medicine, Khon Kaen, Thailand

Stroke in young adults is one of major problem in Thailand. This study describes patient characteristics and factors associated with poor treatment outcomes.

Descriptive retrospective study of data from the National Health Security Office (NHSO), Thailand during 2004–2013. Risk factors, type, comorbidities and complication that associated with poor treatment outcomes in stroke patients (15-45 years) were analyzed.

The total 47,728 of stroke in young adults compare to 444,992 patients (10.73%) of all adult stroke above 15 years was done. The mean age was 37.8 years, male gender 63.5%, hemorrhagic stroke 50.9%, ischemic stroke 38.3%, unclassified 10.8%. Related risk factors were hypertension 27.6%, coagulation abnormalities 3.3%, atrial fibrillation 3%, drug abuse 4.8%. Good outcomes group were complete recovery 0.38% and improved 69.30%. Poor outcomes group were not improved 15.42% and dead 14.90%. Male gender and increasing age associated with poor outcomes. Comorbidities associated with poor outcomes were coagulation abnormalities (AOR 3.08, 95% CI 2.78, 3.40), chronic kidney disease (AOR 1.75,95% CI 1.40, 2.18), neoplasms (AOR 1.48, 95% CI 1.06, 2.08). Complications associated with poor outcomes were septicemia (AOR 10.74, 95% CI 8.25, 13.99), herniation of brain (AOR 6.43, 95% CI 5.01, 8.25), gastrointestinal bleeding (AOR 3.72, 95% CI 2.67, 5.19), intracerebral hemorrhage (AOR 3.59, 95% CI 3.11, 4.15).

Factors associated with poor treatment outcomes were male gender, increasing age, coagulation abnormalities, chronic kidney disease, neoplasms, septicemia, herniation of brain, gastrointestinal bleeding, intracerebral hemorrhage. Physicians in Thailand should be aware of these predictors for poor treatment outcomes.

N/A

J. Ruuskanen¹, V. Kytö², J. Posti¹, P. Rautava³ and J. Sipilä¹

¹Turku University Hospital, Clinical neurosciences, Turku, Finland

²University of Turku, Research Centre of Applied and Preventive Cardiovascular Medicine, Turku, Finland

³Turku University Hospital, Clinical Research Center, Turku, Finland

Cerebral venous thrombosis (CVT) is a rare cause of stroke with an annual incidence of approximately 1.3/100 000 in Western Europe. Changes in incidence and mortality have been suggested, although there are few comprehensive population-based studies. Therefore, we studied all cases of CVTs in Finland between the years 2005 and 2014.

All CVT related admissions and in-hospital deaths were retrieved from a mandatory registry covering mainland Finland. All patients aged ≥18 years with acute CVT with or without infarction as discharge diagnosis were included. One admission per patient per year was allowed.

A total of 563 patients with 727 admissions were found. Majority of patients were women (n = 318, 56.5%) and they were significantly younger (median 38.0 years, IQR 32 years) than male patients (54.0, IQR 23) (p < 0.001). Yearly incidence for the study period was 1.67/100 000 (95% CI 1.51-1.84) in women and 1.33/100 000 (95% CI 1.19-1.49) in men. In Poisson regression, a 4.8% increase per year in admissions was seen. Most common comorbidity was malignancy in 7.2% of the patients. In-hospital mortality was 1.7% with no sex difference. Patients who died were older than survivors (median age 65.5 vs 47.0 years, p = 0.018).

In this nationwide, population-based study, incidence of CVT was higher than previously reported in western Europe, with an increase in number of admissions over the study period while in-hospital mortality remained low. This may be due to improving diagnostics and hospitalizations of milder forms of the disease.

N/A

CRP AND CLAUDIN-5 ARE ASSOCIATED WITH VENOUS INFARCTION IN CEREBRAL VENOUS THROMBOSIS

J. Duan¹, Z. Han², Y. Cai³, C. Wang⁴, G. Rajah⁵, H. Song², Y. Ding⁵ and X. Ji²

¹Xuanwu Hospital-Capital Medical University, Neurology, B, China

²Xuanwu Hospital-Capital Medical University, Neurology, Beijing, China

³Xuanwu Hospital-Capital Medical University, Department of Neurobiology, Beijing, China

⁴Xuanwu Hospital-Capital Medical University, Evidence-based Medicine, Beijing, China

⁵Wayne State University School of Medicine, Neurosurgery, Detroit, USA

To explore whether blood brain barrier (BBB) permeability-related proteins and high-sensitivity C-reactive protein (hs-CRP) reflecting inflammation are associated with venous infarction in CVT patients.

A total of 59 patients were divided into four groups: acute/subacute CVT patients with venous infarct (ASVI group, n = 30), acute/subacute CVT patients without venous infarct (ASOVI group, n = 13), chronic CVT group (n = 9) and control subjects (n = 7). Hs-CRP, claudin-5, occludin, matrix metalloproteinase-9, glial fibrillary acidic protein, and S100B were measured in serum and/or cerebrospinal fluid upon admission. CVT severity on admission was assessed using the National Institutes of Health Stroke Scale (NIHSS) and modified Rankin Scale (mRS).

Only serum hs-CRP and claudin-5 were higher in the ASVI group than in the ASOVI group (hs-CRP, 19.13 ± 15.22 vs. 5.65 ± 6.19 mg/l, p = 0.001; claudin-5, 3.15 ± 1.20 vs. 2.41 ± 0.57 mg/l, p = 0.025). Areas under the curves (AUCs) of serum claudin-5, hs-CRP and both biomarkers combined discriminating venous infarct were respectively 0.700, 5% CI 0.54-0.86, p = 0.027; 0.795, 95% CI 0.66-0.93, p = 0.003; and 0.862, 95% CI 0.75-0.97, p < 0.001. Both proteins were independently associated with venous infarct after adjusting covariates (serum hs-CRP, odds ratio [OR] 49.35, 95% CI 2.87-847.14, p = 0.007; claudin-5, OR 39.03, 95% CI 1.74-876.44, p = 0.021). In 43 acute/subacute CVT patients, only serum hs-CRP was positively associated with NIHSS (r = 0.710, 95% CI 0.533-0.824, p < 0.001) and mRS (r = 0.689, 95% CI 0.493-0.830, p < 0.001) on admission.

Serum hs-CRP and claudin-5 may be clinically valuable biomarkers associated with venous infarction in acute/subacute CVT patients. Inflammation may play an important role in the occurrence of venous infarction and CVT severity.

N/A

ISCHEMIC MANIFESTATIONS OF REVERSIBLE CEREBRAL VASOCONSTRICTION SYNDROME; FREQUENCY, FEATURES AND RISK FACTORS.

O. Osman¹, J. Mawet², J. Franc³, H. Chabriat¹, A. Ducros⁴ and P. Reiner¹

¹Lariboisiere Hospital- Assistance Publique des Hô pitaux de Paris, Neurology, Paris, France

²Lariboisiere Hospital- Assistance Publique des Hô pitaux de Paris, Emergency Headache Center, Paris, France

³Clinique Turin, Neuroradiology, Paris, France

⁴CHU de Montpellier, Neurology, Montpellier, France

Stroke represent the major complication of reversible cerebral vasoconstriction syndrome (RCVS). Haemorrhages have been frequently reported but ischemic lesions may also complicate the RCVS course and are less known. The aim of this study was to determine the frequency, patterns and risk factors of ischemic lesions in RCVS.

We prospectively analysed data on 271 patients with confirmed RCVS. Ischemic lesions were confirmed by a neuroradiologist. Standard bivariate statistical tests were applied to compare patients with and without ischemic lesions.

Among 271 patients, 14 patients (5%) developed ischemic lesions. One occured during postpartum and 4 patients used vasoactive drugs. These patients presented more frequently with a single thunderclap headache (57% vs 15%, p = 0.0005), associated haemorrhagic manifestations (71% vs 24%, p = 0.0003) and cervical artery dissections (29% vs 9%, p = 0.04). Eight patients had multifocal lesions and infarcts were located in watershed (7), deep (4) and territorial (6) areas. Four patients with isolated ischemic lesions were treated with antiplatelet without haemorrhagic complication. Among patients with ischemic and haemorrhagic lesions, the absence of antiplatelet use did not result in the appearance of new ischemic lesions.

In RCVS, ischemic lesions are rare and generally associated with a good outcome. However, frequent coexistence of ischemic lesions with haemorrhagic complications suggests that infarcts might be a marker of the severity of the disease. Patients present more frequently with single thunderclap headache, suggesting that RCVS clinical spectrum could be larger than usually thought. Pathophysiological mechanisms of ischemic lesions and the benefit of antithrombotic treatment in RCVS remain unknown.

N/A

PREDICTORS OF CLINICAL OR CEREBRAL LESION PROGRESSION IN MOYAMOYA ANGIOPATHY AMONG ADULT PATIENTS: A PROSPECTIVE COHORT STUDY

D. Herve¹, N. Ibos-Augé², L. Calviere³, C. Rogan⁴, M.A. Labeyrie⁵, J.P. Guichard⁶, O. Godin², M. Kossorotoff⁷, M.O. Habert⁸, E. Tournier Lasserve⁹, S. Chevret¹⁰ and H. Chabriat¹¹

¹CERVCO- Hopital Lariboisiere- APHP/INSERM UMR 1141, Neurology, Paris, France

²Hopital Lariboisiere- APHP, Neurology, Paris, France

³Unité neurovasculaire- hô pital Pierre-Paul-Riquet- Toulouse- France, Neurology, Paris, France

⁴CERVCO- Hopital Lariboisiere- APHP, Neurology, Paris, France

⁵CERVCO- Hopital Lariboisiere- APHP, Neuroradiology, Paris, France

⁶Hopital Lariboisiere- APHP, Neuroradiology, Paris, France

⁷Centre national de référence de l’AVC de l’enfant- Hô pital universitaire Necker-Enfants malades- AP–HP, Neuropediatry, Paris, France

⁸Hopital Salpetrière- APHP- Paris- France, Nuclear Medicine, Paris, France

⁹CERVCO- Hopital Lariboisiere- APHP/INSERM UMR 1141/Université Paris Diderot, Laboratoire de génétique moléculaire, Paris, France

¹⁰Hô pital Saint Louis- APHP- Paris- France, Service de Biostatistique et Information médicale-, Paris, France

¹¹CERVCO- Hopital Lariboisiere- APHP/INSERM UMR 1141/Université Paris Diderot, Neurology, Paris, France

Little is known about the natural history of moyamoya angiopathy (MMA) in adult patients. Prospective studies are needed to define predictive models that can be used to support treatment decisions. We sought to identify independent predictors of clinical or cerebral lesion progression in a large sample of affected individuals prior to decisions regarding revascularization surgery.

Ninety non-operated subjects (median age, 37.5 years) were assessed at baseline and followed for a median time of 42.8 months. Incident strokes, death, as well as incident ischemic and hemorrhagic lesions on MRI were recorded. Multiple demographic, clinical, and cerebral imaging parameters at baseline were considered as potential predictors of clinical and/or cerebral tissue change at follow up. Data were analyzed based on the Andersen-Gill counting process model, followed by internal validation of the prediction model.

Asian origin, a previous history of transient ischemic attacks and a reduction in hemodynamic reserve were found to be associated with an increased risk of clinical and imaging events. While the model estimated the risk of clinical or cerebral lesion progression to be approximately 0.5% per year when none of these factors were present, this risk exceeded 20% per year when all factors were present.

A simple combination of demographic, clinical, and cerebral perfusion imaging parameters may aid in predicting the risk of incident stroke and cerebral lesion progression in adult patients with MMA. These results may help to improve therapeutic decisions and aid in the design of future trials in adults with this rare condition.

N/A

RISK OF STROKE IN A HIGH-RISK POPULATION OF SICKLE CELL DISEASE PATIENTS: RESULTS FROM A FRENCH MULTICENTRIC COHORT (PCDREP)

B. Maier¹, D. Calvet², M. Edjlali-Goujon³, M.P. Gobin-Metteil³, A. Habibi⁴, F. Galacteros⁴ and P. Bartolucci⁴

¹Rothschild Fondation, Interventional Neuroradiology, Paris, France

²Sainte-Anne Hospital- Paris Descartes University- INSERM U1266, Neurology, Paris, France

³Sainte-Anne Hospital, Diagnostic Neuroradiology, Paris, France

⁴Mondor Hospital, Sickle Cell Referral Center- Service de Médecine Interne- UPEC- INSERM U895- Team 2- Laboratoire d’Excellence GRex- Créteil- France, Créteil, France

Sickle cell disease (SCD) prognosis is mainly worsened by the occurrence of strokes, often due to a disease-related vasculopathy. Therefore we assessed the risk of stroke and its predictors in a high-risk population of SCD adults.

From January 2011 to June 2018, 136 consecutive patients were enrolled in the PCDREP cohort, which includes SCD adult patients with history of neurological event (stroke, transient ischemic attack, and seizures) or suspected SCD-related vasculopathy. All patients had a detailed baseline work-up assessment including EchoDoppler ultrasound and MRI with 3D-time-of-flight angiography. Treatments were adapted in multidisciplinary meetings. Patients were followed in referral centers at least twice a year and all cerebrovascular events were prospectively recorded.

Among the 136 patients, mean age was 30.4 ( ± 16.4), genotype was SS in 89%, SC in 7%, Sβ0-thalassemia in 2% and Sβ+-thalassemia in 2%. At inclusion, 48% had a history of stroke, 58% had a vasculopathy and 50% were treated with blood exchange transfusion. During a mean follow-up of 48 months, twelve patients had a stroke (6 ischemic, 6 hemorrhagic), 2 being fatal. The absolute risk of stroke was 6.8% (95% CI = 2.3–11.3) and 9.6% (95% CI = 4.1–15.1) respectively at 1 and 2 years. A history of stroke was independently associated with the occurrence of stroke during follow-up (adjusted HR = 9.9; 95% CI = 1.27–77.1) but not the presence of vasculopathy.

The risk of stroke remains significant in this high-risk population of SCD adults, especially in those with a stroke history, despite a highly specialized management. Assessment of alternative strategy in adult SCD patients is needed.

NA

RECANALIZATION AFTER CEREBRAL VENOUS THROMBOSIS (CVT). A RANDOMIZED CONTROLLED TRIAL OF THE SAFETY AND EFFICACY OF DABIGATRAN ETEXILATE VERSUS DOSE-ADJUSTED WARFARIN IN PATIENTS WITH CVT

J. Férro¹, M. Bendszus², O. Jansen³, J. Coutinho M⁴, F. Dentali⁵, A. Kobayashi⁶, D. Aguiar de Sousa⁷, L. Lucas Neto⁸, J. Caria⁹, M. Fräßdorf¹⁰, H. Huisman¹¹, P. Reilly¹², H.C. Diener¹³; RE-SPECT CVT study group

¹Hospital Santa Maria/Centro Hospitalar Lisboa Norte- Instituto de Medicina Molecular and Faculdade de Medicina- Universidade de Lisboa, Department of Neurosciences and Mental Health- Serviço de Neurologia, Lisbon, Portugal

²Heidelberg University Hospital, Department of Neuroradiology, Heidelberg, Germany

³University Hospital Schleswig-Holstein- Campus Kiel, Department for Radiology and Neuroradiology, Kiel, Germany

⁴Academic Medical Centre, Department of Neurology, Amsterdam, The Netherlands

⁵Insubria University, Department of Clinical and Experimental Medicine, Varese, Italy

⁶Faculty of Health Sciences and Physical Education- Kazimierz Pulaski University of Technology and Humanities- Random, Institute of Psychiatry and Neurology, Warsaw, Poland

⁷Hospital Santa Maria/Centro Hospitalar Lisboa Norte- Instituto de Medicina Molecular and Faculdade de Medicina/Universidade de Lisboa, Department of Neurosciences and Mental Health- Serviço de Neurologia, Lisbon, Portugal

⁸Hospital Santa Maria/Centro Hospitalar Lisboa Norte- Faculdade de Medicina/Universidade de Lisboa, Serviço de Neurorradiologia, Lisbon, Portugal

⁹Boehringer Ingelheim International, Boehringer Ingelheim International, Ingelheim am Rhein, Germany

¹⁰Boehringer Ingelheim GmbH & Co. KG, Boehringer Ingelheim GmbH & Co. KG, Ingelheim am Rhein, Germany

¹¹Boehringer Ingelheim B.V, Boehringer Ingelheim B.V, Alkmaar, The Netherlands

¹²Boehringer Ingelheim Pharmaceuticals- Inc, Department of Clinical Development, Ridgefield- CT, USA

¹³University Duisburg-Essen, Faculty of Medicine, Essen, Germany

The effect of anticoagulants on recanalization after cerebral venous thrombosis (CVT) has not been studied in a randomised controlled trial.

RE-SPECT CVT was a phase III, prospective, randomized, parallel-group, open-label, multicentre, exploratory trial with blinded endpoint adjudication. After 5 to 15 days of heparin, CVT patients were allocated to dabigatran 150 mg twice daily or dose-adjusted warfarin, for 24 weeks. A standardized MR protocol including arterial spin labelling, 3D ToF venography and 3D contrast enhanced MRA was obtained at the end of the treatment period. Cerebral venous recanalization was a secondary efficacy endpoint, assessed by two blinded adjudicators, using the difference in a score of occluded sinus and veins and in the modified Qureshi scale, between baseline and end of treatment.

120 CVT patients were randomised. Cerebral venous recanalization was evaluated in 107 patients. No patient worsened in the score of occluded cerebral veins and sinuses, while 33 (60%) on dabigatran and 35 (67%) on warfarin improved. There were no differences between treatment groups in the mean score change (dabigatran -0.8, SD 0.78; warfarin -1.0, SD 0.92), and in the modified Qureshi score. Full recanalization was adjudicated in 24 (44%), partial recanalization in 23 (42%) patients in the dabigatran arm; in the warfarin arm, full recanalization was adjudicated in 19 (36%), partial recanalization in 26 (49%) patients (p = 0.44).

The majority of patients with CVT anticoagulated with either dabigatran or warfarin for 6 months showed partial or complete recanalization of occluded sinus and veins at the end of treatment.

Clinicaltrials.gov: NCT02913326

Scientific Communications 12-Cerebral Small Vessel Disease

PRESERVE TRIAL: INTENSIVE VERSUS STANDARD BLOOD PRESSURE LOWERING IN CEREBRAL SMALL VESSEL DISEASE ASSESSED BY DTI

H. Markus¹, M. Egle¹, I. Croall¹, D. Tozer¹, A. Blamire², J. O’Brien³, T. Barrick⁴, R. Morris⁵, G. Ford⁶; PRESERVE Trial Investigators

¹University of Cambridge, Clinical Neurosciences, Cambridge, United Kingdom

²Newcastle Magnetic Resonance Centre, University of Newcastle, Newcastle, United Kingdom

³University of Cambridge, Psychiatry, Cambridge, United Kingdom

⁴St George’s University of London, Clinical Neuroscience, London, United Kingdom

⁵Kings College London, Neuropsychology, London, United Kingdom

⁶University of Oxford, Medicine, Oxford, United Kingdom

Hypertension is the major risk factor for cerebral small vessel disease (SVD). Treating BP in midlife prevents SVD, but in severe SVD, when cerebral blood flow and autoregulation are impaired, excessive BP reduction could accelerate white matter damage and worsen cognition. Whether its is safe to intensively treat BP can only be resolved by RCTs. RCTs in SVD using cognition as an endpoint have been inconclusive due to the insensitivity of cognitive testing to change over periods of 1–3 years. Diffusion tensor imaging (DTI), is sensitive to change in SVD over a 1–3 year periods, and has been suggested as a surrogate marker for SVD RCTs. But this hypothesis has not been tested. In PRESERVE intensive versus standard BP lowering regimens were compared in an RCT with DTI as the primary outcome measure.

PRESERVE randomised patients with severe SVD (lacunar stroke and confluent WMH) between intensive (systolic BP target < 125 mmHg) and standard (target 130–140) BP management for 48 months. Primary endpoint was change in white matter damage on DTI between baseline and 24 months. Cognition was also assessed. 111 patients from six UK centres were recruited (intensive 55, standard 56).

Target BP difference was achieved by 3 months and maintained for 2 years; Systolic BP; mean (SD). Standard: baseline 150.0(12.9), 3 months 140.2(13.4), 12 months 138.0(12.6), 24 months 135.1(11.5) mmHg. Intensive: Baseline 149.5(13.4), 3 months 126.6(10.7), 12 months 126.7(13.3), 24 months 126.3(15.2) mmHg.

First results from the primary endpoint analysis will be presented

Results will be discussed.

CRN Number: 10962

DIFFERENT LOCAL ALTERATIONS TO THE MICROVASCULATURE GIVE RISE TO MICROBLEEDS AND MICROINFARCTS IN CEREBRAL AMYLOID ANGIOPATHY

S.J. van Veluw¹, A.A. Scherlek¹, W.M. Freeze¹, A. Ter Telgte¹, A.J. Van der Kouwe², B.J. Bacskai¹, M.P. Frosch³ and S.M. Greenberg¹

¹Massachusetts General Hospital, Neurology, Boston, USA

²Massachusetts General Hospital, Radiology, Boston, USA

³Massachusetts General Hospital, Pathology, Boston, USA

The aim of this study was to assess the relationship between vascular amyloid β (Aβ) severity and microbleeds and microinfarcts in the context of cerebral amyloid angiopathy (CAA) at the whole-brain, regional, local, and single-vessel level.

Intact formalin-fixed hemispheres of 12 CAA cases were scanned at 3T MRI, followed by histopathological examination in pre-defined areas and serial sectioning in targeted areas with multiple lesions.

In total 1,168 cortical microbleeds (mean 97 ± 39 per case) and 456 cortical microinfarcts (mean 38 ± 46 per case) were observed on ex vivo MRI. Increasing CAA severity at the whole-brain or regional level was not associated with the number of microbleeds or microinfarcts detected on MRI or microscopically. However, locally, the density of Aβ positive cortical vessels was lower surrounding microbleeds compared to simulated control lesions, and higher surrounding microinfarcts. Serial sectioning revealed that for (n = 28) microbleeds both Aβ (4%) and smooth muscle cells (4%) were almost never present in the vessel wall at the site of bleeding, but Aβ was frequently observed upstream or downstream (71%), as was extensive fibrin (ogen) build-up (87%) (Fig.1). In contrast, for (n = 22) microinfarcts vascular Aβ was almost always observed at the core of the lesion (91%, p < 0.001) as well as upstream or downstream (82%), but few vessels associated with microinfarcts had intact smooth muscle cells (9%) (Fig.2).

Our findings suggest that microbleeds happen due to extensive vessel wall remodeling and loss of Aβ from a single vessel, whereas microinfarcts are the result of increased Aβ and presumed vessel stiffening locally.

n/a

HISTOPATHOLOGICAL VERIFICATION OF EX VIVO DWI+ LESIONS IN CEREBRAL AMYLOID ANGIOPATHY: EVIDENCE FOR ACUTE ISCHEMIA

A. ter Telgte^1,2, A. Scherlek², Y. Reijmer³, A. van der Kouwe⁴, M. Frosch^2,5, F.E. de Leeuw¹, B. Bacskai², S. Greenberg⁶ and S. van Veluw^2,6

¹Department of Neurology- Donders Institute for Brain- Cognition and Behaviour, Radboud University Medical Center, Nijmegen, The Netherlands

²MassGeneral Institute for Neurodegenerative Disease, Massachusetts General Hospital and Harvard Medical School, Charlestown- MA, USA

³Department of Neurology- Brain Center Rudolf Magnus-, University Medical Center Utrecht, Utrecht, The Netherlands

⁴Athinoula A. Martinos Center for Biomedical Imaging- Department of Radiology, Massachusetts General Hospital, Charlestown- MA, USA

⁵Neuropathology Service- C.S. Kubik Laboratory for Neuropathology, Massachusetts General Hospital and Harvard Medical School, Boston- MA, USA

⁶J. Philip Kistler Stroke Research Center- Department of Neurology, Massachusetts General Hospital and Harvard Medical School, Boston- MA, USA

Small diffusion-weighted imaging positive (DWI+) lesions, frequently observed in patients with cerebral amyloid angiopathy (CAA), are suggestive of acute microinfarcts. However, histopathological confirmation remains absent. To verify the acute ischemic nature of DWI+ lesions, we (1) investigated whether ex vivo DWI can be used to detect acute lesions, and (2) validated our findings in a positive control case that demonstrated multiple small DWI+ lesions on in vivo MRI ≤2 weeks before death.

Formalin-fixed whole hemispheres of 12 definite CAA cases and three non-CAA controls, and two posterior brain slabs from the positive control case, underwent ex vivo 3T DWI (1.0 × 1.0 × 1.0 mm³) and T2 (0.5 × 0.5 × 0.5 mm³). Small cortical lesions were visually assessed and classified as either DWI+/T2+ or DWI-/T2+. From three CAA cases representative DWI+/T2+ and DWI-/T2+ lesions were sampled for histopathology.

136 DWI+/T2+ lesions were observed in 10/12 CAA cases vs. 0 in non-CAA controls (p = .022). DWI+/T2+ lesions corresponded to focal areas of tissue pallor, with hypoxic neurons and few reactive astrocytes, indicative of acute microinfarcts (n = 7 verified) (Fig.1). DWI-/T2+ lesions corresponded to focal areas of tissue loss or cavitation, and reactive astrocytes, indicative of chronic microinfarcts (n = 4 verified) (Fig.2). The positive control case exhibited many ex vivo DWI+/T2+ lesions, with the same histopathological features as DWI+/T2+ lesions in the CAA cases.

This MRI-histopathology study provides evidence that small DWI+ lesions in CAA represent acute microinfarcts. Our findings also support the use of ex vivo DWI as a method to identify these acute lesions, pending confirmation in larger studies.

N/A

MIGRAINE PREVALENCE AND SYMPTOMATOLOGY IN HEREDITARY CEREBRAL AMYLOID ANGIOPATHY DUTCH TYPE

E.A. Koemans¹, S. Voigt¹, I. Rasing¹, E.S. van Etten¹, E.W. van Zwet², M.A.A. van Walderveen³, M.J.H. Wermer¹ and G.M. Terwindt¹

¹Leiden University Medical Center, Neurology, Leiden, The Netherlands

²Leiden University Medical Center, Biomedical Data Sciences, Leiden, The Netherlands

³Leiden University Medical Center, Radiology, Leiden, The Netherlands

To determine whether migraine is a symptom of amyloid pathology and is correlated with age of first intracerebral hemorrhage (ICH), occipital located ICH and cortical superficial siderosis (cSS) in patients with Hereditary-Cerebral-Hemorrhage-With-Amyloidosis-Dutch-type (HCHWA-D), a hereditary model of cerebral amyloid angiopathy (CAA).

We recruited HCHWA-D mutation-carriers who visited the Leiden University Medical Center between 2012–2018. Migraine symptoms were evaluated by interview and classified according to the International Classification of Headache Disorders (ICHD-3). Available MRI scans were scored for ICH location and presence of cSS by two independent observers. Kaplan-Meier survival-analysis was used to assess age of ICH onset in patients with and without migraine. The relation between migraine and presence of cSS and ICH location was calculated with Poisson regression-analysis, adjusted for confounders.

We included 95 participants with HCHWA-D (56% women, mean age 58 years, 66 with a history of ICH and 27 presymptomatic mutation-carriers). Overall 48 (51%) participants had a lifetime history of migraine, all with visual aura. In 75% of participants migraine was the inaugural symptom of HCHWA-D. There was no difference in age of first ICH for participants with and without migraine (P = 0.75). Migraine was not increased in relation to cSS (RR 1.55 (95% CI 0.68-3.57)) or occipital ICH (RR 1.22 (95% CI 0.66-2.26)).

Migraine with aura is an important symptom in HCHWA-D, observed in more than half of the participants. Migraine is often inaugural and is not related with earlier ICH onset, occipital ICH or cSS. Migraine with aura seems an early marker of disease in hereditary CAA.

N/A

ADVANCED DIFFUSION MODELING FOR CEREBRAL SMALL VESSEL DISEASE

M. Konieczny¹, A. ter Telgte², K. Wiegertjes², S. Finsterwalder¹, B. Gesierich¹, M. Huebner¹, M.A. Caballero¹, M. Ewers¹, J.P. Marques³, D.G. Norris³, A.M. Tuladhar², R. Schmidt⁴, M. Dichgans¹, F.E. de Leeuw² and M. Duering¹

¹University Hospital- LMU Munich, Institute for Stroke and Dementia Research ISD, Munich, Germany

²Donders Institute for Brain- Cognition and Behaviour- Radboud University Medical Center, Department of Neurology, Nijmegen, The Netherlands

³Donders Institute for Brain- Cognition and Behaviour- Radboud University Medical Center, Center for Cognitive Neuroimaging, Nijmegen, The Netherlands

⁴Medical University of Graz, Department of Neurology, Graz, Austria

Diffusion imaging detects subtle microstructural changes in cerebral small vessel disease (SVD). Measures from diffusion tensor imaging (DTI) are highly sensitive for tissue alterations and strongly associated with clinical deficits but provide only limited insight into underlying microstructure. Therefore, we evaluated the utility of advanced diffusion models (diffusion kurtosis imaging [DKI], free water imaging, neurite orientation dispersion and density imaging) and hypothesized that, compared to DTI, advanced measures explain more variance in SVD-related cognitive impairment.

We studied two samples: sporadic SVD (spSVD, n = 54) for primary analysis and genetically defined SVD (CADASIL, n = 59) for independent validation. Diffusion data was acquired at 3 Tesla using a multi-band, multi-shell acquisition protocol. Cognitive performance was assessed with trail making test (TMT) and a computerized reaction time task (RT). Univariate associations between diffusion and cognitive measures were analyzed by linear regressions. For multivariable analysis including multiple imaging parameters, age and sex, we used random forest regression.

In spSVD, linear regression models showed strongest association between DKI measures and both cognitive tests (mean kurtosis: R² = 0.21, p = 5.3e-04 for TMT; axial kurtosis: R² = 0.09, p = 0.017 for RT). In CADASIL, mean diffusivity from DTI explained more variance (mean diffusivity: R² = 0.28, p = 1.2e-05 for TMT; followed by radial kurtosis: R² = 0.21, p = 5.5-04 for RT). In multivariable models, DKI measures (mean kurtosis for spSVD; radial kurtosis for CADASIL) showed the highest contribution for explaining cognitive deficits.

By taking information on microstructural complexity into account, the advanced diffusion model DKI allows to characterize SVD and related cognitive deficits more accurately than previous diffusion methods.

N/A

THE DYNAMIC NATURE OF WHITE MATTER HYPERINTENSITIES IN ALZHEIMER’S DISEASE SUPPORTS VASCULAR CONTRIBUTIONS TO ALZHEIMER’S DISEASE

O. Hamilton¹, J. Ramirez², K. Walker², A.A. McNeely², J.M. Wardlaw¹ and S.E. Black²

¹University of Edinburgh, Centre for Clinical Brain Sciences, Edinburgh, United Kingdom

²Sunnybrook Research Institute, LC Campbell Cognitive Neurology Research Unit- Hurvitz Brain Sciences Program, Toronto, Canada

White Matter Hyperintensities of presumed vascular origin (WMH) represent pathological cerebrovascular changes and contribute to the pathophysiology of Alzheimer’s disease (AD). Recent studies have reported WMH regression over time, but variability in lesion regression estimates is high, and little is known about potential clinical determinants.

We included participants diagnosed with AD (n = 124) and controls (n = 45) with repeated MR imaging (mean interscan interval = 1.2 years), from the Sunnybrook Dementia Study, Toronto, Canada. We used spatial overlap images to quantify volumes of WMH shrinkage and growth for each individual.

Although WMH regression and progression occurred to some degree in all participants, AD patients demonstrated greater dynamic changes than controls (shrinkage: median 19.4cc, range 1.0-241.4cc, p = 2x10⁻⁴; growth: median 23.6cc, range 0.3-225.2cc, p = 0.005). In both the AD and the control group, greater WMH shrinkage and growth were associated with greater baseline WMH volume (p = 1x10⁻¹⁵); furthermore, greater shrinkage was strongly associated with greater growth, and vice versa (p = 1x10⁻¹⁵). Participants with the most WMH shrinkage/growth (top 50%) were older than those with the least WMH changes (p = 0.002). Those with the most WMH shrinkage had higher baseline prevalence of hypertension (p = 0.045), and greater volumes of lacunes (p = 2x10⁻⁵) and enlarged perivascular spaces (p = 0.013).

We show that WMH are dynamic in AD, not just in patients with vascular disease as demonstrated in several recent studies. WMH regression could reflect interstitial fluid fluctuations due to various mechanisms, such as oedema. The associations are similar to those found in cerebrovascular disease populations, providing further evidence for the vascular nature of AD.

N/A

INCREASED BLOOD BRAIN BARRIER PERMEABILITY AND INFLAMMATION IN CEREBRAL SMALL VESSEL DISEASE

J. Walsh¹, D.J. Tozer¹, H. Sari¹, Y.T. Hong¹, T.D. Fryer¹, G.B. Williams¹, F.I. Aigbirhio¹, J. O’Brien² and H.S. Markus¹

¹University of Cambridge, Clinical Neurosciences, Cambridge, United Kingdom

²University of Cambridge, Psychiatry, Cambridge, United Kingdom

Cerebral small vessel disease (SVD) is responsible for 20% of strokes and is the most common cause of vascular dementia. It is characterised by both white matter lesions and more diffuse white matter damage outside the lesions. Despite its importance, the mechanisms causing white matter damage are incompletely understood. Both inflammation and increased blood brain barrier (BBB) permeability have been suggested to play a role. We hypothesised that BBB permeability and inflammation would be increased in the normal appearing white matter of SVD patients compared to healthy controls.

A PET/MR system was used to acquire simultaneous dynamic contrast enhanced MRI with Patlak modelling to measure BBB leakage, and PK-11195 PET to assess microglia activation as a marker of inflammation. 20 sporadic SVD patients, 20 genetic SVD patients (CADASIL) and 20 controls were studied.

There was higher permeability in the normal appearing white matter of sporadic SVD patients compared to both CADASIL (p = 0.016) and control (p < 0.001) groups. There was also an increase in the volume of permeability hotspots in sporadic SVD patients compared to control (p = 0.024). The number of inflammatory hotspots was increased in the normal appearing white matter of both sporadic SVD (p = 0.031) and CADASIL (p = 0.020) compared to control.

Our results suggest that inflammation and increased permeability of the BBB may have a role in the pathophysiology of white matter damage in SVD. Further research is needed to see if these markers predict white matter damage in longitudinal studies, and whether the can be altered therapeutically.

N/A

LEUKOARAIOSIS IN ASSOCIATION WITH LONG-TERM OUTCOMES AFTER ISCHEMIC STROKE: A SYSTEMATIC REVIEW AND META-ANALYSIS

M. Georgakis¹, M. Duering¹, J. Wardlaw M.² and M. Dichgans¹

¹LMU MunichUniversity Hospital, Institute for Stroke and Dementia Research, Munich, Germany

²University of Edinburgh, Division of Neuroimaging Science- Centre for Clinical Brain Science- Edinburgh Imaging and UK Dementia Research Institute, Edinburgh, United Kingdom

Leukoaraiosis is an established neuroimaging marker of cerebral small vessel disease. While common in patients with ischemic stroke, its association with outcomes post-stroke remains to be defined. We aimed to investigate the relationship between baseline leukoaraiosis in ischemic stroke patients and long-term risk of dementia, functional impairment, recurrent stroke, and mortality.

We systematically searched Medline and Scopus for studies of ischemic stroke patients examining longitudinal associations between MRI- or CT-assessed leukoaraiosis at baseline with dementia, functional impairment, recurrent stroke, and mortality at 3 months or later post-stroke. We evaluated study quality with the Newcastle-Ottawa scale and pooled relative risks (RR) for the presence and severity of leukoaraiosis with random-effects meta-analyses.

Pooling 104 studies of 71,298 ischemic stroke patients, we found moderate/severe leukoaraiosis at baseline to be associated with increased risk of cognitive impairment (RR: 2.29, 95% CI: 1.48-3.54), dementia (RR: 2.17, 95% CI: 1.72-2.73), functional impairment (RR: 2.21, 95% CI: 1.83-2.67), recurrent stroke (RR: 1.65, 95% CI: 1.36-2.01), all-cause mortality (RR: 1.72, 95% CI: 1.47-2.01), and cardiovascular mortality (RR: 2.02, 95% CI: 1.44-2.83). Both MRI- and CT-defined leukoaraiosis was associated with the outcomes and all associations followed dose-response patterns for leukoaraiosis severity. Sensitivity analyses of high-quality studies and studies adjusting for age, stroke severity, and cardiovascular risk factors, as well as publication bias-adjusted analyses showed consistently stable results.

Leukoaraiosis is associated with substantially increased risk of dementia, functional impairment, stroke recurrence, and mortality after ischemic stroke. Leukoaraiosis assessment may aid clinical prognostication and the planning of future trials.

N/A

Scientific Communications 13-Carotid Artery Disease Imaging and Prognosis

IPSILATERAL NON-STENOTIC CAROTID PLAQUES AND DETECTION OF ATRIAL FIBRILLATION IN PATIENTS WITH EMBOLIC STROKE OF UNDETERMINED SOURCE: THE AF-ESUS STUDY

G. Ntaios¹, K. Perlepe², G. Sirimarco³, D. Strambo³, A. Eskandari³, E. Karagkiozi¹, A. Vemmou⁴, E. Koroboki⁵, E. Manios⁴, K. Makaritsis¹, P. Michel³ and K. Vemmos⁴

¹University of Thessaly, Department of Medicine, Larissa, Greece

²University of Thessaly- Larissa- Greece, Department of Medicine, Λάρισα, Greece

³Centre Hospitalier Universitaire Vaudois and University of Lausanne, Stroke Center and Neurology Service- Department of Clinical Neurosciences, Lausanne, Switzerland

⁴Medical School of Athens- Alexandra Hospital, Department of Clinical Therapeutics, Athens, Greece

⁵National and Kapodistrian University of Athens, Department of Neurology, Athens, Greece

Carotid plaques causing < 50% stenosis (non-stenotic carotid plaques) are considered as a potential cause underlying embolic strokes of undetermined source (ESUS) but the strength of this association is unclear as the existing evidence is controversial. We hypothesized that if the causal association between non-stenotic carotid plaques and ESUS is strong, atrial fibrillation (AF) would be less frequently detected in ESUS patients with non-stenotic carotid plaques compared to those without.

We pooled data of all consecutive ESUS patients from 3 prospective stroke registries. Multivariate stepwise regression assessed the association between the presence of non-stenotic carotid plaques and AF detection. The 10-year cumulative probabilities of AF detection were estimated by the Kaplan–Meier product limit method.

Among 777 patients followed for 2642 patient-years, 341 (38.6%) patients had an ipsilateral non-stenotic carotid plaque. AF was detected in 112 (14.4%) of patients in the overall population during follow-up. The overall rate of AF detection was 8.5% in patients with non-stenotic carotid plaques (2.9% per 100patient-years), and 19.0% in patients without (5.0% per 100patient-years) (unadjusted HR:0.56, 95% CI:0.37-0.84). The presence of ipsilateral non-stenotic carotid plaques was associated with lower probability for AF detection (adjusted HR:0.57, 95% CI:0.34-0.96, p = 0.03). The 10-year cumulative probability of AF detection was lower in patients with ipsilateral non-stenotic carotid plaques compared to those without (34.5%, 95% CI:21.8-47.2 vs. 49.0%, 95% CI:40.4-57.6 respectively, log-rank-test:11.8, p = 0.001)

AF is less frequently detected in ESUS patients with non-stenotic carotid plaques compared to those without.

N/A

VASCULAR TORTUOSITY PREDISPOSES TO EPIAORTIC VESSEL DISSECTION: CASE-CONTROL STUDY

G. Venturini¹, L. Vuolo², G. Pracucci¹, A. Picchioni¹, Y. Failli¹, F. Benvenuti¹ and C. Sarti¹

¹University of Florence, Neurofarba, Florence, Italy

²Careggi University Hospital, Neuroradiology Unit, Florence, Italy

Recent studies suggest that vascular tortuosity (VT) may represent an endogenous factor interacting with exogenous ones to determine epiaortic vessels dissection (EVD).

We aimed to verify this hypothesis comparing VT in stroke patients with and without EVD.

42 EVD stroke patients (cases) were compared with 41 non EVD stroke (controls).

Vascular Tortuosity Index (VTI) according to Giossi et al algorithm i.e. [(actual distance/straight line distance -1) ×100] between proximal and distal point of both CA and VA (figure 1), and kinking and coiling according to Weibel-Fields-Metz classification were measured on angio CT images.

Differences between groups, analyzed by ANOVA for numeric variables and chi-square test for categoric ones, were adjusted for hypertension, diabetes, smoking, and dyslipidemia. ROC curve was used to find a VTI threshold beyond which the risk of dissection was significantly increased.

VTI was significantly higher in cases than in controls considering CA and VA together (p = 0.031) and CA alone (p = 0.006) but not VA alone (p = 0.239) (Table). VTI threshold value for CAD was 27.5% (p = 0,014) (Figure 2). Cases had significantly higher number of coiling (p = 0.028) but not of kinking (p>0,05) than controls (Table).

Vascular tortuosity appears to be an independent risk factor for dissection specially for carotid district. VA seems to be less susceptible to dissection than CA, perhaps for different embryogenesis. VTI threshold as CAD predictor deserves confirmation in larger studies.

N/A

PLAQUE MORPHOLOGY OF ACUTE INTRACRANIAL ATHEROSCLEROTIC DISEASE

T.W. Leung¹, X. Leng¹, Y.O. Soo¹, V.H. Ip¹, A.Y. Chan¹, L.W. Au¹, F.S. Fan¹, S.H. Ma¹, K.K. Ma¹, A.Y. Lau¹, H. Leung¹, V.C. Mok¹, K.F. Hui², R. Li³, S.H. Li⁴, M. Fu⁵, W.C. Fong⁶ and S.C. Yu⁷

¹The Chinese University of Hong Kong, Department of Medicine and Therapeutics, Hong Kong, Hong Kong S.A.R

²United Christian Hospital, Department of Medicine, Hong Kong, Hong Kong S.A.R

³Pamela Youde Nethersole Eastern Hospital, Department of Medicine, Hong Kong, Hong Kong S.A.R

⁴North District Hospital, Department of Medicine, Hong Kong, Hong Kong S.A.R

⁵Tuen Mun Hospital, Department of Medicine and Geriatrics, Hong Kong, Hong Kong S.A.R

⁶The Queen Elizabeth Hospital, Department of Medicine, Hong Kong, Hong Kong S.A.R

⁷The Chinese University of Hong Kong, Department of Diagnostic and Interventional Radiology, Hong Kong, Hong Kong S.A.R

Intracranial atherosclerotic disease (ICAD) is globally a major ischemic stroke subtype with high recurrence. Understanding the morphologyof cerebral plaques duringthe acute phasemay help identify vulnerable lesions prone to relapse.

Patients with recent ischemic stroke or transient ischemic attack attributed to high-grade ICAD (60-99% stenosis) underwent MRI brain and 3-dimensional rotational angiography (3DRA). We delineated theplaque geometry and correlated infarct topography with plaque morphology.

Among the recruited180 patients (median age = 60 years), ICAD lesions distributed in M1 middle cerebral artery (MCA)(n = 132), C6/C7 intracranial internal carotid artery (ICA)(n = 26), across MCA and ICA (n = 6), basilar artery (n = 14) and V4 vertebral artery (n = 2). The median luminal stenosis was 75%. The majority of ICAD lesionsshowed an irregular or ulcerativesurface contour (n = 131; 72.7%). Plaque ulcers indicative of fibrous cap rupture (n = 30) were most commonly found at the proximal (n = 13; 44%) and middle one-third (n = 10; 33%) of the stenoses. Compared with non-ulcerative lesions, ulcerative plaques were thicker (p = 0.003), of a steeper upstream shoulder (p < 0.001), had more adjoining branch atheromatous disease (p = 0.024), and were associated with coexisting acute and established infarcts in the corresponding border-zones (odds ratio = 5.29; p = 0.009).

UlcerativeICAD plaques entailed a higher atherosclerosis burden in vulnerable geometry. The larger infarct load at the borderzones of ulcerative plaques might signify artery-to-artery thrombo-embolism with impaired washout as a prevailing stroke mechanism, justifying an initial intensified antiplatelet regimen. The frequent coexistence of branch atheromatous disease explained the association of perforator stroke with ICAD, and underscored the risk of ‘snowplow effect’ during angioplasty/stenting in this subgroup.

N/A

A SIMPLE CLINICAL SCORE (SCAIL SCORE) INCLUDING PLAQUE INFLAMMATION-RELATED METABOLISM ON 18FDG-PET AND STENOSIS SEVERITY IDENTIFIES PATIENTS WITH CAROTID STENOSIS AND EARLY RECURRENT STROKE

P. Kelly¹, P. Camps-Renom², N. Giannotti³, J. Marti-Fabregas², J.C. Baron⁴, S. Cronin⁵, S.B. Coutts⁶, E. Dolan⁷, A. Fernandez-Leon⁸, S. Foley³, J. Harbison⁹, G. Horgan¹, J. McCabe¹, C. McDonnell¹⁰, J. McNulty¹¹, V.K. Sharma¹², D. Williams¹³, M. Marnane¹, M. O’Connell¹⁴ and S. Murphy¹⁵

¹Stroke Clinical Trials Network Ireland and University College Dublin/Mater University Hospital, Neurology, Dublin, Ireland

²Hospital de la Santa Creu i Sant Pau, Servi de Neurologia, Barcelona, Spain

³University College Dublin, Diagnostic Imaging Dept, Dublin, Ireland

⁴Hospital Sainte-Anne- Universite Paris 5, Neurology, Paris, France

⁵Cork University Hospital, Neurology, Cork, Ireland

⁶University of Calgary, Neurology, Calgary, Canada

⁷Connolly Hospital/RCSI, Stroke Dept, Dublin, Ireland

⁸Hospital de la Santa Creu i Sant Pau, Servi de Medicina Nuclear, Barcelona, Spain

⁹St James Hospital/Trinity College Dublin, Stroke, Dublin, Ireland

¹⁰Stroke Clinical Trials Network Ireland and University College Dublin/Mater University Hospital, Vascular Surgery, Dublin, Ireland

¹¹University College Dublin, Diagnostic Imaging, Dublin, Ireland

¹²National University Hospital Singapore, Neurology, Singapore, Singapore

¹³Beaumont Hospital/Royal College of Surgeons Ireland, Stroke, Dublin, Ireland

¹⁴Mater University Hospital/University College Dublin, Radiology, Dublin, Ireland

¹⁵Stroke Clinical Trials Network Ireland and University College Dublin/Mater University Hospital, Stroke, Dublin, Ireland

Lumen stenosis predicts recurrent stroke in patients with carotid atherosclerosis but provides no information about plaque inflammation, which is associated with thromboembolism. ¹⁸F-FDG-PET identifies inflammation-related metabolism in carotid plaque. We aimed to derive a clinical score incorporating stenosis and inflammation to improve risk stratification of patients with recently-symptomatic carotid atheroma.

We pooled individual-patient data from three prospective studies with similar methodology. Included patients had recent stroke/TIA, with ipsilateral carotid stenosis (≥50%). Patients had co-registered carotid ¹⁸F-FDG PET/CT angiography, expressed as maximum Standardised Uptake Value (SUV_max) and 90-day follow-up. To derive a clinical score, points were assigned based on adjusted hazard ratios on multivariable Cox regression.

In the pooled cohort (n = 185, 37 recurrent strokes), the adjusted HR of recurrent stroke per 1g/mL SUVmax was 2.15 (CI 1.39-3.32, p = 0.001). The SCAIL (Symptomatic Carotid Atheroma Inflammation Lumen-stenosis) score (range 1–5) was derived, based on SUVmax and stenosis categories. The adjusted HR of early recurrent stroke associated with 1-point increase in SCAIL score was 2.1 (CI 1.4–3.0, p < 0.001). Recurrent stroke risk increased linearly with higher SCAIL score (ptrend < 0.001). When the score was collapsed as Low (1–2), Medium (3), High (4–5) SCAIL categories, recurrent stroke risk increased from 9.9%, 24.6%, 53.9% (ptrend < 0.001). Compared to stenosis severity alone (c-statistic 0.60, CI 0.51–0.69), the c-statistic for recurrent stroke improved with the SCAIL score (0.72, CI 0.63–0.81).

The SCAIL score, incorporating plaque inflammation-related 18F-FDG uptake and stenosis, was independently associated with early recurrent stroke. If validated, our findings may lead to improved patient selection for carotid revascularisation.

n/a

EXPLORING THE FEASIBILITY OF MRI INTRACRANIAL VESSEL WALL IMAGING (VWI) IN THE DIAGNOSES AND MONITORING OF INTRACRANIAL INFLAMMATORY VASCULOPATHY

L. Benjamin¹, J. Markus², M. Sokolska², V. Aggarwal³, E. Nastouli⁴, G. Yen⁴, J. Mason⁵, H.R. Jager¹, R. Simister¹

¹University College London, Stroke Research Centre- Brain Repair and Rehabilitation, London, United Kingdom

²University College London Hospitals, Radiology, London, United Kingdom

³University College London Hospitals, Neurology, London, United Kingdom

⁴University College London Hospitals, Virology, London, United Kingdom

⁵Imperial College London, Medicine, London, United Kingdom

Current stroke guidelines do not facilitate the expedient diagnosis of inflammatory causes of stroke. We report on the feasibility of integrating MRI-intracranial-VWI into a clinical stroke service, to diagnose and manage patients with suspected intracranial inflammatory vasculopathy.

Between Dec 2016-Jun 2018, patients with a stroke or TIA and intracranial stenosis were recommended for investigation of possible intracranial inflammatory vasculopathy at our weekly case review meeting. Selected cases underwent MRI-VWI together with autoimmune and infection screening. A dedicated MDT was set up to establish the diagnosis and determine treatment. Six-month outcomes were determined.

Following MDT review, sixteen patients had intracranial vessel wall inflammation. The median age was 41-yrs (IQR:29-51), and 7 (43%) were female. Eight (50%) had an autoimmune disorder, and 7 (44%) had at least one conventional vascular risk factor. Six (38%) had evidence of systemic PET avidity but none had intracranial avidity. Median CRP was 4.6 (IQR 1.0, 11.6) and only 6 (24%) had an abnormal cerebrospinal fluid. Two VWI-positive cases had brain biopsy in keeping with an inflammatory vasculopathy, one had a confirmed herpesvirus-2 associated vasculopathy. Although not significant, the prevalence of recurrent clinical events was higher among those not treated with immunosuppression [57%] compared with those on immunosuppression [43%] p = 0.43.

We have demonstrated the feasibility of MRI-VWI as a clinical platform to detect vessel wall inflammation, and in turn revealed potentially treatable causes of young stroke. Detailed phenotyping of these inflammatory vasculopathies is warranted to understand the mechanism and tailor response to treatment.

N/A

ASSOCIATION OF PLAQUE INFLAMMATION MEASURED BY 18-FDG POSITRON EMISSION TOMOGRAPHY WITH MARKERS OF UNSTABLE PLAQUE MORPHOLOGY ON MRI IN PATIENTS WITH SYMPTOMATIC CAROTID STENOSIS

N. Giannotti¹, J. McNulty¹, S. Murphy^2,3,4, M. Marnane^2,4, T. Cassidy⁵, E. Dolan⁶, S. Foley¹, G. Horgan², E. Kavanagh⁷, C. McDonnell⁸, M. O’Connell⁷ and P. Kelly^2,4

¹University College Dublin, Radiography & Diagnostic Imaging, Dublin, Ireland

²University College Dublin, HRB Stroke Clinical Trials Network Ireland, Dublin, Ireland

³RCSI, School of Medicine, Dublin, Ireland

⁴Mater University Hospital, Neurovascular Unit for Translational and Therapeutics Research, Dublin, Ireland

⁵St Vincent’s Hospital, Geriatric medicine, Dublin, Ireland

⁶Connolly Hospital, Geriatric medicine, Dublin, Ireland

⁷Mater University Hospital, Radiology, Dublin, Ireland

⁸Mater University Hospital, Vascular Surgery Department, Dublin, Ireland

Pathologic studies suggest that unstable plaque morphology and inflammation are associated with cerebrovascular events. ¹⁸FDG-PET may quantify imaging biomarkers of plaque inflammation and HR-MRI identifies morphologic features of plaque instability. However, almost no studies have investigated the relationship of these biomarkers in patients with symptomatic carotid stenosis. We investigated the association between plaque characteristics measured with HR-MRI and PET.

A sub-group of patients included in the prospective BIOVASC plaque imaging study were included. Inclusion criteria were: (1) Speech/motor TIA or non-severe stroke (Rankin≤3) < 72hours (2) Ipsilateral carotid stenosis ≥50% (3) Age ≥50 (4) Carotid plaque high-resolution HR-MRI and ¹⁸FDG-PET/CTA performed. Semi-automated plaque analysis of HR-MRI axial images was done using PlaqueView software. PET images were co-registered with CTA and inflammation-related metabolism expressed as maximum standardised uptake value (SUV_max) (Figure1).

25 patients met inclusion criteria (72% men, mean age 65 years) (Table1). MRI-measured plaque volume was lower in patients receiving statin (p = 0.016). Plaque area (78.01 vs 61.26mm2, p = 0.006) and thickness (1.95 vs 1.71mm, p = 0.05) were greater in patients with hypertension. 18FDG-PET SUVmax was associated with serum LDL-cholesterol (p = 0.05) and with MRI-measured plaque lipid-rich necrotic core (p < 0.001) in the corresponding axial images. SUVmax was inversely associated with fibrous cap thickness (p = 0.04) and calcium volume (p = 0.039) (Table2).

We demonstrated novel correlations of non-invasive imaging markers of plaque inflammation with morphological MRI markers of plaque instability. If replicated, our findings may support the application of combined HR-MRI/PET to detect vulnerable plaque in future clinical practice and randomised trials.

N/A

JUXTALUMINAL CAROTID PLAQUE ECHODENSITY AND ITS PREDICTIVE VALUE FOR CEREBROVASCULAR SYMPTOMATOLOGY

T. Tegos¹, C. Xerras¹, G. Konstantinidis¹, G. Stefanou¹, I. Papagianis¹, V. Siokas¹, A. Papadimitriou¹, M. Arnaoutoglou¹ and M. Tsolaki¹

¹AHEPA University Hospital, A Neurology Department, Thessaloniki, Greece

Previous studies suggested that symptomatic carotid plaques are echolucent on ultrasound, whereas asymptomatic ones are echogenic. The aim of this study was to determine whether juxtaluminal plaque echodensity constitutes a better discriminator of the symptomatic and asymptomatic status, as compared to global plaque echodensity

Analysis performed on ultrasound images of 407 carotid plaques with 50%-99% stenosis (312 patients, 189 symptomatic and 218 asymptomatic plaques). The global (100%) plaque Grey Scale Median (GSM₁₀₀) and also the juxtaluminal 10%, 25%, 50% plaque area GSM (GSM₁₀, GSM₂₅, GSM₅₀) were evaluated in a computer software.

Symptomatic plaques were associated with median GSM100 of 9 whereas the asymptomatic ones of 32.5 (p = 0.0001). The corresponding values for the median GSM10, GSM25, GSM50 were: (3,5,6) for symptomatic plaques and (28.5,32.5,33) for asymptomatic ones respectively (p = 0.0001 for each analysis). ROC curves demonstrated a more adequate ability of GSM10 and GSM25 over GSM100 in separating the symptomatic from the asymptomatic plaques (difference between areas: 0.053 and 0.22, p = 0.0011 and p = 0.0001 respectively). ROC curves failed to demonstrated a more adequate ability of GSM50 over GSM100 in separating the symptomatic from the asymptomatic plaques (difference between areas: 0.0141, p = 0.4).

Our results suggested that superficial juxtaluminal plaque echodensity (GSM10, GSM25) proved as a more adequate index, compared with global plaque echodensity (GSM100), in the separation of symptomatic and asymptomatic carotid plaques. This concept might be of use in natural history studies of asymptomatic individuals with carotid plaques, having stroke as an exit point.

N/A

CAROTID ATHEROSCLEROSIS AND PROGRESSION OF WHITE MATTER LESIONS IN STROKE SURVIVORS: A SEVEN YEAR FOLLOW-UP STUDY

H. Ihle-Hansen¹, H. Ihle-Hansen², B. Fure³, B. Thommessen⁴, A. Øksengård¹, M. Beyer⁵, T.B. Wyller² and G. Hagberg¹

¹Bærum Hospital- Vestre Viken Hospital Trust, Department of Medicine, Drammen, Norway

²Oslo University Hospital, Department of Medicine, Oslo, Norway

³Karlstad Central Hospital and Institute of Public Health, Department of Internal Medicine, Karlstad, Sweden

⁴Akershus University Hospital, Department of Neurology, Lørenskog, Norway

⁵Oslo University Hospital, Department of Radiology and Nuclear Medicine, Oslo, Norway

We aimed to explore whether carotid atherosclerosis assessed with ultrasound during hospitalization for acute stroke is associated with progression of white matter lesions and cerebral atrophy measured on magnetic resonance imaging (MRI) in long-term follow-up after stroke.

All patients with a first-ever stroke or transient ischemic attack admitted to the stroke unit of Bærum Hospital, Norway in 2007/2008, were recruited and followed for seven years. Carotid ultrasound was performed at inclusion, with measurement of stenosis in the internal carotid artery (ICA). MRI was performed one and seven years after the index event and analyzed according to Fazekas scale (0–3), medial temporal lobe atrophy (MTA) (0–4) score and global cortical atrophy (GCA) scale (0–3).

Of 227 subjects recruited, 116 participants completed the seven years follow-up, of whom 75 had both the MRI examinations. Mean age 66.7 ± 10.8, 41% women and 9% had ≥50% ICA stenosis. Mean Fazekas scale was 1.7 ± 0.9 and 1.8 ± 1.0, mean MTA score 1.0 ± 1.0 and 1.7 ± 1.0, and mean GCA scale 1.4 ± 0,7 and 1.4 ± 0.6 after one and seven years, respectively. Mean change in Fazekas scale was 0.2 ± 0.6, 71% retained the same score, while 21% showed progression. Deterioration in GCA scale was seen in 18% and increasing MTA score in 61%. Carotid stenosis was associated with progression in Fazekas score (beta 0.25, 95CI 0.04-0.90, p < 0.05), but not with change in MTA score or GCA scale.

Carotid atherosclerosis is associated with white matter lesions progression in stroke survivors, and observed in one out of five patients seven years post-stroke.

Clinicaltrials.gov (NCT00506818)

RECURRENCE OF CEREBRAL ISCHEMIC EVENTS IN SYMPTOMATIC CAROTID STENOSIS ARE RELATED TO PLAQUE INFLAMMATION

V. Sharma¹, P. Paliwal², B. Tan², L. Yeo², L. Wong², J. Chen², Z. Du², C.S. Hong², Y. Chee², H. Teoh², B. chan² and A. Sinha³

¹National University of Singapore, Neurology, Singapore, Singapore

²National University Hospital, Medicine, Singapore, Singapore

³National University Hospital, Diagnostic Imaging, Singapore, Singapore

Severity of symptomatic carotid stenosis often determines the treatment approach. However, severity explains only the regional hypoperfusion as the mechanism of cerebral ischemia. Artery-to-artery embolisation remains an important pathogenic mechanism in patients with high-risk carotid plaques. Inflammation is considered as the initiating event for plaque rupture and cerebral embolism. Using 18F-fluorodeoxyglucose positron-emission tomography (FDG-PET)/computed tomography (CT) and high-resolution magnetic resonance imaging (HR-MRI), we investigated the role of plaque imaging in stroke recurrence in our patients with recently symptomatic carotid stenosis.

This prospective study included consecutive patients within 30-days of recent stroke and ipsilateral carotid stenosis (≥50%). FDG uptake was quantified as mean standardized uptake values (SUV, g/ml). The ratio of T1 hyperintensity of carotid plaque to the ipsilateral sternocleidomastoid muscle (SCM) was recorded on T1-weighted fat suppressed HR-MRI. Patients were followed-up for stroke recurrence within 90-days.

Of the 125 patients included in the study, 19 (15%) suffered from recurrent cerebral ischemic event in the stenosed carotid territory within 90-days. Compared to patients who remained asymptomatic, patients with recurrent cerebral ischemia showed higher mean T1 carotid-SCM ratio (2.18 versus 1.39; p < 0.0001) and higher mean SUV value in the carotid plaque (3.19g/ml versus 1.22g/ml; p < 0.0001). Higher T1 carotid-SCM ratio on HR-MRI (OR 3.105, 95% CI 1.412-5.634; p < 0.0001) and higher mean SUV on FDG-PET (OR 3.815, 95% CI 2.832-7.461; p = 0.005) were independent predictors of recurrent cerebral ischemia.

FDG-PET/CT and HR-MRI of carotid stenosis are useful tools for risk stratification and may aid in better therapeutic decision-making.

N/A

Scientific Communications 14-Rehabilitation – Interventions

FACTORS ASSOCIATED WITH YOUNG STROKE SURVIVORS’ RETURN TO WORK IN A VERY EARLY REHABILITATION TRIAL (AVERT)

S. Cain^1,2, L. Churilov^1,2, J. Collier¹, K. Borschmann¹, V. Thijs^1,2, J. Bernhardt¹; The AVERT Collaboration Group

¹The University of Melbourne, Florey Institute of Neuroscience and Mental Health, Heidelberg, Australia

²Austin Health, Medicine, Heidelberg, Australia

Worldwide, the proportion of strokes occurring in the working-age is increasing. Return to work (RTW) is a key rehabilitation outcome. This study aimed to describe working-age stroke survivors and identify factors associated with RTW.

Data was collected in A Very Early Rehabilitation Trial (AVERT) at 56 acute stroke units across Australia, New Zealand, the United Kingdom, Singapore and Malaysia (2006-2015). For this analysis of AVERT patients, inclusion criteria were: age less than 65, working at time of stroke, and complete 12 month RTW data. The primary outcome was return to paid work at 12 months post stroke. Logistic regression analysis was conducted to determine the association of sociodemographic, stroke-related, psychological, and work-related factors on RTW.

Of 2104 AVERT patients, 376 met inclusion criteria. By 12 months, 221 (58.8%) participants had returned to work. RTW rates in Australia, the United Kingdom and Asia were not significantly different. On univariable analysis, the odds of RTW were higher with younger age (OR 0.95, 95% CI 0.92-0.97), no past diabetes (0.4, 0.24-0.67), lower stroke severity (NIHSS) (0.82, 0.78-0.86), less depressive traits (IDA Scale) at 3 months (0.87, 0.80-0.93), less disability (mRS) at 3 months, and full-time work prior to stroke (2.04, 1.23-3.38). On multivariable analysis, age (OR 0.94, 95% CI 0.91-0.98), stroke severity (0.92, 0.86-0.99), full-time work (2.33, 1.24-4.40), and 3-month disability retained their association with RTW.

Age, stroke severity, 3-month disability and premorbid work status are important factors associated with RTW. These results could help inform future interventional research, and target vocational rehabilitation.

ACTRN12606000185561

ALTERING THE REHABILITATION ENVIRONMENT TO IMPROVE STROKE SURVIVOR ACTIVITY (AREISSA TRIAL): A PATIENT-DRIVEN MODEL OF ENVIRONMENTAL ENRICHMENT

H. Janssen¹, L. Ada², J. Bernhardt³, L. Churilov⁴, S. Middleton⁵, M. Nilsson⁶, M. Pollack⁷, N. Spratt J⁸; AREISSA Trial Collaboration

¹University of Newcastle & Hunter New England Local Health District, Faculty of Health and Medicine, Newcastle, Australia

²University of Sydney, Faculty of Health Sciences, Sydney, Australia

³The Florey Institute of Neuroscience and Mental Health, Stroke, Melbourne, Australia

⁴University of Melbourne, Biostatistics, Melbourne, Australia

⁵Australian Catholic University, Faculty of Health Sciences, Sydney, Australia

⁶University of Newcastle, Faculty of Health and Medicine, Newcastle, Australia

⁷Hunter New England Health Local Health District, Rehabilitation, Newcastle, Australia

⁸Univeristy of Newcastle, Faculty of Health and Medicine, Newcastle, Australia

Use of environmental enrichment significantly improves sensorimotor recovery in experimental stroke. The AREISSA trial aimed to determine whether a patient-driven model of environmental enrichment reduced inactivity and was safe in the clinical setting.

A multi-site, cluster, cross-over trial with blinded measurement was conducted. Environmental enrichment included access to communal and individual physical, cognitive and social stimulation. Major outcomes were (i) level of inactivity (percentage of observations where patients were inactive on Day 10 +/- 3 days following rehabilitation admission), and (ii) safety (number of adverse events including falls).

At baseline, participants (n = 190) were a median of 69 years old (IQR 59,81), had mild stroke symptoms (median NIHSS score 5, IQR 3,7) and had moderately severe disability (median modified Rankin Scale score 4, IQR 3,4). After adjusting for age and stroke severity, there was no difference between groups in inactivity levels (5% less inactivity in the environmental enrichment group, 95% CI -2 to 13). Environmental enrichment resulted in less inactivity in those participants with already low levels of inactivity (9%, 95% CI 5 to 13). There was no observed difference between groups in adverse events (OR 0.5, 95% CI 0.2 to 1.1).

This patient-driven model of environmental enrichment was safe but reduced inactivity only in stroke survivors with already low levels of inactivity. Future testing of similar models of environmental enrichment requires consideration of strategies to enable activity in more disabled stroke survivors.

ACTRN12613000796785

HIGH-INTENSITY INTERVAL TRAINING AFTER STROKE (HIIT-STROKE STUDY) – EFFECTS ON PEAK OXYGEN UPTAKE – A RANDOMIZED CONTROLLED TRIAL

T.I. Gjellesvik¹, F. Becker², A.E. Tjønna³, B. Indredavik¹, H. Nilsen⁴, B. Brurok⁵, T. Tørhaug¹, M. Busoladzic⁶, S. Lydersen⁷ and T. Askim¹

¹Faculty of Medicine and Health Science, Department of Neuromedicine and Movement Science- NTNU, Trondheim, Norway

²Faculty of Medicine, Institute of Clinical Medicine- University of Oslo, Oslo, Norway

³Faculty of Medicine and Health Science, Department of Circulation and Medical Imaging- NTNU, Trondheim, Norway

⁴Ålesund Hospital, Department of Physical Medicine and Rehabilitation, Ålesund, Norway

⁵St. Olav’s University Hospital, Department of Physical Medicine and Rehabilitation, Trondheim, Norway

⁶Sunnaas Rehabilitation Hospital, Department of Research, Oslo, Norway

⁷Faculty of Medicine and Health Science, Regional Centre for Child and Youth Mental Health and Child Welfare- NTNU, Trondheim, Norway

High-intensity interval training (HIIT) is a promising training mode to increase peak oxygen uptake (VO_2peak) in the healthy population and in patients with cardiovascular diseases. This study examined if 8 weeks of HIIT would increase VO_2peak more than standard care in patients with stroke.

This was a multicenter, single-blinded, parallel group, randomized controlled trial. Participants with first ever stroke between 18 and 75 years of age, 3 months to 5 years post stroke with no serious co-morbidities were invited to participate. The intervention consisted of 24 supervised treadmill sessions with work periods of 4x4 minutes at an intensity between 85–95% of peak heart rate interspersed with 3 minutes of active recovery at 50–70% of peak heart rate. The control group received standard care. Primary outcome, analyzed by intention-to-treat, was VO_2peak 12 months after inclusion.

Seventy consenting participants were randomized to the intervention (n = 36) or the control group (n = 34), 41% were women, mean (SD) age was 57.6(9.3) years and 26.4(14.5) months had passed since stroke. The mean (SE) baseline VO2peak was 2.60(0.15) versus 2.86(0.15) (L/min-1), p = 0.198. At 12-month follow-up VO2peak was 2.63(0.15) versus 2.73(0.15) (L/min-1), p = 0.657, in the intervention (n = 28) and control group respectively (n = 28). The interaction between time and group (Figure 1) was significant, p = 0.014.

HIIT was not superior to standard care to improve VO2peak at end of follow-up in this sample of relatively well functioning stroke survivors. However, secondary analysis showed a significant interaction between time and group, indicating different trajectories for the groups.

NCT02550015

A PHYSICAL ACTIVITY PROGRAM IS NO MORE EFFECTIVE THAN NO INTERVENTION AT MAINTAINING UPPER LIMB ACTIVITY IN COMMUNITY-DWELLING PEOPLE WITH STROKE, THE LAST TRIAL

B. Langhammer¹, M. Gunnes², H. Ihle-Hansen³, B. Indredavik², T. Askim²; the LAST collaboration group

¹Oslo Metropolitan University, Physiotherapy Department, Oslo, Norway

²NTNU, Department of Neuroscience, Trondheim, Norway

³Oslo University Hospital, Department of Medicine, Oslo, Norway

The aim of this study was to determine whether an 18-month, home-based exercise program additional to usual care is more effective than usual care in terms of upper limb activity.

The design was a prospective, randomised controlled trial in three municipalities in Norway.

The experimental group received coaching on physical activity and exercise by a physiotherapist each month for 18 months after inclusion. The control group received standard care.

The primary outcome, in this secondary analysis, was Motor Assessment Scale (MAS) items 6, 7 and 8. Secondary outcomes were National Institute of Health Stroke Scale (NIHSS) item 5, the Stroke Impact Scale (SIS) item 7 and the Modified Ashworth Scale in flexion/extension of the elbow.

In total, 380 persons with stroke were included, with mean age 72 (11) years and baseline scores total NIHSS was 1.4 (2.2)/1.6 (2.4) and MAS items 6, 7 and 8 in the intervention/control group was 5.5 (1.2)/5.5 (1.2), 5.4 (1.4)/5.4 (1.3) and 3.6 (2)/3.5 (2), respectively. There was no significant difference between groups in terms of upper limb function in any of the MAS items, nor in SIS item 7 or tonus. In this population with minor stroke, upper limb activity was good at 3 months post stroke (74% of the maximum) and remained good 18 months later (77% of maximum).

After intervention, there was no difference between the experimental and control group in terms of upper limb activity.

NCT01467206

NON-INVASIVE REPEATED THERAPEUTIC STIMULATION FOR APHASIA RECOVERY NORTHSTAR): RESULTS ON PRIMARY LANGUAGE OUTCOME MEASURES

A. Zumbansen¹, S.E. Sandra², J.L. Chen^2,3, D.J. Edwards^4,5,6, A. Hartmann⁷, W.D. Heiss^8,9, S. Lanthier¹⁰, P. Lesperance¹¹, G. Mochizuki², C. Paquette¹², E.A. Rochon¹³, I. Rubi-Fessen⁷, J. Valles¹⁴, H. Vogt⁷, S. Wortman-Jutt^5,14 and A. Thiel¹

NORTHSTAR study group: Alexander Thiel Anna Zumbansen Sharon Shapiro Stephanie Houston Mica Vincent Dominique Gillis and Caroline Paquette (Jewish General Hospital Montreal) Alexander Hartmann Ilona Rubi-Fessen Heike Vogt and Thomas Rommel (RehaNova Cologne) Wolf-Dieter Heiss (Max Planck Institute für Stoffwechsel Forschung – MPI for Metabolism Research & Universität zu Köln Cologne) Josef Kessler (Universität zu Köln Cologne) Dylan J. Edwards Jennie Valles and Susan Wortman-Jutt (Burke Medical Research Institute White Plains) Sylvain Lanthier Marlène Lapierre Diana Mina Liliana Jastrzabek Paul Lesperance and Walid El-Abyad (Hô pital Notre-Dame Montreal) Elizabeth Rochon Laura Laird Fiona Höbler Lisa McQueen Ruth Tannenbaum Joanna Wong Amy Lewis and Alyssa Bobkin (Toronto Rehabilitation Institute Toronto) Sandra E. Black George Mochizuki Joyce L. Chen Valerie Closson Andrew Centen and Tyler Saumur (Sunnybrook Hospital Toronto)

¹McGill University, Jewish General Hospital, Montreal, Canada

²University of Toronto, Sunnybrook Research Institute, Toronto, Canada

³University of Toronto, Faculty of Kinesiology and Physical Education, Toronto, Canada

⁴Moss Rehabilitation Research Institute, Moss Rehabilitation Research Institute, Elkins Park, USA

⁵Burke Neurological Institute, Burke Neurological Institute, White Plains, USA

⁶Edith Cowan University, Edith Cowan University, Joondalup, Australia

⁷RehaNova, Rehabilitation Hospital, Cologne, Germany

⁸Max Planck Institute für Stoffwechsel Forschung, MPI for Metabolism Research, Cologne, Germany

⁹Universität zu Köln, University of Cologne, Cologne, Germany

¹⁰Université de Montréal, Hô pital du Sacré-Cœur de Montreal, Montréal, Canada

¹¹Université de Montréal, Hô pital Notre-Dame, Montreal, Canada

¹²McGill University, Department of Kinesiology and Physical Education, Montreal, Canada

¹³University of Toronto, Toronto Rehabilitation Institute, Toronto, Canada

¹⁴Burke Rehabilitation Hospital, Burke Rehabilitation Hospital, White Plains, USA

The clinical efficacy of non-invasive brain stimulation in combination with speech and language therapy (SLT) in post-stroke aphasia has been suggested in case series and small randomized controlled trials (RCT). This international multicenter proof-of-concept trial (Canada-US-Germany) compared the efficacy of inhibitory repetitive transcranial magnetic stimulation (rTMS), cathodal transcranial direct current stimulation (ctDCS) and sham stimulation on the right-hemispheric homologue of Broca’s area, as adjuvant therapy to SLT in sub-acute post-stroke aphasia.

Sixty-five aphasic patients within 45 days after ischemic stroke were randomly assigned to receive rTMS, ctDCS or sham stimulation in combination with SLT for 10 days. Standardized scores on the Boston Naming Test (BNT), the Token Test (TT) and the Semantic Fluency test (SF) were the primary language outcome measures at Baseline, and 1 and 30 days after the last treatment session. The primary intention-to-treat analysis tested for an interaction between time and intervention using 2-way mixed ANOVAs (3x3), with lesion location as co-variate.

Forty-nine participants completed the study. The preliminary analysis for main effect showed a statistically significant difference at the different time points but not between intervention groups. The interaction of time and intervention was not significant (BNT, F(3.203, 99.105) = .574, p = .664, partial η2 = .019, ε = .801; SF, F(3.396, 101.874) = .965, p = .420, partial η2 = .031, ε = .849; TT, F(3.364, 100.917) = .397, p = .777, partial η2 = .013, ε = .841).

This first multilingual and international RCT in post-stroke aphasia did not find a significant effect of rTMS or tDCS on primary language outcomes. Per-protocol and planned subgroup analyses are pending.

ClinicalTrials.gov-NCT02020421

CLINICAL AND BIOMECHANICAL EFFECTS OF TRANSCUTANEOUS ELECTRICAL STIMULATION ON CHRONIC POST-STROKE OROPHARYNGEAL DYSPHAGIA: RESULTS AT ONE YEAR FOLLOW UP OF A RANDOMIZED CONTROLLED TRIAL

V. Arreola¹, D. Alvarez-Berdugo², L. Rofes², O. Ortega², D. Muriana³, P. Clavé⁴

¹speech language therapist, Gastrointestinal Physiology Laboratory- Hospital de Mataró- Consorci Sanitari del Maresme., Barcelona, Spain

²Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas CIBERehd- Instituto de Salud Carlos III., Gastrointestinal Physiology Laboratory, Barcelona, Spain

³Neurology Unit- Hospital de Mataró- Consorci Sanitari del Maresme, Neurology Unit, Barcelona, Spain

⁴Gastrointestinal Physiology Laboratory- Hospital de Mataró- Consorci Sanitari del Maresme., Gastrointestinal Physiology Laboratory, Barcelona, Spain

Chronic post-stroke oropharyngeal dysphagia (OD) is associated with severely impaired pharyngeal sensory and motor function. We assessed the clinical outcome and biomechanical effect of sensory (SES) and motor (NMES) electrical stimulation on chronic post-stroke OD patients at one-year follow up.

Ninety post-stroke patients with chronic OD were randomly assigned to a) a control group with compensatory treatment including fluid thickening (CT), b) SES group: CT plus sensory electrical stimulation c) NMES group: CT plus neuromuscular electrical stimulation. Patients were treated with up to two cycles (6 months apart) of 15 sessions and followed up with up to five clinical and videofluoroscopic (VFS) assessments during 1 year.

Swallow function significantly improved in patients of the SES and NMES groups for both prevalence of safety impairment (p < 0.001), PAS (p < 0.001) and laryngeal vestibule closure time (p < 0.01) after the first treatment and at 1 year follow up. This improvement caused a significant reduction in the need for thickening agents (p < 0.001). At 1 year follow up, patients in the control group also presented reduced prevalence of signs of safety and efficacy impairment, however, the reduction was significantly smaller than that observed with SES and NMES and not related to any reduction in the laryngeal vestibule closure time. No major adverse events related to SES or NMES were observed

Transcutaneous electrical stimulation is a safe and effective therapy for post-stroke OD. After one year follow up, both SES and NMES greatly improved the safety of swallow and reduced the need for fluid thickening in these patients

N/A

EFFECTS OF A SENSORIMOTOR SHORT-TERM REHABILITATION PROGRAMME ON DYSPHAGIA AFTER CEREBRAL INFRACTION. A COMPARATIVE STUDY

C. Ghemulet Politi¹, S. Hamdy², I. Papathanasiou³ and E. Michou⁴

¹OLYMPION Rehab. Center of Patras, Speech Language Depart., Patra, Greece

²Faculty of Medical and Human Sciences- University of Manchester, Centre for Gastrointestinal Sciences- Institute of Inflammation and Repair, Manchester, United Kingdom

³TEI Western Greece- Patras, Department Speech Language Therapy, Patra, Greece

⁴TEI Western Greece- Patras- University of Manchester- UΚ, Department Speech Language Therapy- Honorary Lecturer- GI Sciences-, Patra, Greece

The aim of the study was to compare the effects of a short rehabilitation training programme on dysphagia after stroke between a sensorimotor therapeutic tool and traditional dysphagia therapy (TDT). The therapeutic protocol consisted of alternating exercises aiming to muscle strengthening (Chin Tuck Against Resistance, CTAR) and sensory stimulation of swallowing (Thermal Tactile Stimulation, TTS).

In this prospective study 30 eligible stroke patients with dysphagia were included. Two (2) groups were matched in terms of demographic characteristics (p>0,05). The experimental group (N = 15, 10 men, 11 with subcortical lesions, mean ( ± SD) age 63,9 ± 5.6, days post-stroke 10 ± 3.7) received the intense therapeutic protocol while the control group (N = 15) received TDT. Fiberoptic endoscopic evaluation of swallowing (FEES) was conducted at baseline and 2 weeks post, measuring standing pharyngeal secretions, spillage, penetration-aspiration, pharyngeal swallow onset time and residue. Dysphagia severity (Dysphagia Severity Rating Scale) and Functional Oral intake scale (FOIS) were also scored. Parametric statistical test Anova was used to compare the variables between the 2 groups (SPSS-v. 23).

Significant changes were observed in the experimental group: vallecula residue (measured with Yale residue scale) (p = 0.002), penetration/aspiration scores (p < 0.001), dysphagia severity (Dysphagia Severity Rating Scale) (p = 0.007 και p < 0.001). Delayed pharyngeal swallow was observed in 46.7% of patients from experimental group vs. 80% in control group.

This study demonstrated that sensorimotor treatment approach in the form of combined CTAR and TTS is effective in improving the pharyngeal swallow, reducing pharyngeal residue, and improving the swallowing function efficacy in post-stroke dysphagic patients.

N/A

TREATING ANXIETY AFTER STROKE (TASK): PROOF-OF-CONCEPT OF TELEMEDICINE COGNITIVE BEHAVIOURAL THERAPY (TASK-CBT) IN A STREAMLINED RANDOMISED CONTROLLED TRIAL

H.Y.Y. Chun¹, A. Carson¹, T. Tsanas², M. Dennis¹, G. Mead¹ and W. Whiteley¹

¹University of Edinburgh, Centre for Clinical Brain Sciences, Edinburgh, United Kingdom

²University of Edinburgh, Usher Institute of Population Health Sciences and Informatics Contact details, Edinburgh, United Kingdom

Anxiety is common and disabling after stroke. We designed TASK-CBT, a telemedicine intervention for anxiety after stroke with efficient recruitment and outcome methods.

To test the feasibility of i) an efficient clinical trial workflow, ii) the TASK-CBT intervention, and iii) smartwatch actigraphy in the trial

We recruited community-based stroke and transient ischaemic attack patients with remote screening, consent, and enrolment. Participants were randomly allocated to TASK-CBT (six weekly telephone sessions and a website) (n = 14) or TASK-Relax (online relaxation exercises) (n = 13). All participants were offered a smartwatch to wear. We collected generalized anxiety (GAD-7), phobic anxiety (Fear Questionnaire) and independence (modified Rankin Scale) via automated electronic questionnaires and actigraphy with a smartwatch at week six and 20 after enrolment.

All 27 participants were recruited remotely (mean age 65[+/−10]; 56%men; 63%stroke, 37%TIA). All 14 TASK-CBT participants completed all telephone sessions. There were significantly lower levels of anxiety in TASK-CBT compared to TASK-Relax at weeks 6 and 20 (Figure 1). Mean smartwatch wearing time was 33 days (+/-15). Further results will be presented. Duration and quality of sleep and activity are visualised (Figure 2). Further results will be presented.

Figure 1. A reduction of anxiety in TASK-CBT compared to TASK-Relax at weeks six and 20

Figure 2.

Post stroke anxiety was reduced with a remote CBT intervention, and efficient methods of recruitment and follow-up were feasible. Continuous actigraph data are potentially useful measures of stroke outcomes in a clinical trial. A larger definitive trial of remotely delivered CBT for post-stroke anxiety is planned.

Clinicaltrials.gov NCT03439813.

Scientific Communications 15-Vascular Cognitive Impairment

COGNITIVE PROFILE OF THALAMIC LACUNAR STROKES RESULTS OF THE SPS3 TRIAL

A. Honig¹, G. Alejandra², A.A. Sepehry³, G. Al-momen², S. Gosselin-Lefebvre², T. Field², S. Yip² and O. Benavente²

¹Hadassah-Hebrew University Medical center, Neurology, Jerusalem, Israel

²University of British Columbia, Neurology, Vancouver, Canada

³Adler University, Clinical Psychology, Vancouver, Canada

Despite their small size, lacunar strokes are associated with cognitive impairment, particularly those in strategic locations such as the thalamus. We aimed to characterized cognitive function in thalamic lacunar strokes.

Data from the Secondary Prevention of Small Subcortical Strokes (SPS3) trial. Participants were enrolled with a recent lacunar stroke verified by MRI. Detail neuropsychological assessment (with multi-domain tests) was conducted at study entry and yearly. Cognitive function at baseline was classified as: a) normal or with mild cognitive impairment (MCI), b) amnestic (aMCI) c) non-amnestic (naMCI) and multidomain (mdMCI).

Out 2911 SPS3 participants, 770 (26%) had a thalamic lacunar infarct as index event. Thalamic infarcts were most likely to be in women, Caucasian, and had lower association with preexisting hypertension and age-related white matter changes (ARWMC) score (P < 0.001). Of 2911 participants, 1254 (43%) with index events outside thalamus met criteria for MCI (12% aMCI, 17% naMCI, 14% mdMCI). Out of 770 thalamic lacunar strokes, 350 (45%) had MCI (14% aMCI, 14% naMCI, 17% mdMCI). MCI in thalamic lacunes was independently associated with black race (OR 4.1, 95% CI 1.8-9.1, p < 0.001), left side location (OR 1.6, 95% CI 1.1-2.2, p = 0.006), and multiple old infarcts on MRI (OR 1.6, 95% CI 1.1-2.2, p = 0.01). Patients with aMCI were more likely to have thalamic infarcts than those in other localizations (OR 1.5, 95% CI 1.1-2.1, p = 0.016).

Thalamic lacunar infarcts were independently associated with aMCI. MCI was more prevalent in those with left thalamic involvement and multiple old infarcts.

N/A

MODULATING EFFECT OF COGNITIVE RESERVE ON COGNITIVE IMPAIRMENT AND RECOVERY AFTER STROKE: THE KOSCO STUDY

Y.H. Kim¹, M. Shin¹, M.K. Sohn², J. Lee³, D.Y. Kim⁴, S.G. Lee⁵, Y.I. Shin⁶, G.J. Oh⁷, Y.S. Lee⁸, M.C. Joo⁹, S.Y. Lee¹⁰, J.H. Han¹¹, J. Ahn¹², W.H. Chang¹, M.A. Shin¹, J.Y. Choi¹³, S.H. Kang¹³, K.H. Lee¹³ and Y.T. Kim¹⁴

¹Samsung Medical Center, Physical and Rehabilitation Medicine, Seoul, Republic of Korea

²Chungnam National University School of Medicine, Department of Rehabilitation Medicine, Daejon, Republic of Korea

³Konkuk University School of Medicine, Department of Rehabilitation Medicine, Seoul, Republic of Korea

⁴Yonsei University College of Medicine, Department and Research Institute of Rehabilitation Medicine, Seoul, Republic of Korea

⁵Chonnam National University Medical School, Department of Physical and Rehabilitation Medicine, Gwangju, Republic of Korea

⁶Pusan National University Yangsan Hospital, Department of Rehabilitation Medicine, Pusan, Republic of Korea

⁷Wonkwang University School of Medicine, Department of Preventive Medicine, Jeollabuk-do, Republic of Korea

⁸Kyungpook National University Hospital, Department of Rehabilitation Medicine, Gyeongsangbuk-do, Republic of Korea

⁹Wonkwang University School of Medicine, Department of Rehabilitation Medicine, Jeollabuk-do, Republic of Korea

¹⁰Department of Rehabilitation Medicine, Jeju National University Hospital, Jeju, Republic of Korea

¹¹Hallym University, Department of Statistics, Gangwon-do, Republic of Korea

¹²Ewha Womans University, Department of Health Convergence, Seoul, Republic of Korea

¹³Korea Centers for Disease Control and Prevention, Division of Chronic Disease Prevention- Center for Disease, Chungcheongbuk-do, Republic of Korea

¹⁴Korea Centers for Disease Control and Prevention, Division of Chronic Disease Control- Center for Disease Prevention, Chungcheongbuk-do, Republic of Korea

A theory of cognitive reserve (CR) was introduced to account for individual differences in clinical manifestation of neuropathology. This study was an interim analysis of the Korean Stroke Cohort for Functioning and Rehabilitation (KOSCO) designed as 10-year long-term follow-up study of first-ever stroke patients. We investigated whether CR had moderating effect on cognitive impairment and recovery after stroke.

The 10,636 patients with first-ever stroke participated. Education, premorbid occupation, and composite CR score were used as proxies of CR. Performance on Korean Version of Mini-Mental State Examination (K-MMSE) was obtained at 7 day after stroke and at every 3–6 months until 30 months of follow up. Data from 7,459 patients who had K-MMSE score at least one time point during follow up period were included (Figure 1).

Level of education and occupation were negatively associated with the severity and frequency of cognitive impairment (K-MMSE < 16%ile). Predictive effect of CR was significant in logistic regression; low CR increased risk of cognitive impairment (Table 1). In multi-level model analysis, K-MMSE total score was increased until 3 months and then reached to plateau, but the slope of the increment was steeper in patients with higher levels of education or occupation (Figure 2).

Enriched lifetime experiences of an individual, including education and occupation, can buffer against cognitive impairment due to stroke, and promote rapid cognitive recovery after stroke.

Supported by a research program funded by the Korea Centers for Disease Control and Prevention (2016E-33003-02) and a NRF grant funded by the Korea government (NRF-2017R1A2A1A05000730 and NRF-2017M3A9G5083690).

NCT03402451

APOLIPOPROTEIN E GENOTYPE AND DEMENTIA ASSOCIATED WITH TRANSIENT ISCHAEMIC ATTACK AND STROKE: PROSPECTIVE POPULATION-BASED COHORT STUDY

S.T. Pendlebury¹, D. Poole¹, A.I. Burgess¹, J. Duerden¹ and P.M. Rothwell¹

¹Centre for Prevention of Stroke and Dementia, Nuffield Department of Clinical Neurosciences- University of Oxford, Oxford, United Kingdom

Apolipoprotein E4 (apoE4) is a genetic risk factor for Alzheimer’s disease and for reduced recovery from brain injury. Since data on apoE genotype and dementia associated with TIA/stroke are sparse, we determined the associations in a longitudinal population-based cohort.

All patients with TIA or stroke (2002–2012) in a defined population of 92 728 (Oxford Vascular Study) had follow-up to 5-years. Pre-event and incident post-event dementia were ascertained through direct patient assessment, supplemented by review of primary care records. Associations between pre- and post-event dementia and apoE status (homozygous (E4/E4), heterozygous (E4/-) and non-carrier) were examined using logistic regression and Cox regression models respectively, adjusted for age, sex, and education and also for vascular burden (event severity and white matter disease).

Among 1549 tested patients (mean/SD age = 72.7/13.0 years, 798/51.5% male, 535/34.5% TIA), 1168 (75%) were apoE4 negative, 354 (22.8%) were heterozygous, and 27 (1.7%) were homozygous. Although homozygous patients tended to be younger (mean/sd age = 70.7/12.4) than heterozygous or non-carrier patients (72.1/13.1 and 72.9/13.0 years), homozygosity was associated with both pre-event (OR = 4.63, 95% CI = 1.43-15.00, p = 0.01) and post-event (HR = 3.04, 1.49-6.19, p = 0.002) dementia with associations unaffected by further adjustment for vascular burden. Heterozygosity for apoE4 was weakly associated with pre-event dementia but not post-event dementia.

In the largest study to date in patients with TIA and stroke, apoE4 homozygosity was associated with both pre- and post-event dementia. The association appeared to be independent of vascular burden and may be mediated through increased neurodegenerative pathology.

N/A

VASCULAR AND NEURODEGERATIVE MARKERS FOR THE PREDICTION OF POST-STROKE COGNITIVE IMPAIRMENT – RESULTS FROM THE TABASCO STUDY

J. Molad¹, H. Hallevi¹, A. Korczyn², E. Kliper¹, E. Auriel³, N. Bornstein⁴, E. Ben Assayag¹

¹Tel Aviv Sourasky Medical Center, Neurology, Tel Aviv, Israel

²Tel-Aviv University, Neurology, Tel Aviv, Israel

³Rabin Medical Center, Neurology, Petach Tikva, Israel

⁴Shaarei Zedek Medical Center, Neurology, Jerusalem, Israel

Stroke is a major cause of cognitive impairment and dementia. We sought to understand the contribution of vascular pathology measures to post-stroke cognitive impairment (PSCI), separate from and in conjunction with pathologic changes associated with Alzheimer disease (AD).

Data from 397 cognitively intact first-ever ischemic stroke patients were available. All patients underwent White matter hyperintensity volume (WMHV) and integrity, ischemic lesions, brain, hippocampal and cerebro-fluid (CSF) volumes were measured at baseline using 3T MRI Along with comprehensive cognitive tests on admission and after two years. We used a multistage multiple regression approach to evaluate the contributions of vascular pathology measures (Framingham risk score for stroke, WMHV and existence of small vessel disease) and AD associated markers (apolipoprotein E4 status and hippocampal volume).

During two years follow-up period, 80 participants (20.2%) developed PSCI. Low gray matter and cortex volume and high WMHV and CSF predicted the development of PSCI in a dose-dependent relationship (P = 0.001). The newly detected lesion volume was not associated with PSCI. Vascular related imaging markers and risk factors predicted PSCI better than AD related markers (p < 0.001).

Brain structural measures, including total gray matter volume, cortex, WMHV and CSF volume were independently associated with PSCI and may serve as early biomarkers for risk prediction. In our sample vascular pathology measures contributed significantly better to prediction of PSCI than pathologic changes associated with AD. The newly detected ischemic lesion itself has not emerged as biomarker for PSCI risk, thus maybe a part of the ongoing vascular pathology.

NCT01926691

INCREASED BETWEEN-VISIT BLOOD PRESSURE VARIABILITY IN ACUTE STROKE PREDICTS WORSE COGNITIVE SCORES AT 90 DAYS

J. Appleton¹, L.J. Woodhouse¹, Z.K. Law¹, N. Sprigg¹, J.M. Wardlaw², P.M. Bath¹; for the ENOS investigators

¹University of Nottingham, Stroke- Division of Clinical Neuroscience, Nottingham, United Kingdom

²University of Edinburgh, Centre for Clinical Brain Sciences, Edinburgh, United Kingdom

Increased between-visit blood pressure variability (BPV) is associated with poor clinical outcomes after acute stroke, but there is limited data on the effect of BPV on cognition. We assessed the effect of between-visit BPV in the first days after acute stroke on cognitive outcomes at day 90 using data from the Efficacy of Nitric Oxide in Stroke (ENOS) trial.

ENOS trialrandomised 4011 patients with acute stroke and raised systolic BP (140-220 mmHg) to transdermal GTN or no GTN within 48 hours of onset. Between-visit BPV was defined as the standard deviation of systolic BP over days 1 to 7 and split into equal quintiles. Cognition was assessed by blinded assessors at day 90 as telephone mini-mental state examination (t-MMSE), modified telephone interview for cognition scale (TICS-M) and verbal fluency. Data are mean difference (MD) with 95% confidence intervals (CI). Analyses were adjusted for baseline prognostic factors.

The highest quintile of between-visit BPV over days 1 to 7 was associated with worse cognitive scores at day 90 compared with the lowest quintile: t-MMSE MD −2.03, 95% CI −2.84 to −1.22, p < 0.001; TICS-M MD -2.68, 95% CI −3.84 to −1.51, p < 0.001; verbal fluency MD −1.87, 95% CI −2.73 to −1.00, p < 0.001. These associations were maintained in a sensitivity analysis excluding participants with atrial fibrillation.

Increased between-visit BPV in the first days after acute stroke was associated with worse cognitive scores at day 90. Further research is needed to establish whether BPV is a modifiable target in acute stroke.

ISRCTN99414122

PREDICTORS OF LONG-TERM RISK OF DELIRIUM IN PATIENTS WITH PREVIOUS TIA AND STROKE: PROSPECTIVE POPULATION-BASED COHORT STUDY

S.T. Pendlebury¹, S.J. Welch¹, A.M. Haigh¹, R.J. Thomson¹ and P.M. Rothwell¹

¹Centre for Prevention of Stroke and Dementia, Nuffield Department of Clinical Neurosciences- University of Oxford, Oxford, United Kingdom

History of stroke is associated with delirium in general hospital and cardiac surgery cohorts, but there are few prospective studies of long-term risk of delirium in patients who have undergone detailed phenotyping of TIA and stroke events. We determined the predictors of delirium during general hospital admissions on long-term follow-up after TIA and stroke.

In TIA and stroke patients recruited from 2002–2012 to the population-based Oxford Vascular Study (OXVASC), admissions for any reason to the regional hospital or to the local community hospital during Oct 2013 to May 2015 were identified. Patients were assessed prospectively and delirium diagnoses identified. Using OXVASC baseline data on stroke severity, white matter disease, vascular risk factors and MMSE/MoCA scores together with admission diagnosis, we determined predictors of delirium using logistic regression adjusted for age and sex.

Among TIA/stroke patients initially recruited, 1565 (mean/SD age 68.9/13.3, 751 female, 676 TIA) were still alive on 01 October 2013. Of 139 subsequent hospital admissions in 123 patients (mean/SD age 77.6/11.7 years, 57 female), 122 (88%) were unplanned. Delirium was associated with age (OR = 1.08/year 95% CI = 1.05-1.33, p < 0.0001), illness severity (1.61,1.17-2.21, p = 0.003), dementia (8.70,4.0-19.0, p < 0.0001), prior cognitive impairment (MoCA – 0.86,0.80-0.94, p < 0.0001; MMSE – 0.88,0.81-0.96, p = 0.003), depression (2.15,1.0-4.6, p = 0.05), and prior stroke severity (NIHSS – OR/point = 1.09,1.01-1.17, p = 0.03).

Markers of cerebrovascular disease burden and pre-existing cognitive impairment predispose to long-term delirium risk possibly through increasing cerebral susceptibility to systemic factors.

N/A

ASSOCIATION BETWEEN STRIATAL BRAIN IRON DEPOSITION, MICROBLEEDS AND COGNITION OVER 12 MONTHS AFTER MINOR ISCHAEMIC STROKE

M. Valdés Hernández¹, T. Case², F. Chappell¹, S. Makin³, P. Armitage⁴, A. Glatz¹, F. Doubal¹ and J. Wardlaw¹

¹University of Edinburgh, Neuroimaging Sciences, Edinburgh, United Kingdom

²University of Edinburgh, College of Medicine and Veterinary Medicine, Edinburgh, United Kingdom

³Cardiovascular Research Centre, Institute of Cardiovascular and Medical Sciences, Glasgow, United Kingdom

⁴University of Sheffield, Infection- Immunity and Cardiovascular Sciences, Edinburgh, United Kingdom

Brain iron deposits (IDs) are inversely associated with cognitive function in community-dwelling older people, but their association with cognition after ischaemic stroke, and whether that differs from microbleeds, is unknown.

We quantified basal ganglia IDs (BGID) and microbleeds (BMBs) semi-automatically on brain magnetic resonance images from patients with minor stroke (NIHSS < 7), at presentation and 12 months after stroke. We administered the National Adult Reading Test (NART, estimates premorbid or peak adult cognition) and the Revised Addenbrooke’s Cognitive Examination (ACE-R; current cognition) at 1 and 12 months after stroke. We adjusted analyses for baseline cognition, age, gender, white matter hyperintensity (WMH) volume and vascular risk factors.

In 200 patients, mean age 65 years, striatal IDs and BMBs volumes did not change over the 12 months. Baseline BGID volumes correlated positively with NART scores at both times (ρ = 0.19, p < 0.01); baseline BMB volume (% in ICV) was associated negatively only with the 12 month NART scores (β = -0.205, p = 0.012). Baseline and follow-up BGID volumes correlated positively with age (ρ = 0.248, p < 0.001 and ρ = 0.271, p < 0.001 respectively), but only baseline BMB volume correlated with age (ρ = 0.129, p < 0.05). Both Smoking and baseline WMH burden predicted verbal fluency and visuospatial abilities scores (B = -1.13, p < 0.02 and B = -0.22, p = 0.001 respectively) at 12 months after stroke.

BGIDs and BMBs are associated differently with cognition post-stroke; studies of imaging and post-stroke cognition should adjust for premorbid cognition. The positive association of BGID with NART is likely to reflect the lower premorbid cognition in patients with stroke at younger vs older ages.

N/A

CORTICAL ATROPHY IS ASSOCIATED WITH STRENGTH OF CONNECTION TO WHITE MATTER HYPERINTENSITIES

C. Mayer¹, B.M. Frey¹, E. Schlemm¹, M. Petersen¹, J. Fiehler², C. Gerloff¹, B. Cheng¹ and G. Thomalla¹

¹University Medical Center Hamburg-Eppendorf, Department of Neurology, Hamburg, Germany

²University Medical Center Hamburg-Eppendorf, Department of Neuroradiology, Hamburg, Germany

White matter hyperintensities (WMH) resulting from cerebral small vessel disease are known to be associated with cortical atrophy. However, investigations of structural factors linking vascular WMH with the topography of cortical thinning are scarce. We therefore investigated whether pattern and extent of cortical thickness is associated with WMH by mediation of fiber tract connection strength in a population at risk of cerebrovascular disease.

We analyzed 930 MRI data sets of participants from a population-based single-center cohort study. Cortical thickness was measured on T1-weighted imaging in 210 distinct brain regions. WMH were segmented on FLAIR images. Whole-brain probabilistic tractography based on diffusion tensor imaging data was used to define connection strength of WMH to cortical regions. The effect of relative connectivity to WMH on cortical thickness was analyzed by a linear mixed-effects model. Known factors associated with cortical thickness were included as fixed effects (age, sex, hemisphere) and random effects (region and individual level of cortical thickness) were added.

Mean age was 62.8 ± 8.1 years. Visual inspection of cortical surface maps revealed distinct connectivity patterns of cortical regions to WMH (see Figure). The connectivity of cortical areas to WMH was significantly associated with reduced cortical thickness (p = .003).

Our results indicate a specific effect of WHM on cortical thickness mediated by connecting fiber tracts. These results point to a mechanism of specific neurodegeneration in brain regions distant, yet connected to subcortical lesion.

N/A

Scientific Communications 16-Endovascular Stroke Treatment – Who will Benefit?

FRAME: FRENCH ACUTE CEREBRAL MULTIMODAL IMAGING TO SELECT PATIENT FOR MECHANICAL THROMBECTOMY

J.M. Olivot¹, J.F. Albucher¹, I. Sibon², M. Mlynash³, A. Viguier¹, T. Tourdias⁴, F. Bonneville⁵, N. Raposo¹, L. Calviere¹, A. Guenego⁶, A. Drif⁷, S. Christensen³, C. Thalamas⁷, A. Sommet⁷, V. Rousseau⁷, M. Mazighi⁸, A.C. Januel⁶, P. Menegon⁴, G. Albers³, C. Cognard⁶; on behalf of the FRAME Investigators

¹Hopital Toulouse Purpan, Acute Stroke Unit- Toulouse Neuro Imaging Center- Clinical Investigation center, Toulouse, France

²Hopital Bordeaux Pellegrin, Neurology, Bordeaux, France

³Stanford University, Stanford Stroke Center, Stanford, USA

⁴Hopital Bordeaux Pellegrin, Diagnostic and Interventionnal Neuroradiology, Bordeaux, France

⁵Hopital Toulouse Purpan, Diagnostic and Interventionnal Neuroradiology- Toulouse Neuro Imaging Center, Toulouse, France

⁶Hopital Toulouse Purpan, Diagnostic and Interventionnal Neuroradiology, Toulouse, France

⁷Hopital Toulouse Purpan, Clinical Investigation center, Toulouse, France

⁸Lariboisière Hospital- AP-HP- Paris7-Denis Diderot University- Sorbonne Paris Cité, Neurology and Stroke Centre- France, Paris, France

Recent meta-analyses suggest that patients with a large infarct core volumes may benefit from mechanical thrombectomy (MT) within 6 hours after stroke onset. The presence of target mismatch (TMM) on multimodal imaging appears to be a good surrogate for salvageable penumbra. Whether patients without imaging evidence of salvageable penumbra on multimodal imaging before MT benefit from thrombectomy is unknown. We aim to investigate in a prospective multicentric cohort study the relationship between the success of MT performed within 6 hours after onset, the rate of good functional outcome at 3 months and the presence or absence of TMM on pretreatment multimodal imaging.

Prospective Multicentric Cohort Study. Consecutive patients eligible for MT within 6 hrs after onset will systematically undergo multimodal imaging (CTP or MRI) before treatment and will be treated according to the current recommendations. Treating physicians are blinded to perfusion maps and core volume measurement at the time of the treatment decision. Automatically processed maps using RAPID software are sent to a study server and will be assessed by investigators blinded of the success of thrombectomy and clinical outcomes. TMM is defined according to the EXTEND-IA study criteria.

Population: 220 consecutive patients evaluated and treated at investigating centers for thrombectomy. (enrollment to be finished end of January)

Outcome: The rate of good functional outcome defined by a modified Rankin Scale of 0–2, 3 months after treatment and the success of MT (TICI IIb-III) will be compared with the prevalence of a TMM on baseline MRI/CTP.

Ongoing

NA

NCT03045146

IMAGING SELECTION CRITERIA AND TREATMENT EFFECT OF ENDOVASCULAR THROMBECTOMY IN ENDOVASCULAR TREATMENT TRIALS

K.W. Muir¹, P. White², B.S. Brown³, B.C.V. Campbell⁴, P. Mitchell⁵, A.M. Demchuk⁶, M.D. Hill⁶, T.G. Jovin⁷, A. Dávalos⁸, F. Guillemin⁹, S. Bracard¹⁰, J.L. Saver¹¹, G. Albers¹², C. Majoie¹³, D.W.J. Dippel¹⁴, M. Goyal⁶; on behaf of the HERMES Collaborators

¹University of Glasgow, Institute of Neuroscience & Psychology, Glasgow, United Kingdom

²University of Newcastle, Institute of Neuroscience, Newcastle, United Kingdom

³BRIGHT Research Partners, Statistics, Minneapolis, USA

⁴University of Melbourne, Department of Medicine and Neurology, Melbourne, Australia

⁵University of Melbourne, Department of Radiology, Melbourne, Australia

⁶University of Calgary, Department of Clinical Neurosciences, Calgary, Canada

⁷University of Pittsburgh Medical Center, Department of Neurology, Pittsburgh, USA

⁸Universitat Autònoma de Barcelona, Department of Neuroscience, Barcelona, Spain

⁹Centre Hospitalier Régional et Universitaire de Nancy, Inserm, Nancy, France

¹⁰University of Lorraine and University Hospital of Nancy, Department of Diagnostic and Interventional Neuroradiology, Nancy, France

¹¹University of California- Los Angeles, Department of Neurology, Los Angeles, USA

¹²Stanford Stroke Center, Department of Neurology, Palo Alto, USA

¹³Amsterdam University Medical Center, Department of Radiology and Nuclear Medicine, Amsterdam, The Netherlands

¹⁴Erasmus University Medical Center, Department of Neurology, Rotterdam, The Netherlands

Endovascular thrombectomy (EVT) trials within 12h of onset selected patients using different imaging methods. Four trials (MR CLEAN, REVASCAT, THRACE, PISTE) required only brain and angiographic imaging; three trials required additional favourable penumbral perfusion (EXTEND-IA, SWIFT Prime) and one trial additional favourable computed tomography (CTA) collaterals (ESCAPE). We explored the relationship between imaging selection criteria and treatment effect in these trials.

Individual data from seven trials were analysed, comparing simple imaging (non-invasive angiography selection (CTA/MRA) with physiological imaging (CT perfusion or multiphase CTA). All analyses were adjusted for age, sex, NIHSS, occlusion location, alteplase administration, time from onset to randomization, and baseline ASPECTS. We compared the adjusted odds ratios of modified Rankin Scale (mRS) 0–2, mRS distribution, mortality up to day 90 and SICH incidence at 24 hours.

Physiological imaging criteria were used in 601 participants (312 assigned EVT and 289 assigned control) and simple imaging alone in 1163 participants (559 EVT and 604 control). Favourable outcome (mRS 0–2 at day 90) with physiological imaging-selection was seen in 57.2% EVT versus 33.0% control and with simple imaging-selection 42.5% EVT versus 29.4% control alone (absolute treatment difference 24.0% versus 13.1%). Physiological imaging selection favourably modified the treatment effects on mortality (p = 0.016), day 90 mRS 0–2 (p = 0.010) and mRS distribution (p = 0.001).

EVT improved outcome in both simple and physiological imaging selected patients. After adjustment for baseline prognostic factors, selection by physiological imaging was associated with significantly improved odds of independent recovery, reduced disability and lower mortality compared to angiographic selection alone.

N/A

ARTIFICIAL INTELLIGENCE BASED DETECTION OF LARGE VESSEL OCCLUSIONS ON CT ANGIOGRAPHY IN STROKE

M. Koopman¹, M.L. Tolhuisen², E. Ponomareva³, H. Lin³, R. Sales Barros³, I.G.H. Jansen³, A.M.M. Boers³, Y.B.W.E.M. Roos⁴, H.A. Marquering² and C.B.L.M. Majoie¹

¹Amsterdam UMC- location AMC, Department of Radiology and Nuclear Medicine, Amsterdam, The Netherlands

²Amsterdam UMC- location AMC, Department of Biomedical Engineering and Physics, Amsterdam, The Netherlands

³Nico.lab, www.nico-lab.com, Amsterdam, The Netherlands

⁴Amsterdam UMC- location AMC, Department of Neurology, Amsterdam, The Netherlands

Rapid and accurate detection of a large vessel occlusion (LVO) in acute ischemic stroke may be aided by artificial intelligence (AI) to accelerate therapeutic decision making. We aimed to evaluate the use of an AI-based algorithm for automated detection of LVO on computed tomography angiography (CTA).

In this retrospective study, we included patients with LVO (carotid artery terminus and middle cerebral artery segments M1 and M2) and stroke mimics from a tertiary stroke center. Ground truth was established by consensus between two separate observers with both >5y of experience. LVO was detected by an algorithm based on convolutional neural networks developed by Nico.lab (www.nico-lab.com). Its accuracy was compared to that of another expert observer (>5y experience), both blinded to all other imaging and clinical parameters. Detection was considered correct if the delineated region overlapped with the ground truth. From this overlap, sensitivity and specificity were calculated.

A total of 100 patients were analyzed, of which 50 had proven LVOs (ICA-T n = 12;M1 n = 25; M2 n = 13) and 50 were stroke mimics. The accuracy of the algorithm was 89% vs. 94% of human observer. Sensitivity and specificity were 0.84 vs. 0.88 and 0.93 vs. 1.00 for the algorithm vs. human observer, respectively (Figure 1). Maximum running time of the algorithm for LVO detection was 0.1 seconds per patient.

AI-based LVO detection on CTA showed comparable accuracy as an expert observer. This supportive tool could lead to a faster and more objective detection of LVO in acute stroke.

N/A

GREATER BENEFIT FROM MECHANICAL THROMBECTOMY IN FAST PROGRESSORS

X. Urra¹, A. Renú¹, C. Laredo¹, S. Rudilosso¹, L. Llull¹, V. Obach¹, S. Amaro¹ and A. Chamorro¹

¹Hospital Clinic, Institut Clinic de Neurociències, Barcelona, Spain

The speed of progression of the ischemic lesion is variable in stroke patients. We tested the hypothesis that the tissue saving effect of mechanical thrombectomy over best medical treatment (BMT) may be more significant in fast progressors.

Single-center cohort of consecutive patients (n = 285) with proximal occlusions in the carotid territory, known time from onset, and full imaging available (baseline CTP and follow-up MRI). The infarct growth rate was calculated with the non-viable tissue defined in CBF maps and the time to imaging, and patients were categorized as fast/slow progressors according to the median growth rate of 4.1cc/h. The primary outcome measure was the interaction between the progression rate category and MT on infarct growth. The secondary outcomes assessed the treatment effect of MT over BMT on final infarct volume and functional status in fast and slow progressors. The safety outcomes were mortality and symptomatic intracranial hemorrhage.

The progression rate category had a modifying effect of MT on infarct growth (p = 0.015), that was significantly reduced only in fast progressors. There was also a significant interaction on final infarct volume (p < 0.007), with a greater reduction after MT in fast progressors (25.7cc with MT vs. 68.9cc with BMT, p = 0.022). The functional status improved with MT (with no significant interaction) both in fast and slow progressors, and there were also no significant interactions regarding mortality and symptomatic hemorrhage.

The use of MT in large vessel stroke results in a more substantial limitation of infarct growth in fast progressor patients.

N/A

RATE AND BENEFIT OF 'FIRST-PASS EFFECT' IN THE ASPIRATION VERSUS STENTRETRIEVER (ASTER TRIAL)

C. Ducroux¹, B. Lapergue², J. Labreuche³, M. Piotin¹, R. Blanc¹, M. Ben Maacha¹, B. Gory⁴, R. Fahed¹; The ASTER collaborators

¹Fondation Rothschild Hospital, Interventional Neuroradiology, Paris, France

²Foch Hospital, Neurovascular Unit, Suresnes, France

³Lille Hospital, Department of Biostatistics, Lille, France

⁴CHRU de Nancy, Department of Neuroradiology, Nancy, France

In the context of thrombectomy, the “First-Pass Effect” (FPE), which was originally described with stentretrievers, designates a (near-) complete revascularization obtained after a single device pass with no rescue therapy, and is associated with an improved clinical outcome and a decreased mortality. We report the rate and benefits of FPE in the Aspiration versus Stentretriever (ASTER) trial.

ASTER is a randomized trial comparing stentretriever (SR) technique with contact aspiration (CA), in terms of angiographic revascularization, assessed with the modified thrombolysis in cerebral infarction (mTICI) scale by an external core laboratory. Patients with FPE (defined by a mTICI = 2c/3 after a single pass with no rescue therapy) were compared with the other patients in terms of efficacy and safety outcomes.

A FPE was achieved in 97/336 patients (28.9%, 95% CI, 24.0% to 33.7%), with no significant difference between SR and CA (53/169 patients [31.3%] versus 44/167 patients [26.3%], p = 0.31). After multivariate analysis, FPE was associated with a significantly improved clinical outcome and a decreased mortality. Moreover, FPE was also associated with a significantly decreased rate of hemorrhagic transformation and procedural complications.

In the ASTER trial, First-Pass Effect was achieved in similar rates with Stentretriever and Contact Aspiration, and was associated with a significantly improved outcome. New techniques and devices to improve the rate of FPE are warranted.

N/A

2B OR 2C: WHAT SHOULD BE THE TARGET FOR ENDOVASCULAR TREATMENT?

N. LeCouffe¹, M. Kappelhof², K. Treurniet², G. Zhang³, B. Emmer², C. Majoie², Y. Roos¹, J. Coutinho¹; on behalf of the MR CLEAN Registry Investigators

¹Amsterdam UMC- location AMC, Neurology, Amsterdam, The Netherlands

²Amsterdam UMC- location AMC, Radiology & Nuclear Medicine, Amsterdam, The Netherlands

³The First Affiliated Hospital of Harbin Medical University, Neurosurgery, Harbin, China

eTICI 2B or more is the current target angiographic endpoint for endovascular treatment (EVT) of ischemic stroke due to a large vessel occlusion. We examined whether eTICI 2C (90-99% reperfusion) leads to better clinical outcomes than eTICI 2B in daily clinical practice.

We included patients from the MR CLEAN Registry who had both anteroposterior and lateral runs of last DSA. All images were assessed by a core laboratory. We compared the distribution of the modified Rankin Scale (mRS) score per eTICI grade, with specific focus on comparing patients with eTICI 2C vs 2B.

Out of 1526 patients, 1310 met imaging criteria. Compared to patients with eTICI 2B (n = 276), patients with eTICI 2C (n = 138) were more often male (60% vs 50%, p = 0.06) and more often underwent general anesthesia during EVT (42% vs 30%, p = 0.04). Rate of IV rtPA prior to EVT did not differ between groups. Functional outcome improved per eTICI grade (Figure 1). However, after adjustments, mRS-shift analysis showed no statistically significant difference in outcome at 90 days between patients with eTICI 2C and 2B (acOR 1.19, 95% CI 0.81–1.75, mRS 0–2 47% vs 41%). Likewise, mortality did not differ significantly between eTICI 2C and 2B.

Clinical outcomes did not clearly differ between patients with eTICI 2C and 2B.

N/A

LESS IS MORE: THE DETRIMENTAL EFFECT OF DEVICE-PASSES ON OUTCOME IN ACUTE ISCHEMIC STROKE SECONDARY TO LARGE VESSEL OCCLUSION

A. Garcia-Tornel¹, M. Rubiera¹, N. Rodriguez¹, S. Boned¹, D. Rodriguez-Luna¹, J. Pagola¹, J. Juega¹, M. Muchada¹, D. Hernandez², A. Tomasello², M. Requena¹, M. Olivé¹, C.A. Molina¹ and M. Ribo¹

¹Hospital Universitari Vall d’Hebron, Stroke Unit, Barcelona, Spain

²Hospital Universitari Vall d’Hebron, Neurorradiology department, Barcelona, Spain

Substantial proportion of patients who achieve successful recanalization of acute ischemic stroke due to large vessel occlusion do not achieve good functional outcome. We aim to analyze the effect of number of thrombectomy device passes and degree of the recanalization (by modified Thrombolysis in Cerebral Infarction (mTICI)) on the clinical and functional outcome

Five-hundred and forty-five consecutive patients underwent mechanical thrombectomy for acute large vessel occlusion in the anterior circulation at a single tertiary stroke center. Baseline characteristics, number of passes, degree of recanalization, clinical outcome at 24 hours (measured by National Institutes of Health Scale score (NIHSS)) and functional outcome (measured by modified Rankin scale (mRs) at 90 days) were registered. Multivariate analysis was performed to determine the association of number of passes and degree of recanalization with dramatical clinical recovery (final NIHSS < = 2 or decrease in 8 or more NIHSS points in 24hours) and good functional outcome (mRs < = 2 at 90 days).

Four-hundred and fifty-nine patients (84%) achieved successful recanalization (mTICI 2B-3), 213(39%) of them after first device pass. In the multivariate analysis, first-pass recanalization and mTICI 3 were independent predictors of good functional outcome (OR2.5, CI 1.4-4.5, p = 0.002 and OR2.6 CI 1.5-4.7, P = 0.001, respectively) and dramatical clinical recovery (OR1.8, CI 1.1-3, p = 0.032 and OR2.9, CI 1.7-5.1, p < 0.001, respectively). A linear association between number of passes and good functional outcome was observed: 1 pass (58.6%), 2 passes (50.5%), 3 passes (48.4%), 4 passes (38.5%) or 5 or more passes (25.6%)(p < 0.001).

Number of passes and degree of recanalization are both independent predictors of favorable outcome in stroke patients undergoing endovascular treatment who achieve recanalization

N/A

THROMBECTOMY RESISTANT ISCHAEMIC STROKE: THE MATRIS STUDY

P. Alcantara Miranda¹, S. Trillo¹, C. Gómez-Escalonilla², C. Matute³, J. Egido², J. Martínez Poles⁴, C. Ramos Martín¹, C. Díaz Pérez¹, P. Simal², A. De Felipe⁴, M. Moreu⁵, R. Vera⁴, S. Rosati⁵, A. Cruz-Culebras⁴, Á. Ximénez-Carrillo¹, J. Caniego⁶, E. Bárcena⁶, J. Masjuan⁴ and J. Vivancos¹

¹Hospital La Princesa, Neurology, MADRID, Spain

²Hospital Clínico San Carlos, Neurology, Madrid, Spain

³Hospital Universitario Ramón y Cajal, Neurosciences, Madrid, Spain

⁴Hospital Universitario Ramón y Cajal, Neurology, Madrid, Spain

⁵Hospital Clínico San Carlos, Radiology, Madrid, Spain

⁶Hospital La Princesa, Neurosciences, Madrid, Spain

Mechanical thrombectomy (MT) in the treatment of large vessel occlusion MCA stroke is not always successful despite multiple attempts, leaving an important amount of Thrombectomy Resistant Ischaemic Stroke (TRIS) patients behind. The goal of the MAdrid TRIS (MATRIS) is to identify the number of attempts that best describe these patients, their clinical profile and the features associated with their final recanalization (TICI>2a).

Multicentric retrospective study with clinical and radiological data collection from January-2016 to December-2017 in three different stroke centres.

N = 376. Up to 80% of successful recanalization was observed within 4 attempts, but this percentage decreases with successive attempts being insignificant over 6(<2%), thus, patients who underwent >4 attempts were considered as TRIS, n = 48(13%). Univariate analysis found significant differences (p < 0.005) between TRISp and non-TRISp in prior antiplatelet drug uptake (37,5%vs 23,5%), post-MT H1-H2 cerebral hemorrhage rates (45,8%vs 26,5%), and functional independence at 3m (19%vs59%). 44% of TRISp recanalized successfully. Significant differences were also found between recanalized and non-recanalized TRISp in smoking rates (5%vs37%), serum creatinine levels (median = 0,76[0,3] vs 0,96[0,36]), previous thrombolysis (29% vs 60%), intraprocedural anticoagulation therapy administration (14% vs 0%) and functional outcome at 3 months (38% vs 3%).

The number of attempts that best describes TRIS patients (TRISp) is >4, with a very low probability of recanalization over 6. TRISp were more frequently on antiplatelet therapy and had worse outcome at 3 months than non-TRISp. Successful recanalization in TRISp was associated with non-smokers, lower creatinine levels, no previous thrombolysis and intraprocedural anticoagulation. Final recanalization in TRISp was associated with better outcome compared to non-recanalized TRISp.

N/A

INTRACRANIAL ARTERIAL GEOMETRY BASED RISK FACTORS OF UNSUCCESSFUL ENDOVASCULAR TREATMENT OF ACUTE ISCHEMIC STROKE

P. Konduri^1,2, K. Treurniet^2,3, O. Berkhemer^2,4,5, A. Yoo⁶, G. Lycklama³, Y. Roos⁷, R. van Oostenbrugge⁸, M. Kappelhof^1,2, A. van der Lugt⁵, J. Boiten⁹, H. Marquering^1,2, C. Majoie²; On behalf of MRCLEAN Registry investigators

¹Amsterdam University Medical Centre – Location AMC, Biomedical Engineering and Physics, Amsterdam, The Netherlands

²Amsterdam University Medical Centre – Location AMC, Department of Radiology and Nuclear Medicine, Amsterdam, The Netherlands

³Haaglanden Medisch Centrum, Department of Radiology, The Hague, The Netherlands

⁴Erasmus MC University Medical Centre, Department of Neurology, Rotterdam, The Netherlands

⁵Erasmus MC University Medical Centre, Department of Radiology, Rotterdam, The Netherlands

⁶Texas Stroke Institute, Department of Radiology, Dallas, USA

⁷Amsterdam University Medical Centre – Location AMC, Department of Neurology, Amsterdam, The Netherlands

⁸Maastricht University Medical Centre, Department of Neurology, Maastricht, The Netherlands

⁹Haaglanden Medisch Centrum, Department of Neurology, The Hague, The Netherlands

Identifying intracranial arterial geometry based risk factors that can lead to challenging/unsuccessful endovascular treatment (EVT) can aid in patient selection and provide recommendations to improve stent design.

Four-hundred-eight patients were retrospectively selected based on availability of good quality thin-slice NCCT and CTA from the MRCLEAN-Registry, a nationwide EVT-database. Anterior intracranial vasculature was segmented using a convolutional neural network developed by Nico.lab (www.nico-lab.com). Artery labelling and geometry analysis were performed with a semi-automated tool, iCAFE (Figure). Proximal arterial length, average radius, tortuosity and angle of bifurcation between the parent and proximal artery in the ipsilateral side were determined. Unsuccessful treatment was defined as < 50% reperfusion.

In this ongoing study, we analysed 15 patients with distal M1 occlusions. Thirty-three percent of the patients had an unsuccessful treatment. Median length, radius, tortuosity, and angle of bifurcation between the ipsilateral ICA and the proximal M1 segment were 20 mm (IQR: 17–21mm), 2.1 mm (IQR: 1.9- 2.4 mm), 1.1 (IQR: 1.05-1.3), 124° (77° – 135°) in patients with unsuccessful treatment and 15 mm (IQR: 13–17mm), 2.0mm (IQR: 1.9- 2.1 mm), 1.1 (IQR: 1.0-1.2), 149° (97° – 163°) in patients with successful treatment respectively.

Patients with unsuccessful treatment had longer proximal M1 segment and smaller angle of bifurcation. We have set up an efficient pipeline for the semi-automatic analysis of the geometry of intracranial arteries.

N/A

NEW ANTI-DIABETIC DRUGS AND RISK OF STROKE: A SYSTEMATIC REVIEW AND META-ANALYSIS OF RANDOMISED CONTROLLED TRIALS

H. Azhari¹, C. Haig² and J. Dawson³

¹University of Glasgow- UK. – Amm Al-Qura University- SA., Institute of Cardiovascular and Medical Science College of Medical- Veterinary & Life Science, Glasgow, United Kingdom

²Institute of Health and Wellbeing College of Medical- Veterinary and Life Sciences, Robertson Centre for Biostatistics University of Glasgow Boyd Orr Building- Level 11 University of Glasgow Glasgow G12 8QQ, Glasgow, United Kingdom

³Institute of Cardiovascular and Medical Sciences College of Medical- Veterinary & Life Sciences University of Glasgow Room M0.05 Office Block Queen Elizabeth University Hospital Glasgow G51 4TF, Professor of Stroke Medicine Consultant Physician Clinical Lead Scottish Stroke Research Network/NRS Stroke Research Champion Chair MVLS Research Ethics Committee, Glasgow, United Kingdom

Diabetes is associated with increased stroke risk. There are several new classes of anti-diabetic drugs. This review aimed to assess the effects of these drugs on stroke risk.

A systematic search of databases was conducted to identify relevant studies. These included the Cochrane Register of Controlled Trials, MEDLINE, EMBASE, and Web of Science (from their inception to 28˗week, 2017). Published and unpublished trials comparing DPP4-Is, GLP1-As, SGLT2-Is with placebo or other AHGs were included. The protocol was published in PROSPERO-CRD 42017067889.

From 960 titles, 249 studies were identified. These included 61,932 people treated with DPP4-Is, 19,923 with GLP1-As, 29,713 with SGLT2-Is, 85,489 treated with comparator agents and 6,538 stroke outcomes. Twenty-four studies had high–risk of bias, 30 unclear, and 195 low–risk with high GRADE quality of evidence. In trials of DPP4-Is, (92 trials, 91,790 participants, 1,019 cases of stroke) there was no reduction in stroke vs. control (RR 0.92; 95% CI 0.81–1.03; P = 0.16; I2 = 0%). In trials of GLP1-As, (29 trials, 33,144 participants, 846 cases of stroke) there was a trend towards a reduction in stroke (RR 0.88; 95% CI 0.77–1.01; P = 0.06; I2 = 0%). In trials of SGLT2-Is, (45 trials, 35,789 participants, 373 cases of stroke) there was no reduction in stroke (RR 1.02; 95% CI 0.83–1.26; P = 0.81; I2 = 0%).

The GLP1-As may reduce the risk of stroke compared to comparator agents but no evidence of a reduction in stroke risk was seen with other agents. These findings may help physicians guide treatment for type 2 diabetes after stroke.

N/A

HIGH ON-CLOPIDOGREL PLATELET REACTIVITY IN STROKE: A SYSTEMATIC REVIEW AND META-ANALYSIS

V. Alakbarzade¹, X. Huang¹, S. Drury², I. Chis Ster³, M. McEntagert² and A.C. Pereira¹

¹St George’s University Hospitals NHS Foundation Trust, Neurology, London, United Kingdom

²St George’s University of London, Neuroscience, London, United Kingdom

³St George’s University of London, Statistics, London, United Kingdom

Emerging studies highlight clopidogrel resistance or high on-clopidogrel platelet reactivity (HCPR) as a hindrance to prevention of ischaemic stroke or TIA (S/TIA). The aim of this study was to assess the prevalence of HCPR, recurrence risk and mechanism of action in S/TIA patients.

Studies were searched and identified using PUBMED (from inception to Dec 25, 2018) and were included if absolute numbers/percentages of HCRP were reported (Fig_1). 10 studies were included, a total of 1599 S/TIA patients.

The pooled prevalence of HCPR was 26% (95% CI: 18–33.6%). However, the prevalence rates reported between studies were heterogenous (Fig_2). There was a higher S/TIA recurrence rate in clopidogrel resistant patients (RR = 3.8, 95% CI:1.6–9.3; p = 0.003; heterogeneity across studies was low: I2 = 33.6%, Cochran Q: p = 0.2) (Fig_3). From the genotyping data, a significant proportion of clopidogrel resistant S/TIA patients were CYP2C19 reduced-function allele carriers (p = 0.02).

This systematic review shows a high prevalence of clopidogrel resistance in S/TIA and increased S/TIA recurrence rate in these patients. CYP2C19polymorphisms significantly influence clopidogrel resistance.

N/A

EFFECT OF COMBINED ASPIRIN AND CLOPIDOGREL VERSUS ASPIRIN ALONE ON FUNCTIONAL OUTCOME AND RECURRENCE IN ACUTE ISCHEMIC STROKE PATIENTS

M. Ouyang¹, L. Song¹, L. Billot², M.L. Hackett³, P.M. Lavados⁴, V.V. Olavarría⁴, P. Muñoz-Venturelli⁵, S. Middleton⁶, O.M. Pontes-Neto⁷, T.H. Lee⁸, C. Watkins⁹, T. Robinson¹⁰ and C.S. Anderson¹

¹The George Institute China at Peking University Health Science Centre, Stroke Division, Beijing, China

²The George Institute for Global Health, Statistic Division, Sydney, Australia

³The George Institute for Global Health, Neurology & Mental Health Division, Sydney, Australia

⁴Clinica Alemana Universidad del Desarrollo, Department of Neurology and Psychiatry- Clínica Alemana de Santiago- Facultad de Medicina, Santiago, Chile

⁵Clínica Alemana Universidad del Desarrollo, Centro de Estudios Clínicos- Instituto de Ciencias e Innovación en Medicina- Facultad de Medicina, Santiago, Chile

⁶Australian Catholic University, Nursing Research Institute- St Vincent’s Health Sydney Australia, Sydney, Australia

⁷University of Sao Paulo, Department of Neuroscience and Behavioral Sciences- Ribeirão Preto Medical School, Sao Paulo, Brazil

⁸Chang Gung University, Stroke Center and Department of Neurology- Linkou Chang Gung Memorial Hospital and College of Medicine, Taoyuan, Taiwan R.O.C

⁹University of Central Lancashire, Faculty of Health and Wellbeing, Preston, United Kingdom

¹⁰University of Leicester, Department of Cardiovascular Sciences and NIHR Leicester Biomedical Research Centre, Leicester, United Kingdom

Evidence indicates benefits of that combination (aspirin+clopidogrel) antiplatelet therapy (APT) improves outcome in minor acute ischemic stroke (AIS) and TIA. We aimed to determine effects of combination and single APT use among AIS patients across different levels of stroke severity.

Post-hoc analyses of 9467 AIS patients randomised to lying-flat vs sitting-up head positioning in an international multi-centre, cluster cross-over, randomised controlled trial. Multilevel binomial logistic regression analysis was used to determine associations of dual APT and clinical outcomes of death or disability (modified Rankin Scale 3–6) and recurrence of stroke events within 90-days.

Overall, 4272 (45%) patients had APT at baseline, of whom 27% (1148/4272) were on combination aspirin and clopidogrel. Combined APT was more frequent in younger patients with mild stroke, and lower systolic blood pressure. Recurrent stroke and poor outcome did not differ between treatment groups (adjusted odds ratio [aOR] 0.87, 95% CI 0.55-1.37 and aOR 0.89, 95% CI 0.72-1.09, respectively). In patients under dual APT, the rate of death or dependency of mild stroke (National Institutes of Health Stroke Scale[NIHSS] < 5) was lower compared to more severe patients (NIHSS≥5) (30.5% vs. 69.5%, p < 0.0001). There is a significant interaction between treatment and stroke severity on functional outcome (p < 0.0001). Minor stroke patients on combination APT had significantly decreased risks of 90-day death or dependency compared to patients on aspirin alone (aOR 0.70, 95% CI 0.51-0.95).

AIS patients with minor stroke on combined APT had a decreased risk of poor outcome compared to patients on Aspirin alone.

N/A

EFFICACY AND SAFETY OF RIVAROXABAN VERSUS ASPIRIN IN PATIENTS WITH EMBOLIC STROKE OF UNDETERMINED SOURCE AND CAROTID ATHEROSCLEROSIS: AN EXPLORATORY SUBGROUP ANALYSIS OF THE NAVIGATE-ESUS

G. Ntaios¹, B. Swaminathan², S.D. Berkowitz³, R.J. Gagliardi⁴, W. Lang⁵, J.E. Siegler⁶, P. Lavados⁷, H. Mundl⁸, N.M. Bornstein⁹, E. Meseguer¹⁰, P. Amarenco¹¹, B. Cucchiara⁶, P. Camps-Renom¹², K. Makaritsis¹, E. Korompoki¹³, J. Marti-Fabregas¹², H. Milionis¹⁴, K. Vemmos¹, S.J. Connolly¹⁵, R.G. Hart¹⁶; The NAVIGATE ESUS Investigators

¹University of Thessaly, Department of Internal Medicine, Larissa, Greece

²Population Health Research Institute, Hamilton, Ontario, Canada

³Bayer U.S. LLC, Pharmaceuticals Clinical Development Thrombosis, New Jersey, USA

⁴Irmandade, da Santa Casa de Misericórdia de São Paulo, Sao Paulo, Brazil

⁵Sigmund Freud Private University- Medical Faculty-, Hospital St. John of God-, Vienna, Austria

⁶Hospital of the University of Pennsylvania, Department of Neurology, Pennsylvania, USA

⁷Clínica Alemana, de Santiago, Santiago, Chile

⁸Bayer AG, Bayer AG, Wuppertal, Germany

⁹Shaare-Zedek, Medical Center, Jerusalem, Israel

¹⁰Bichat Hospital, Neurology Service, Paris, France

¹¹Sorbonne Paris Cité university, APHP- Bichat hospital- Paris-Diderot, Paris, France

¹²Hospital de la Santa Creu i Sant Pau, Unitat de Malalties Vasculars Cerebrals- Servei de Neurologia-, Barcelona, Spain

¹³Imperial College London- Charing Cross Hospital, Division of Brain Science, London, United Kingdom

¹⁴Faculty of Medicine University of Ioannina, School of Health Sciences, Ioannina, Greece

¹⁵McMaster University/Population Health Research Institute, Department of Medicine Cardiology, Hamilton- Ontario, Canada

¹⁶McMaster University/Population Health Research Institute, Department of Medicine Neurology, Hamilton- Ontario, Canada

In this exploratory subgroup analysis of patients with carotid atherosclerosis in NAVIGATE-ESUS, we assessed whether the treatment effect in this subgroup is consistent with the overall trial population and investigated the association of carotid atherosclerosis with recurrent ischemic stroke.

Carotid atherosclerosis was analyzed separately as either the presence of mild (i.e. 20–49%) atherosclerotic stenosis or the presence of carotid plaque. The primary efficacy outcome was ischemic stroke recurrence. The safety outcomes were major bleeding and symptomatic intracerebral bleeding.

Carotid plaque was present in 40% of participants, and mild carotid stenosis in 11%. There was no significant difference in ischemic stroke recurrence rate between rivaroxaban- and aspirin-treated patients among 490 patients with carotid stenosis (5.0 versus 5.9/100 patient-years respectively, HR:0.85, 95% CI:0.39-1.87; p-value for interaction with patients without 0.78) and among 2905 patients with carotid plaques (5.9 versus 4.9/100 patient-years respectively, HR:1.20, 95% CI:0.86-1.68; p-value for interaction with patients without 0.2). Among patients with carotid plaque, major bleeding was more frequent in rivaroxaban-treated patients compared with aspirin-treated (2.0 versus 0.5/100patient-years, HR:3.75, 95% CI:1.63-8.65).

Patients with carotid stenosis had similar rate of ischemic stroke recurrence compared to those without (5.4 versus 4.9/100 patient-years respectively, HR:1.11, 95% CI:0.73-1.69), but there was a strong trend of higher rate of ischemic stroke recurrence in patients with carotid plaque compared to those without (5.4 versus 4.3/100patient-years respectively, HR:1.23, 95% CI:0.99-1.54).

We found no difference in efficacy between rivaroxaban and aspirin for prevention of recurrent stroke in ESUS patients with carotid atherosclerosis. Aspirin was safer, consistent with the overall trial results.

NCT02313909

THE EFFECT OF ASPIRIN ON ISCHEMIC STROKE SUB-TYPES IN THE HEALTHY ELDERLY

G. Cloud¹, G. Donnan², J. Williamson³, C. Eaton⁴, D. Brauer⁵, J. Lockery⁶, M. Nelson⁷, R. Wolfe⁶, A. Murray⁸, R. Shah⁹, J. Ryan⁶, R. Woods⁶, J. McNeil⁶; On behalf of the ASPREE Investigators

¹Alfred Health, Department of Neurology, Melbourne, Australia

²Florey Institute of Neuroscience and Mental Health, Department of Neurology, Melbourne, Australia

³Wake Forest Baptist Medical Center, Geriatric Medicine and Gerontology, Winston-Salem, USA

⁴Brown University, Center for Primary Care and Prevention, Providence- Rhode Island, USA

⁵Minnesota State University, School of Nursing, Minnesota, USA

⁶School of Public Health and Preventive Medicine- Monash University, Department of Epidemiology and Preventive Medicine, Melbourne, Australia

⁷Menzies Institute for Medical Research- University of Tasmania, General Practice School of Medicine, Hobart, Australia

⁸Hennepin Health Research Institute- University of Minnesota, Berman Center for Clinical Research, Minneapolis, USA

⁹Rush University, Department of Family Medicine- Rush Medical College, Chicago, USA

Aspirin is established for secondary prevention in ischemic stroke. However, the benefit of aspirin in primary stroke prevention is unclear. The Aspirin in Reducing Events in the Elderly (ASPREE) trial recently published the results of a randomized placebo controlled trial of a daily low dose of aspirin in 19,114 healthy people over the age of 65. Here we specifically consider the effect of aspirin on different sub-groups of ischemic stroke within ASPREE.

All suspected stroke events within ASPREE were independently and blindly adjudicated by two members of an expert committee. Disagreements were adjudicated by a third expert and consensus reached. Stroke sub-types were attributed based on available diagnostic reports from hospitals.

After a median of 4.7 years follow up in 19,114 people recruited, there were 302 first ischemic stroke events.

Table of time to first ischemic stroke sub-type compared between aspirin and placebo randomized groups, summarized as rates per 1000 person years

Table

There was no significant benefit in aspirin over placebo for any ischemic stroke sub group. Previously published ASPREE results demonstrated increased rates of hemorrhagic stroke with aspirin compared to placebo. Aspirin cannot be recommended for the primary prevention of stroke in healthy older adults.

ISRCTN83772183

THE INCREASE IN ANTICOAGULATION PRESCRIBING AND THE IMPACT ON NATIONAL STROKE RATES

R. Doyle¹, J. McCormack², R. Collins³, P. Marsden⁴, S. Clarke⁵, M. Barry⁵, P. Kelly⁶ and J. Harbison⁷

¹St Vincent’s University Hospital- St Columcille’s Hospital and University College Dublin, Medicine for the Elderly, Dublin, Ireland

²Royal College of Physicians of Ireland, National Clinical Programme for Stroke, Dublin, Ireland

³Royal College of Physicians of Ireland and Tallaght University Hospital, National Clinical Programme for Stroke, Dublin, Ireland

⁴Health Service Executive, Department of Public Health, Dublin, Ireland

⁵St. James’s Hospital and Trinity College Dublin, Department of Pharmacoeconomics, Dublin, Ireland

⁶Health Research Board and Mater Misericordiae hospital, Stroke Clinical Trials Network, Dublin, Ireland

⁷St. James’s Hospital and Trinity College Dublin, Mercer’s Institute for Successful Ageing, Dublin, Ireland

The number of people receiving anticoagulation more than doubled (33073 – 69256) from January (2009 – 2018) in the Republic of Ireland.

The aims were to look at 1. the increase in anticoagulation prescribing and its impact on ischaemic and haemorrhagic strokes and 2. the rate of anticoagulation in the over 80’s.

Hospital In-Patients Enquiry (HIPE) data was collected on ischaemic (ICD 163) and haemorrhagic strokes (161) admitted to public hospitals (2009-2017) inclusive and compared with national medicines prescribing anticoagulation data. This was compared with baseline incidence period 2004–2009.

The 209% increase in anticoagulation prescribing revealed a shift from warfarin (99% (2009) to 28% (2018)) to DOAC’s (72%) in 2018. Baseline average of ischaemic stroke increased 3.6% p.a. (4174-4781 (14.5%)) in the base line period but only 0.5% from 2009–2016. Rates returned to 2009 levels in 2017 with 4760 ischaemic strokes compared to the predicted 6500. The rate of haemorrhagic stroke fell 2.1% p.a in the baseline period (1374-1256(8.6%)) but increased 1.2% on average from 2009–2016(8.1%). This fell to 858 in 2017 below 2004 figures.

Trials of DOAC’s included < 20% of patients aged over 80. Our prescribing data show 35%-45% of people receiving anticoagulation > 80 years. In 2017, 34% of ischaemic and 32% of haemorrhagic strokes occurred in > 80’s.

The increase in anticoagulation prescribing was associated with slowing of ischaemic stroke rates.

Both ischaemic and haemorrhagic strokes decreased in 2017 despite an ageing population

Over 80’s traditionally underrepresented in clinical trials account for 35%-45% of anticoagulation prescribing.

N/A

DOES PRIOR ANTICOAGULATION RELATE TO STROKE SUB-TYPE AND OUTCOME IN PATIENTS WITH ATRIAL FIBRILLATION IN SCOTLAND? A NATIONAL DATABASE STUDY

M.J. Macleod¹, K.A. Sterling¹, P. Langhorne² and M. Turner³

¹University of Aberdeen, Institute of Medical Sciences, Aberdeen, United Kingdom

²University of Glasgow, Department of Cardiovascular and Clinical Medicine, Glasgow, United Kingdom

³Institute of Applied Health, Aberdeen, United Kingdom

Atrial fibrillation (AF) increases the risk of stroke fivefold. Anticoagulation therapy reduces the stroke risk in this population. This study aims to examine the pre-stroke use of anticoagulants in stroke patients with AF, and the subsequent outcomes after the index stroke.

A novel linked dataset of the Scottish Stroke Care Audit, Prescribing Information System and the Scottish Morbidity Records was used to identify retrospectively stroke patients with AF prescribed oral anticoagulants (OAC) between January 2010 and December 2015. Statistical analyses included chi-square tests for proportions and logistic regression to determine odds ratios.

There were 11,442 patients with confirmed AF upon admission to hospital for the index stroke. Prior treatment with OAC included Warfarin 2545 (22.2%), Dabigatran etexilate 21 (0.2%), Apixaban 78 (0.7%) and Rivaroxaban 126 (1.1%). For recorded stroke type, 1971 (83.7%) AF patients on OAC experienced an ischaemic stroke, 348 (14.8%) experienced a Haemorrhagic stroke, 25 (1.1%) had a Haemorrhagic transformation of infarct and 10 (0.4%) were unclassified. OAC patients were more likely to have a Haemorrhagic stroke (adjusted OR: 3.874, CI: 3.314 – 4.528; P < 0.001) compared to patients not on an OAC. There was no difference in the chance of discharge home by 30 days (aOR: 0.888, CI 0.786 – 1.003; P = 0.055). There was an increased 30-day mortality with Warfarin (aOR: 1.421, CI: 1.221 – 1.654; P < 0.001), not seen with the use of Direct OACs.

Pre-stroke use of OAC was associated with more haemorrhagic strokes and increased 30-day mortality, mainly due to warfarin use.

N/A

DABIGATRAN INITIATION IN PATIENTS WITH NON-VALVULAR ATRIAL FIBRILLATION (NVAF) AND ACUTE ISCHAEMIC STROKE: AN ANALYSIS FROM THE SAFE IMPLEMENTATION OF TREATMENTS IN STROKE (SITS) REGISTRY

M. Mazya^1,2, C. Teutsch³, Z. Gdovinova⁴, L. Barbarini⁵, E. Filip⁶, W. Fryze⁷, M. Karlinski⁸, A. Kobayashi⁹, G. Krastev¹⁰, A.P. Nunes¹¹, K. Pasztoova¹², A. Peeters¹³, P. Sobolewski^14,15, A. Vilionskis^16,17, D. Toni¹⁸ and N. Ahmed^1,2

¹Karolinska University Hospital, Department of Neurology, Stockholm, Sweden

²Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden

³Boehringer Ingelheim International GmbH, Medical Department- TA Cardiology and Metabolics, Ingelheim, Germany

⁴Pavol Jozef Šafárik University in Košice, Department of Neurology- Faculty of Medicine, Košice, Slovak Republic

⁵Department of Neurology-, Ospedale Vito Fazzi, Lecce, Italy

⁶St. Queen Jadwiga Clinical Regional Hospital no. 2, Stroke Unit, Rzeszow, Poland

⁷M. Kopernik Hospital, Copernicus Pl, Gdańsk, Poland

⁸Institute of Psychiatry and Neurology, Second Department of Neurology, Warsaw, Poland

⁹University of Technology and Humanities, Faculty of Medical Sciences and Physical Education, Radom, Poland

¹⁰Faculty Hospital Trnava, Department of Neurology, Trnava, Slovak Republic

¹¹Centro Hospitalar Lisboa Central, Stroke Unit, Lisbon, Portugal

¹²Forlife General Hospital, Department of Neurology, Komarno, Slovak Republic

¹³Cliniques Universitaires St-Luc, Department of Neurology, Brussels, Belgium

¹⁴Jan Kochanowski University, The Faculty of Medicine and Health Sciences, Kielce, Poland

¹⁵Holy Spirit Specialist Hospital in Sandomierz, Department of Neurology and Stroke Unit, Sandomierz, Poland

¹⁶Republican Vilnius University Hospital, Department of Neurology, Vilnius, Lithuania

¹⁷Vilnius University, Institute of Clinical Medicine, Vilnius, Lithuania

¹⁸Sapienza University of Rome, Department of Human Neurosciences, Rome, Italy

The optimal time range for initiation of dabigatran after acute ischemic stroke is not established. We aimed to evaluate the timing and clinical outcomes of dabigatran initiation in NVAF patients hospitalized after acute ischemic stroke.

This observational study analyzed SITS Registry data (NCT03258645; July 2014 to July 2018) from European NVAF patients (≥18 years) hospitalized after first-ever ischemic stroke (index event), receiving dabigatran within 3 months. Primary outcome: time from index event to dabigatran initiation. Additional outcomes included physicians’ reasons for delaying dabigatran initiation beyond acute hospital discharge, and clinical outcomes within 3 months of index event.

The analysis included 1240 patients from 18 European countries. At baseline (hospital admission), median age was 76 years; 53% of patients were female, 15% were receiving an oral anticoagulant, 29% acetylsalicylic acid, and 4% clopidogrel. Most patients (72%) initiated dabigatran 2–14 days after the index event (Figure). Patients initiating earlier had lower stroke severity: from median NIHSS 8 (IQR 6–13) if initiated within 7 days, to NIHSS 15 (9-19) if initiated between 28 days and 3 months. Most common reasons for delaying initiation were hemorrhagic transformation or intracranial hemorrhage, stroke severity, infarct size. Few thrombotic/hemorrhagic events occurred within 3 months post–index event: 2,5%, (20/926 with available data).

In clinical practice, dabigatran was most commonly initiated within 2 weeks of hospitalization for first-ever acute ischemic stroke in NVAF patients. The main reasons for delayed initiation were hemorrhagic transformation and stroke severity/infarct size. Few thrombotic/hemorrhagic events occurred within 3 months.

N/A

ETIOLOGY OF ISCHEMIC STROKE IN PATIENTS WITH ATRIAL FIBRILLATION DESPITE ORAL ANTICOAGULATION

J. Purrucker¹, K. Hölscher¹, J. Kollmer² and P.A. Ringleb¹

¹Heidelberg University Hospital, Department of Neurology, Heidelberg, Germany

²Heidelberg University Hospital, Department of Neuroradiology, Heidelberg, Germany

Patients with atrial fibrillation suffer ischemic stroke at a rate of 0.9 to 1.8% per year despite anticoagulation with vitamin K antagonists (VKA) or non-VKA oral anticoagulants (NOAC). The etiology of strokes despite oral anticoagulation is unknown.

Within all acute ischemic stroke cases treated from January 2016 to June 2018 at the Heidelberg University Hospital, patients with known atrial fibrillation and oral anticoagulation were included. Potential underlying causative diseases other than cardio-embolism were classified according to the ASCOD (atherosclerosis; small-vessel disease; cardiac pathology; other causes) phenotyping. Patients with known NOAC cessation > 3 days before admission were excluded.

Of 2615 patients, 301 (11.5%) with known atrial fibrillation and oral anticoagulation were identified. More patients received NOACs than VKA (n = 196, 65.1% vs. n = 105, 34.9%, p < 0.001). NOAC concentration assays were performed in 175 of the NOAC patients (89.3%). At admission, 52/175 patients (29.7%) had levels < 10 ng/ml. Dosing errors within 3 days before admission were reported in 51 (26%) of the NOAC patients. Subtherapeutic INR values (<2.0) were observed in 70 (66.7%) of the VKA patients. While atherothrombotic stroke was frequent (19.6%), small-vessel-disease was only graded as potentially causative in 3.3%. No differences between VKA and NOAC patients was observed.

Insufficient oral anticoagulation according to coagulation parameters at admission occurred more often in VKA than NOAC patients. However, dosing errors were frequent in NOAC receiving patients. One-fifth of the patients suffered from atherothrombosis as a potentially secondary cause despite atrial fibrillation, presenting a challenge to secondary prevention.

NA

ISCHEMIC STROKE DESPITE ORAL ANTICOAGULANT THERAPY IN PATIENTS WITH AF: RECURRENCE RISK AND OPTIMAL ANTICOAGULATION STRATEGY

D. Seiffge^1,2,3, G.M. De Marchis¹, M. Koga⁴, M. Paciaroni⁵, D. Wilson², M. Cappellari⁶, K. Macha⁷, G. Tsivgoulis^8,9, B. Kallmünzer⁷, S. Yaghi¹⁰, C. Traenka¹, P. Lyrer¹, B. Bonetti⁶, S. Arihiro⁴, S. Engelter¹, D. Werring²; on behalf of rhe RAF/RAF-NOAC CROMIS-2 SAMURAI NOACISP Erlangen and Verona registry investigators

¹University Hospital Basel, Neurology and Stroke Center, Basel, Switzerland

²UCL Queen Square Institute of Neurology, Department of Brain Repair and Rehabilitation- Stroke Research Center, London, United Kingdom

³University Hospital Berne, Department of Neurology and Stroke Center, Berne, Switzerland

⁴National Cerebral and Cardiovascular Center, Department of Cerebrovascular Medicine, Suita, Japan

⁵University of Perugia, Stroke Unit and Division of Cardiovascular Medicine, Perugia, Italy

⁶Azienda Ospedaliera Universitaria Integrata, Stroke Unit – Department of Neuroscience, Verona, Italy

⁷University of Erlangen-Nueremberg, Department of Neurology, Erlangen, Germany

⁸National & Kapodistrian University of Athens School of Medicine- “Attikon” University Hospital, Second Department of Neurology, Athens, Greece

⁹University of Tennessee Health Science Center, Department of Neurology, Memphis, USA

¹⁰The Warren Alpert Medical School of Brown University, Department of Neurlogy, Rhode Island, USA

It is unknown whether patients with atrial fibrillation (AF) with ischemic stroke despite oral anticoagulant therapy are at increased risk for further recurrent strokes and how ongoing secondary prevention should be managed.

We conducted a pooled individual patient data analysis of 7 prospective cohort studies recruiting patients with AF and an index event (ischemic stroke or TIA). We compared patients taking oral anticoagulants (Vitamin K antagonists [VKA] or direct oral anticoagulants [DOAC]) prior to index event (OAC_prior) with those without prior anticoagulation (OAC_naive). We further compared those who changed the type (i.e. from VKA or DOAC, vice versa or DOAC to DOAC) of anticoagulation (OAC_changed) with those who continued the same anticoagulation as secondary prevention (OAC_unchanged). Time to recurrent ischemic stroke (AIS) was analysed using multivariate cox proportional hazard regression models with a study frailty term.

We included 5413patients (median age 78years[IQR71-84years], 5136[96.7%] had ischemic stroke as the index event (median NIHSS-on-admission 6[IQR 2–12]). The median CHA2DS2-Vasc score was 5(IQR4-6) and not different between OACprior (n = 1195) and OACnaive (n = 4119, p = 0.103). During 6128patient-years of follow-up,289 patients had AIS(4.7%/year,95% CI 4.2-5.3%). OACprior was associated with an increased risk of AIS(HR1.6,95% CI 1.1-2.1, p = 0.006). OACchanged (n = 307) was not associated with decreased risk of AIS(HR1.3,95% CI 0.8-2.2, p = 0.326) compared with OACunchanged (n = 585).

Patients with AF who have a stroke despite oral anticoagulation once are at a higher risk for further ischemic strokes despite a similar CHA2DS2-Vasc score to those without prior oral anticoagulation. Switching the type of OAC after treatment failure was not associated with a decreased risk. Better prevention strategies are needed for this high risk patient group.

n/a

BASELINE VASCULAR OCCLUSION MARKEDLY ENRICH THE PREDICTING MODEL FOR 30-DAY DETECTION OF ATRIAL FIBRILLATION

J. Pagola¹, J. Juega¹, A. Bustamante², E. Pala², M. De Lera³, J.F. Arenillas³, R. De Torres⁴, J. Montaner⁴, J.A. Cabezas⁵, F. Moniche⁵, J. Francisco⁶, T. Gonzalez Alujas⁷, C.A. Molina¹; CryptoAF investigators

¹Vall d’Hebrón hospital, Stroke Unit Vall d’ Hebron Hospital, Barcelona, Spain

²Vall d’Hebron Research Institute, Neurovascular Lab, Barcelona, Spain

³Valladolid Hospital, Stroke Unit. Neurology department, Valladolid, Spain

⁴Virgen Macarena Hospital, Stroke Unit. Neurology department, Sevilla, Spain

⁵Virgen del Rocio Hospital, Stroke Unit. Neurology department, Sevilla, Spain

⁶Vall d’Hebrón hospital, Arrhythmia Unit. Cardiology department, Barcelona, Spain

⁷Vall d’Hebrón hospital, Echocardiography Lab. Cardiology department, Barcelona, Spain

Crypto-AF was a prospective multicenter study of cryptogenic stroke patients for detecting atrial fibrillation (AF) based on acute clinical data, stroke imaging features and atrial dysfunction markers.

We included cryptogenic stroke patients after standard systematic work up from Jan 2015 to Jan 2017. All patients underwent 30-day ECG Holter monitoring starting within 72 hours of symptoms onset. We collected demographic data, stroke imaging characteristics, ECG analysis (P-wave terminal force V1 PTFV₁), left atria size and functionality (longitudinal strain analysis) and blood biomarkers (NTPBNP, BNP). We developed a model and regression tree analysis for AF risk stratification. Patients were followed up for at least 2 years for vascular events (recurrent stroke, peripheral embolism or myocardial infarction).

A total of 264 patients of 320 patients included fulfilled the 30-day monitoring. The multivariate regression model including (age, NIHSSS, CHADSVASC score, atrial volume, vascular occlusion and strain analysis) confirmed that age (OR 1.11), NIHSSS (OR 0.92), vascular occlusion (OR 9.51), and strain analysis OR 0.92 were independent predictors of AF detection (ROC curve 0.84). Regression tree analysis showed that the risk of AF was increased in more than 2 fold in patients with baseline vessel occlusion (OR 2.37). Adding markers of atrial dysfunction (LAVI>28.5, PTFV1 >3900) to the model further increase the predictive accuracy OR 4.96 to detect AF. On the other hand, only CHADSVASC score emerged as independent predictor for vascular events (OR 2.1).

The detection of baseline vascular occlusion improved the accuracy of the AF risk model based on atrial dysfunction.

N/A

HEART RATE AS A PREDICTOR FOR STROKE RECURRENCE IN ATRIAL FIBRILLATION-RELATED ACUTE ISCHEMIC STROKE

K.J. Lee¹, B.J. Kim¹, J.T. Kim², D.I. Shin³, J.K. Cha⁴, D.E. Kim⁵, J.M. Park⁶, S.J. Lee⁷, M.S. Oh⁸, B.C. Lee⁸, K.S. Hong⁹, Y.J. Cho⁹, J.C. Choi¹⁰, S.I. Sohn¹¹, T.H. Park¹², W.J. Kim¹³, J. Lee¹⁴, J.S. Lee¹⁵, J. Lee¹⁶, H.J. Bae¹; on the behalf of CRCS-K investigators

¹Seoul National University Bundang Hospital, Neurology, Seongnam, Republic of Korea

²Chonnam National University Hospital, Neurology, Gwangju, Republic of Korea

³Chungbuk National University Hospital, Neurology, Cheongju, Republic of Korea

⁴Dong-A University Hospital, Neurology, Busan, Republic of Korea

⁵Dongguk University Ilsan Hospital, Neurology, Goyang, Republic of Korea

⁶Eulji General Hospital, Neurology, Seoul, Republic of Korea

⁷Eulji University Hospital, Neurology, Daejeon, Republic of Korea

⁸Hallym University Sacred Heart Hospital, Neurology, Anyang, Republic of Korea

⁹Inje University Ilsan Paik Hospital, Neurology, Goyang, Republic of Korea

¹⁰Jeju National University Hospital, Neurology, Jeju, Republic of Korea

¹¹Keimyung University Dongsan Medical Center, Neurology, Daegu, Republic of Korea

¹²Seoul Medical Center, Neurology, Seoul, Republic of Korea

¹³Ulsan University Hospital, Neurology, Ulsan, Republic of Korea

¹⁴Yeungnam University Medical Center, Neurology, Daegu, Republic of Korea

¹⁵Asan Medical Center, Clinical Research Center, Seoul, Republic of Korea

¹⁶Korea University, Biostatistics, Seoul, Republic of Korea

Rate control is well established as a management for patients with atrial fibrillation (AF). However, association between heart rate and recurrent stroke events has not been discovered in patients with AF-related ischemic stroke.

From a multicenter prospective registry of stroke patients, acute ischemic stroke patients with AF who admitted within 48 hours of stroke onset were included. Heart rates during the first 24 hours after admission were analyzed to investigate the effect of heart rate in acute period, and heart rates in subacute period, from 72 hours after onset to discharge, were further analyzed in a subgroup who had available data. Primary outcome was stroke recurrence which was prospectively captured up to 1 year after stroke onset.

A total of 2,046 patients were analyzed for the effect of acute heart rate and 1,597 patients were further analyzed for the effect of subacute heart rate. There were 102 (5.0%) stroke recurrences, 9 (0.4%) myocardial infarctions and 440 (21.5%) deaths within 1 year after stroke onset. Proportional hazards regression models were constructed and the non-linearity of the effect of mean heart rate was examined. While mean heart rate in acute period was not associated outcome measures, mean heart rate in subacute period had a “U-shape” relationship with stroke recurrence (p for quadratic effect = 0.012).

This study suggests mean heart rate during the subacute period of ischemic stroke in patients with AF is a predictor for stroke recurrence and optimal rate control might be a potential target for secondary stroke prevention.

N/A

AUTOMATIC ANALYSIS OF PROLONGED HOLTER-ECG DATA TO IDENTIFY PAROXYSMAL ATRIAL FIBRILLATION IN ACUTE ISCHEMIC STROKE PATIENTS: READY TO DISPLACE PHYSICIANS?

T. Uphaus¹, B. Lange², M. Grond³, M. Jauss⁴, S. Gröschel¹, P. Kirchhof⁵, T. Rostock², R. Wachter⁶ and K. Gröschel¹

¹University Medical Center of the Johannes Gutenberg University Mainz, Department of Neurology, Mainz, Germany

²University Medical Center of the Johannes Gutenberg University, Cardiology, Mainz, Germany

³Kreisklinikum Siegen, Neurology, Siegen, Germany

⁴Hainich Klinikum, Neurology, Mühlhausen, Germany

⁵University of Birmingham, Institute of Cardiovascular Sciences, Birmingham, United Kingdom

⁶University Hospital Leipzig, Cardiology, Leipzig, Germany

In order to prevent recurrent cerebral ischemia the detection of paroxysmal atrial fibrillation (pAF) is of major clinical importance. Prolonged monitoring times are capable to increase the diagnostic yield of pAF but analysis of these data is resource-consuming. We investigated the effectiveness of pAF-detection with an automated software algorithm (SA) in comparison to the routine staff-based analysis (RA) of 72h-Holter-monitoring after acute ischemic stroke.

The prospective multicenter IDEAS-Trial applied 72h-Holter-ECG in acute stroke patients presenting with sinusrhythm. Assessment of the Holter-data was performed either RA-based by an experienced cardiologist or SA-based by a commercially available software (SRAclinic®, Apoplex Medical Technologies). In the case of discordant results concerning the occurrence of pAF between the two methods, the raw-data underwent a blinded rating for reference diagnosis.

pAF could be finally diagnosed in 5.9%(n = 61) of the patient cohort (n = 1034). SA more frequently correctly diagnosed pAF(n = 59,96.7%) compared to RA(n = 45,73.8%). SA misclassified atrial arrhythmias as pAF in 40 patients, whereas RA classified 2 patients as false positive. SA reached a higher sensitivity (SA 96.7%, vs. RA 73.8%) and RA a better specificity (SA 95.9% vs. RA 99.8%) for the detection of pAF. The agreement between the two methods classified by Kappa coefficient was moderate (0.592, p < 0.001).

SA-based evaluation of Holter-ECG increases the diagnostic yield of pAF with the drawback of more false-positive findings. This technique could be used for a resource-saving analysis of increasing ECG-data, however a final rating by an experience physician is still needed and may be omitted in cases of a negative SA analysis.

N/A

RISK OF STROKE RECURRENCE IN ESUS PATIENTS WITH IMPLANTABLE LOOP RECORDER (ILR) DEVICE. A COMPARATION BETWEEN AFIB AND NON-AFIB GROUP

M. de Lera¹, M. Sandin², A. Calleja¹, E. García-Morán², E. Cortijo¹, S. Bombin¹, L. Bautista¹, R. Alcaide¹, J.J. Navarrete¹, E. Martinez-Pías¹, B. Gómez-Vicente¹, J. Trigo¹, B. Talavera¹, A. Chavarría¹, J. Reyes¹, J. Rubio² and J.F. Arenillas¹

¹University Clinic Hospital of Valladolid, Stroke Unit. Neurology Department, Valladolid, Spain

²University Clinic Hospital of Valladolid, Cardiology Department, Valladolid, Spain

Whether implantable loop recorders (ILR) prevent recurrent hard end-points in patients with embolic stroke of unknown source (ESUS) remains unproven. Our aim was to analyse the stroke recurrence rate in a cohort of ESUS patients implanted with ILR, and to compare the recurrence rate between the groups with detected and non-detected atrial fibrillation (AF).

Prospective, observational, long-term follow-up study conducted at our Stroke Center. We included stroke patients with ESUS criteria who were implanted with a ILR and then followed up for a minimum of 12 months. AF was diagnosed after a confirmed > 2 minutes episode. After AF detection, a clinical decision regarding antithrombotic treatment was made within the next 48 hours. Recurrent stroke was considered the primary end point, and survival analyses were performed.

We included 238 ESUS patients implanted with ILR, 131 men, followed up for a mean time of 4 years after the index event. AF was detected and confirmed in 106 (45%) patients. Stroke recurrence occurred in 16 patients (6.7% = annual rate 1.7%). AF and non-AF groups did not differ in recurrence rates, as shown on Kaplan-Meier curve (log-Rank p = 0.618) and multivariate Cox-regression (HR: 0.9 [0.15-5.31] p = 0.9).

The annual recurrence rate observed (1.7%) was lower than in the ESUS clinical trials (5%). AF detection with ILR was not associated with a higher recurrence risk in our series. However, randomized clinical trials are needed to generate evidence in favour of this preventive strategy.

Not clinical trial

ATRIAL FIBRILLATION DIAGNOSED AFTER STROKE AND INCREASED RISK OF DEMENTIA: A COHORT STUDY OF FIRST-EVER ISCHAEMIC STROKE PATIENTS

M. Krawczyk¹, S. Fridman¹, C. Yi², J. Fang², S. Gustavo³, L. Sposato¹

¹Western University, Clinical Neurological Sciences, London, Canada

²ICES, na, Toronto, Canada

³University of Toronto, Division of Neurology- Department of Medicine, Toronto, Canada

Recent evidence suggests that AF detected after stroke (AFDAS) is relatively benign compared to AF known beforethe stroke. AF is a risk factor for dementia in the general population and among ischaemic stroke patients in whom the AF was diagnosed before the event. However, it remains unknown if AFDAS is also associated with increased dementia risk. We therefore assessed the association between AFDAS and the incident risk of dementia. We also evaluated whether the use of oral anticoagulants (OAC) was associated with lower dementia risk among AFDAS patients.

In this cohort study, we classified 9791 first-ever ischaemic stroke patients from the Ontario Stroke Registry into four groups: 1) No AF, 2) AF known before the stroke, 3) Inpatient AFDAS (diagnosed only during admission), and 4) Outpatient AFDAS (diagnosed after discharge). We used multivariable Cox proportional models to estimate cause-specific hazard ratios (HR) for the association between AFDAS and incident dementia risk. Dementia was determined through administrative datasets based on previously validated algorithms.

Relative to patients without AF, the risk of dementia was higher for those with inpatient AFDAS (HR 1.78, 95% CI 1.51−2.10) and outpatient AFDAS (HR 1.74, 95% CI 1.47−2.05). The use of OACs was associated with lower risk of dementia among patients with inpatient AFDAS (HR 0.58, 95% CI 0.43−0.79) and outpatient AFDAS (HR 0.60, 95% CI 0.43−0.83).

AFDAS was independently associated with higher risk of dementia relative to no AF. Among patients with AFDAS, the use of OACs was associated with lower dementia risk.

N/A

CEREBRAL SMALL VESSEL DISEASE BURDEN AND INTRACEREBRAL HAEMORRHAGE IN PATIENTS TAKING ORAL ANTICOAGULANTS: A PRE-PLANNED ANALYSIS OF THE CROMIS-2 STUDY

D. Seiffge^1,2,3, D. Wilson¹, G. Ambler⁴, G. Banerjee¹, I. Hostettler¹, H. Houlden⁵, C. Shakeshaft¹, H. Cohen⁶, T. Yousry⁷, R. Al-Shahi Salman⁸, G. Lip⁹, M. Brown¹, K. Muir¹⁰, H.R. Jager⁷, D. Werring¹; on behalf of the CROMIS-2 investigators

¹UCL Queen Square Institute of Neurology, Department of Brain Repair and Rehabilitation- Stroke Research Center, London, United Kingdom

²University Hospital Berne, Department of Neurology and Stroke Center, Berne, Switzerland

³University Hospital Basel and University of Basel, Department of Neurology and Stroke Center, Basel, Switzerland

⁴University College London, Department of Statistical Science, London, United Kingdom

⁵The National Hospital of Neurology and Neurosurgery, Neurogenetics Laboratory-, London, United Kingdom

⁶University College London, Haemostasis Research Unit- Department of Haematology, London, United Kingdom

⁷UCL Queen Square Institute of Neurology, Neuroradiological Academic Unit- Department of Brain Repair & Rehabilitation, London, United Kingdom

⁸University of Edinburgh, Centre for Clinical Brain Sciences- School of Clinical Sciences, Edinburgh, United Kingdom

⁹University of Birmingham, Institute of Cardiovascular Sciences, Birmingham, United Kingdom

¹⁰University of Glasgow- Queen Elizabeth University Hospital, Institute of Neuroscience & Psychology, Glasgow, United Kingdom

We investigated the role of small vessel disease (SVD) in intracerebral haemorrhage (ICH) in patients taking anticoagulants (OAC-ICH).

We analysed two complementary datasets from a prospective multi-centre observational study (CROMIS-2). We first determined the prevalence of SVD markers on CT and MRI in patients with OAC-ICH and non-OAC ICH. We then categorised patients with atrial fibrillation (AF) and recent cerebral ischaemia started on anticoagulant therapy according to baseline MRI SVD markers (using the SVD burden score), and compared the rates of ICH and ischaemic stroke during 2 years follow-up.

Our first analysis included 1060patients with ICH (421OAC-ICH and 609non-OAC ICH, median age76 years,42.6% female) and 1447patients with AF (SVD-absent:668 patients, SVD-present:779 patients, median age77years,42.2% female). SVD markers were highly prevalent in both OAC-ICH and non-OAC-ICH, with no statistically significant difference in SVD marker prevalence, severity or burden score between the groups (all p>0.05 in adjusted analyses). In the AF cohort, during 3366patient-years of follow-up there were 11ICH and 56ischaemic strokes. The rate of ICH was 0.06%/year (IQR0.00-0.35) in patients with no SVD and 0.56%/year (IQR0.27-1.03) in patients with SVD(p = 0.001); SVD presence was independently associated with ICH (HR8.3,95% CI 1.1-65.6, p = 0.046). There was a non-significant association between SVD and ischemic stroke (HR1.8, IQR1.0-3.2, p = 0.06).ICH risk was predicted by models including SVD presence (c-index0.70,95% CI 0.60-0.75) or severity (c-index0.75,95% CI 0.63-0.85); adding age, hypertension or HAS-BLED-score did not improve discrimination.

SVD was highly prevalent in both OAC-ICH and non-OAC ICH. SVD independently predicted incident ICH in patients with AF taking OAC;absence was associated with very low risk of ICH. Our findings confirm SVD has an important association of ICH.

NCT02513316

RISK FACTORS FOR SPONTANEOUS INTRACEREBRAL HAEMORRHAGE ACCORDING TO LOCATION: A SYSTEMATIC REVIEW AND META-ANALYSIS

W. Jolink¹, K. Wiegertjes², G. Rinkel¹, A. Algra³, F.E. de Leeuw² and K. Klijn²

¹University Medical Center Utrecht, Neurology and Neurosurgery, Utrecht, The Netherlands

²Radboud university medical center, Neurology, Nijmegen, The Netherlands

³University Medical Center Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands

Risk factors for spontaneous intracerebral haemorrhage (ICH) may differ for lobar and non-lobar ICH. We performed a systematic review and meta-analysis of studies for risk factors according to the location of the ICH.

We searched Pubmed and EMBASE for case-control and cohort studies published between January 1^st, 1980, and December 7^th, 2018. We included studies that reported risk factors according to ICH location (lobar and non-lobar). Two authors independently extracted data on risk factors (n = 28) of which we here report on hypertension, diabetes, smoking, alcohol abuse and hypercholesterolaemia. Maximally adjusted estimates were pooled with the generic variance-based random effects method.

We included 40 studies with a total of 25,965 patients with ICH (9,118 lobar, 16,847 non-lobar). For hypertension, pooled relative risks (RR) were 1.86 (95% CI 1.26–2.74; 5 studies) for lobar and 4.10 (2.18–7.70; 6 studies) for non-lobar; for diabetes 1.18 (0.86–1.62; 4 studies) for lobar and 1.55 (1.08–2.23; 6 studies) for non-lobar; for smoking 1.08 (0.86-1.34; 3 studies) for lobar and 1.28 (1.03–1.60; 5 studies) for non-lobar; for alcohol abuse 0.94 (0.74–1.20; 2 studies) for lobar and 1.58 (1.11–2.26; 3 studies) for non-lobar, and for hypercholesterolaemia 1.01 (0.27–3.76; 2 studies) for lobar and 0.49 (0.38–0.62; 2 studies) for non-lobar ICH.

Hypertension is a risk factor for both non-lobar and lobar ICH, and diabetes, smoking and alcohol abuse for non-lobar ICH only. Hypercholesterolemia was associated with a low relative risk of non-lobar ICH. Thus, the term ‘hypertensive ICH’ for deep ICH is not appropriate and appears obsolete.

N/A

VESSEL WALL ENHANCEMENT OF INTRACRANIAL ANEURYSMS ON HIGH RESOLUTION MRI: A RUPTURE SIGN? THE ICARUS STUDY

C. Vert Soler¹, L.L. Gramegna², M. Luzi³, F. Arikan Abello⁴, D. Hernandez Morales², P. Coscojuela Santaliestra², E. Martinez Saez⁵, L. Derchi⁶, A. Rovira Cañellas¹, A. Tomasello Weitz² and L. Dinia²

¹Vall D’Hebron Hospital, Section of Neuroradiology, Barcelona, Spain

²Vall D’Hebron Hospital, Interventional Neuroradiology, Barcelona, Spain

³Università Politecnica delle Marche, Neurosurgery, Ancona, Italy

⁴Vall D’Hebron Hospital, Neurosurgery, Barcelona, Spain

⁵Vall D’Hebron Hospital, Neuropathology, Barcelona, Spain

⁶University of Genova, Department of Radiology, Genova, Italy

Instability state in saccular intracranial aneurysm (SIA) is still difficult to estimate. Inflammation plays a major role and aneurysm wall enhancement (WE) on MRI may be a surrogate of vessel wall inflammation.

Main aim of the study is to assess the capability of WE to identify instability of SIA in real danger for rupturing.

Cohort prospective observational study. Consecutive patients with ruptured and un-ruptured aneurysms underwent contrast 3T-MRI and histology to confirm wall inflammation. Statistical analysis was performed to investigate the association of WE with risk factors and histological markers of inflammation.

A preliminary group of 50 patients was included. WE was present in 100% of the ruptured aneurysms, in 50% of unruptured symptomatic and in 33% of asymptomatics (p = 0.003). Cut-offline of 4 in PHASES score was found, above which most of the aneurysms begin to enhance (p < 0.001). Independent predictors of rupture were: WE (p = 0.018) and presence of multiple aneurysms (p = 0.016). Independent predictors of WE: size (p = 0.001), variables of circle of Willis (p = 0.005) and irregular morphology (p = 0.016). Analysis of WE as a predictor of rupture has shown (p = 0.002): 100% sensitivity, 64% specificity, 28%VPP, 100%VPN. At a cut-off of 7, there is a direct correlation between the number of altered immunohistochemical inflammation markers and presence of WE(p = 0.043). A score higher than 7 entails a 100% sensitivity, specificity, VPP and VPN in relation to presence of WE.

Wall enhancement identifies instability-state in SIA. WE is a new and powerful rupture sign that could be crucial in deciding whether and how to treat SIA, even in small aneurysms.

N/A

COMPARISON OF RUPTURE RISK BETWEEN FAMILIAL AND SPORADIC INTRACRANIAL ANEURYSMS: AN INDIVIDUAL PATIENT DATA META-ANALYSIS

C.C.M. Zuurbier¹, L.A. Mensing¹, M.J.H. Wermer², S. Juvela³, A.E. Lindgren⁴, J.E. Jääskeläinen⁴, T. Koivisto⁴, A. Morita⁵, H. Arai⁶, K. Nozaki⁷, Y. Murayama⁸, T. Ishibashi⁸, H. Takao⁸, G.J.E. Rinkel¹, J.P. Greving⁹ and Y.M. Ruigrok¹

¹University Medical Center Utrecht, Neurology, Utrecht, The Netherlands

²Leiden University Medical Center, Neurology, Leiden, The Netherlands

³University of Helsinki, Clinical Neurosciences, Helsinki, Finland

⁴Kuopio University Hospital- and Institute of Clinical Medicine- School of Medicine- Faculty of Health Sciences- University of Eastern Finland-, Neurosurgery of NeuroCenter, Kuopio, Finland

⁵UCAS Japan Coordinating Office- University of Tokyo- Nippon Medical School, Neurological Surgery, Tokyo, Japan

⁶Juntendo University- Medical School-, Neurosurgery, Tokyo, Japan

⁷Shiga University of Medical Science, Neurosurgery, Shiga, Japan

⁸Tokyo Jikei University School of Medicine, Endovascular Neurosurgery, Tokyo, Japan

⁹University Medical Center Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands

The incidence of aneurysmal subarachnoid hemorrhage (aSAH) is higher in persons with than in those without a positive family history of aSAH. This higher incidence is in part explained by a higher prevalence of unruptured intracranial aneurysms (UIA), but it is unclear whether persons with a positive family history also have a higher risk of rupture of UIA. One study showed a 17-times higher rupture risk for patients with a family history of aSAH and a history of smoking or hypertension compared to patients with a sporadic UIA. We assessed rupture risk of familial versus sporadic UIA taking into account other risk factors for rupture.

We pooled individual patient data for 9,143 patients (853 familial patients, 9.3%) with 11,208 UIA and 22.003 person-years follow-up using data from 6 prospective cohort studies. Rupture rates per patient were analyzed with a Cox proportional hazard regression model stratified per cohort and adjusted for the PHASES score.

Rupture occurred in 19 patients with familial UIA (rupture rate: 0.85%/person-year) and 187 patients with sporadic UIA (0.95%/person-year). The hazard rate (HR) for familial aneurysms compared to sporadic aneurysms was HR 0.98 (95% CI: 0.61-1.59) and the adjusted HR 1.39 (95% CI: 0.86-2.25).

We found a slightly but not statistically significantly increased risk of aneurysm rupture for familial compared to sporadic UIA. When assessing the risk of rupture in familial UIA, the family history is not as pivotal as previously assumed.

N/A

MANAGEMENT DECISIONS ON UNRUPTURED INTRACRANIAL ANEURYSMS BEFORE AND AFTER IMPLEMENTATION OF THE PHASES SCORE

L. Hollands¹, A.M. Algra¹, M.D.I. Vergouwen¹, M.J.H. Wermer², J.P. Greving³ and G.J.E. Rinkel¹

¹University Medical Center Utrecht, Department of Neurology and Neurosurgery- Brain Center Rudolf Magnus, Utrecht, The Netherlands

²Leiden University Medical Center, Department of Neurology, Leiden, The Netherlands

³University Medical Center Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands

In treatment decisions on saccular unruptured intracranial aneurysms (UIAs) the risk of rupture is an important determinant. PHASES is a score that provides absolute 5-year risks of rupture and was introduced in 2014. We assessed whether management decisions on UIA’s changed after implementation of the PHASES score.

We included all patients with UIAs who presented to the University Medical Center Utrecht between 2011 and 2017. We extracted patient and aneurysm characteristics needed to calculate PHASES scores and divided the cohort into pre- and post-PHASES implementation groups, excluding a 6-month implementation period in 2014. Analyses were primarily conducted on an aneurysm level and treatment decisions were scored as to observe or to intervene. We compared median PHASES scores before and after implementation. We calculated risk differences with 95% confidence intervals (CI) to assess the relationship between implementation of the PHASES score and the decision to observe or to intervene.

We included 430 patients with 546 UIAs (263 before, 283 after implementation). In the cohort after implementation, the decision to intervene was less often made (26.2% after vs 35.4% before; risk difference 9.2%; 95% CI:1.5%-16.8%). The median PHASES score of UIAs with a decision to intervene increased from 7 points before to 8 points after implementation.

In our center, the decision for preventive treatment was less often made after implementation of the PHASES score. Additional studies are needed to replicate our findings and to assess to what extent changes in decision making influence patient outcome.

N/A

COMPARISON BETWEEN BEST MEDICAL THERAPY AND VARIOUS INTERVENTIONAL THERAPIES FOR UNRUPTURED BRAIN ARTERIOVENOUS MALFORMATIONS: A SYSTEMATIC REVIEW AND META-ANALYSIS

M. Memon¹, S. Majidi², D. Daniel³, C. Chen⁴, Y. Ma⁴ and A. Qureshi⁵

¹University of Miami- Jackson Memorial Hospital, Neurology, Miami, USA

²Icahn School of Medicine at Mount Sinai, Neurosurgery, New York, USA

³George Washington University, Neurology, Washington- DC, USA

⁴George Washington University, School of Public Health, Washington- DC, USA

⁵Zeenat Qureshi Stroke Research Center Institute, Neurology, St. Cloud- Minnesota, USA

Management of unruptured brain arteriovenous malformations (AVM) remains controversial since the publication of ARUBA trial. Our objective was to compare the outcomes of patients with unruptured AVM following various interventional treatments versus medical management through a systematic review of previously published reports.

We identified previously published studies reporting outcome following medical management and/or interventional therapies between January 1966 and December 2017 using literature search based on pre-specified criteria. The rates of subsequent hemorrhage, permanent neurological deficits (PND), AVM related and unrelated death were estimated for best medical treatment, surgical treatment, radiation therapy and endovascular treatment. Random effects model was used to calculate pooled estimates and compare the rates of various outcomes.

A total of 47 reports were included in the analysis. There was no statistical difference in baseline characteristics except a higher proportion of patients with Spetzler-Martin Grade IV-VI were included in the interventional group (p-value = 0.0397). There was no statistical difference in the rates of ICH, PND and death between medical and interventional groups overall [pooled proportions (PP) 0.0869 vs 0.0545 (p-value = 0.3561)], [PP 0.0577 vs 0.0468 (p-value = 0.5848)] and [PP 0.0330 vs 0.0154 (p-value = 0.5778)] respectively. In the subsequent subgroup comparison among interventional treatments, surgical patients had statistically significant lower risk of subsequent ICH than radiosurgery (p-value = 0.0125) and endovascular embolization (p-value = 0.0181).

In this analysis, we did not identify any differences in rates of subsequent ICHs between interventional treatment and medical treatment alone. Among the interventional groups, surgical patients had less subsequent hemorrhages than embolization or radiosurgery patients.

N/A

STEREOTACTIC RADIOSURGERY FOR CEREBRAL CAVERNOUS MALFORMATIONS: A SYSTEMATIC REVIEW AND META-ANALYSIS

M.H.F. Poorthuis¹, L. Rinkel² and R. Al-Shahi Salman³

¹University Medical Center Utrecht, Department of Neurology and Neurosurgery- Brain Center Rudolf Magnus, Utrecht, The Netherlands

²University of Groningen, University Medical Centre Groningen, Groningen, The Netherlands

³University of Edinburgh, Centre for Clinical Brain Sciences, Edinburgh, United Kingdom

Stereotactic radiosurgery (SRS) is used for the treatment of cerebral cavernous malformations (CCMs) in everyday clinical practice, but its efficacy is uncertain. Therefore, we aimed to: (1) estimate the incidence of clinical outcomes after SRS, and (2) compare the outcome of SRS with neurosurgical excision or conservative management in comparative studies.

We systematically searched PubMed and EMBASE from inception until June 1, 2018 for peer-reviewed publications describing clinical outcome after SRS for ≥10 people with CCM (confirmed by MRI or pathological examination) in studies with or without a comparison group treated with neurosurgical excision or conservative management.

We included 25 case series and five non-randomised comparative studies, including 1,576 patients undergoing SRS. Median duration of follow-up was 48 (range 22–98) months (Table 1). After SRS, the incidence (per 100 person-years [95% CI]) of death attributable to CCM and/or its treatment was 0.18 [0.09-0.30], of symptomatic intracranial haemorrhage was 2.40 [2.05-2.80], and of focal neurological deficit was 0.71 [0.53-0.96] (Table 2). We were unable to reliably compare SRS to other treatment strategies due to confounding and bias in both the five non-randomised studies as well as before-and-after comparisons in case series.

Most studies of SRS for CCM are at high risk of bias and confounding, and lack a comparison group. The annual risk of major clinical outcomes after SRS for CCM is ∼3% over around four years of follow-up, which is comparable to outcomes without SRS. A randomised trial of SRS for CCM is needed.

N/A

MORTALITY ANALYSIS FOR THE MINIMALLY INVASIVE SURGERY WITH THROMBOLYSIS IN INTRACEREBRAL HEMORRHAGE EVACUATION (MISTIE) III TRIAL: WHO DIES AND WHY?

W. Ziai¹, N. McBee¹, R. Avadhani¹, T. Chang², P. Camarata³, M. Malkoff⁴, J.R. Carhuapoma¹, C. Kase⁵, J. Jaffe⁶, K. Jonczak⁷, N. Ostapkovich¹, T. Economas¹, C. Koenig⁸, I. Awad⁹, D. Hanley¹; MISTIE III Trial Investigators

¹Johns Hopkins Medical Institutions, Neurology, Baltimore, USA

²The University of Texas, Neurosurgery, Houston, USA

³University of Kansas, Neurosurgery, Kansas, USA

⁴University of Tennessee, Neurology, Memphis, USA

⁵Boston University, Neurology, Boston, USA

⁶North Shore University, Neurosurgery, Chicago, USA

⁷Temple University, Neurosurgery, Philadelphia, USA

⁸Emissary International- LLC, Neurosciences, Austin, USA

⁹University of Chicago, Neurosurgery, Chicago, USA

MISTIE III was a 506-subject, multicenter, randomized trial comparing minimally invasive catheter hematoma evacuation followed by alteplase for treatment of large intracerebral hemorrhage. Mortality for this disease ranges from 40–80%. We evaluated proximate causes of death in the acute and post-hospitalization phase and withdrawal of active treatment (WAT) decisions.

We reviewed serious adverse event reports for 110 deaths within 1 year follow-up. We tested for associations between death, early death (<30 days) vs. later death (30-365d), and WAT with baseline factors, hospital course, and hematoma volumes in multifactorial logistic regression models and compared with 137 deaths in the 500 subject CLEAR III trial.

One year mortality was 110/499 (22.0%) at median[IQR] 27.5[14-85] days from onset. The most common adjudicated primary causes of death were direct effects of primary hemorrhage (61%), cardiac (9%) and respiratory (13%) complications. Mechanism of death was brain death in 2%, WAT (56%), cardiac arrest (7%) and other medical complications (35%). Early vs. later death occurred in 60 (54.5%). Lower percentage ICH volume reduction and higher admission NIHSS were independently associated with early death after adjustment for WAT. Independent associations with WAT for 61 MISTIE and 57 CLEAR III subjects were age, poor GCS, anticoagulation, early pneumonia and enrollment in North America.

The MISTIE III mortality rate remains below literature estimates for severe ICH. In both severe ICH and intraventricular hemorrhage, just over half of known causes of death are directly due to neurologic events. Reduction in hematoma volumes may lower early and overall mortality risk.

NCT01827046

THE ROLE OF HEMATOMA EVACUATION IN SPONTANEOUS CEREBELLAR INTRACEREBRAL HAEMORRHAGE: A SYSTEMATIC REVIEW

S. Singh¹, H.B. Brouwers¹, J.R. Senff¹, M. Pasi², J.N. Goldstein³, A. Viswanathan², C.J.M. Klijn⁴ and G.J.E. Rinkel¹

¹University Medical Center Utrecht, Department of Neurology and Neurosurgery- Brain Center Rudolf Magnus, Utrecht, The Netherlands

²Massachusetts General Hospital- Harvard Medical School, Hemorrhagic Stroke Research Program- Department of Neurology, Boston, USA

³Massachusetts General Hospital- Harvard Medical School, Division of Neurocritical Care and Emergency Neurology- Department of Neurology, Boston, USA

⁴Radboud University Medical Center, Department of Neurology- Donders Institute for Brain Cognition & Behavior, Nijmegen, The Netherlands

Guidelines regarding recommendations for hematoma evacuation in spontaneous cerebellar intracerebral haemorrhage (ICH) differ. We aimed to systematically review the literature to assess treatment strategies and outcomes.

We searched PubMed and Embase between 1970 and 2019 for randomized or otherwise controlled studies and observational cohort studies. We included relevant studies according to predefined selection criteria and assessed their quality and risk of bias according to the Newcastle Ottawa Scale (NOS). We assessed case-fatality and functional outcome in patients treated conservatively or with hematoma evacuation. Favourable functional outcome was defined as a modified Rankin Scale (mRS) score of 0–2 or a Glasgow Outcome Scale (GOS) score of 4–5.

We included 41 non-controlled cohort studies describing 2062 patients (40% female) with spontaneous cerebellar ICH. 1171 patients (57%) underwent hematoma evacuation. The median NOS score (IQR) was 5 (4-6) out of 8 points and the bias score 2 (1-3) out of 8 (indicative of high risk of bias). Overall case-fatality at discharge was 21% [95% CI 17–25%] after conservative treatment and 24% [95% CI 19–29%] after hematoma evacuation. At ≥6 months, overall case-fatality was 30% [95% CI 25–30%] and favourable functional outcome 45% [95% CI 40–50%] following conservative treatment and 34% [95% CI 30–38%] and 42% [95% CI 37–47%] respectively after hematoma evacuation.

Cerebellar ICH is associated with poor outcome irrespective of treatment strategy. However, the risk of bias is high, and there may well be highly beneficial currently available treatment strategies that could be clarified with higher quality observational studies and clinical trials.

N/A

Scientific Communications 20-Large Artery Disease and Treatment

AORTIC ARCH ATHEROSCLEROSIS IN PATIENTS WITH EMBOLIC STROKE OF UNDETERMINED SOURCE: AN EXPLORATORY ANALYSIS OF THE NAVIGATE-ESUS TRIAL

G. Ntaios¹, P. Amarenco², E. Meseguer³, M. Endres⁴, S. Ozturk⁵, W. Lang⁶, N.M. Bornstein⁷, C. Molina⁸, J. Pagola⁸, H. Mundl⁹, S.D. Berkowitz¹⁰, G. Peters¹¹, Y.Y. Liu¹², S. Sen¹³, R. Hart¹⁴; The NAVIGATE ESUS investigators

¹University of Thessaly, Department of Internal Medicine, Larissa, Greece

²Paris-Diderot- Sorbonne Paris Cité university, APHP- Bichat hospital, Paris, France

³Bichat Hospital, Neurology Service-, Paris, France

⁴Charité-Universitätsmedizin Berlin, Klinik und Hochschulambulanz für Neurologie, Berlin, Germany

⁵Selcuk University Faculty of Medicine, Department of Neurology, Konya, Turkey

⁶Sigmund Freud Private University- Medical Faculty, Hospital St. John of God, Vienna, Austria

⁷Shaare-Zedek, Medical Center, Jerusalem, Israel

⁸Vall d´Hebron Stroke Unit, Hospital Universitari Vall d´Hebron, Barcelona, Spain

⁹Bayer AG, Bayer AG, Wuppertal, Germany

¹⁰Pharmaceuticals Clinical Development Thrombosis, Bayer U.S. LLC-, New Jersey, USA

¹¹Janssen, Research and Development, Spring House, USA

¹²Population Health Research Institute, Hamilton, Ontario, Canada

¹³University of South Carolina- School of Medicine, Department of Neurology-, Columbia, USA

¹⁴McMaster University/Population Health Research Institute, Department of Medicine Neurology, Hamilton- Ontario, Canada

Aortic arch atherosclerosis (AAA) is an important source of brain embolism in patients with embolic stroke of undetermined source (ESUS). The optimal antithrombotic strategy for secondary prevention in patients with AAA is uncertain. This exploratory analysis of the NAVIGATE-ESUS trial assessed baseline characteristics, stroke recurrence rates, and response to rivaroxaban versus aspirin of participants with AAA.

Analysis of AAA was restricted to the participants undergoing transesophageal echocardiography (TEE) at baseline. AAA plaques were considered to have complex features when reported as complex or ulcerated, or were ≥4 mm in thickness or had a mobile thrombus present.

Among 7213 participants, 1382 (19.2%) had TEE, among whom 397 (28.7%) had AAA and 112 (8.1%) had complex AAA. Age (OR:1.05, 95% CI:1.02-1.07), diabetes mellitus (OR:1.63, 95% CI:1.02-2.61), coronary artery disease (OR:2.69, 95% CI:1.06-6.78), statin use prior to randomization (OR:2.68, 95% CI:1.62-4.44) and aortic valve disease (OR:1.93, 95% CI:1.09-3.30) were independent predictors of complex AAA. Patients with complex AAA showed a strong trend of having more frequently multiterritorial qualifying stroke compared with patients without complex AAA (20.5% vs. 14.2%, p = 0.06). The annualized rates of recurrent ischemic stroke were 8.1% in patients with complex AAA (9 recurrent strokes) vs. 5.2% without complex AAA (68 recurrent strokes), HR: 1.53, 95% CI: 0.76-3.07, p = 0.23]. There was no interaction with treatment assignment (p = 0.75).

Complex AAA was present in 8% of NAVIGATE-ESUS participants undergoing TEE and was associated with a trend for higher stroke recurrence rate. Too few events were observed among those with complex AAA to meaningfully assess the effect of antithrombotic treatment.

NCT02313909

RATES OF AORTIC ARCH ATHEROMA CORRELATED WITH CAROTID STENOSIS AND VASCULAR RISK FACTORS IN RETINAL ISCHEMIA PATIENTS PRESENTING TO A TERTIARY REFERRAL CENTRE TIA CLINIC

F. Humphries¹, M. Adams², L. Cooper³, M. Delaney³, A. Cheng⁴, A. Zarkali⁵, A. Dados³, R. Simister¹ and A. Chandratheva¹

¹UCL Stroke Research Centre- UCL Institute of Neurology, Department of Brain Repair and Rehabilitation, London, United Kingdom

²National Hospital for Neurology and Neurosurgery, Neuroradiology, London, United Kingdom

³University College London Hospital, Hyperacute Stroke Unit, London, United Kingdom

⁴University College London, Division of Surgery and Interventional Science, London, United Kingdom

⁵University College London, Dementia Research Centre, London, United Kingdom

We aimed to determine the rates of aortic arch atheroma (AAA) in patients with ischaemic transient (TVL) and permanent visual loss (PVL).

We reviewed the records for consecutive patients from June 2013-September 2018 presenting to the daily TIA clinic at UCLH, a regional referral centre for north central London and Moorfields Eye Hospital. We reviewed the CT angiograms (CTA) for TVL and PVL and rated AAA proximal to the ostium of the left subclavian artery using a previously published rating score: grade 1 = 4 mm thick; grade 2 = 4 mm thick; grade 3 = 4 mm thick, encroaching on lumen; grade 4 = 4 mm thick pedunculated/with thrombus. We compared AAA with doppler measured carotid stenosis (CS), and number of vascular risk factors (No.RFs): hypertension, diabetes, atrial fibrillation, stroke, TIA, hypercholesterolaemia, smoking.

Of 485 patients, 256 (53%) had a CTA, 159 (62%) male, mean age 63.1 years. 155 (61%) had TVL and 101 (39%) had PVL. 251 CTAs were of diagnostic quality and were assessed. 34% had AAA = 1, 14% had AAA = 2, 3% had AAA>2. Mean AAA scores: TVL = 0.6, PVL = 0.9 (p = 0.004). Mean No.RFs were similar with and without AAA (1.7 vs.1.8, p = 0.07). There was moderate correlation between CS and AAA (r = 0.354, p = 0.000). 13.7% of CS < 50% had AAA >1 compared to 34.9% of CS>50% (p = 0.001; OR 3.4, 95% CI 1.6-7.1, p = 0.001).

This is the first description of AAA in ischaemic visual loss. 51% have AAA, AAA is more severe in PVL and if CS >50%. Rates of AAA in PVL are comparable to patients with cerebral infarcts.

N/A

PREDICTING RISK OF VASCULAR EVENTS AFTER CAROTID REVASCULARIZATION: MODEL DEVELOPMENT STUDY IN INDIVIDUAL PATIENT DATA FROM FOUR RANDOMISED TRIALS

E.J. Volkers¹, A. Algra¹, L.J. Kappelle², N.P. Zuithoff¹, O. Jansen³, G. Roubin⁴, G. Howard⁵, J. Hendrikse⁶, A. Halliday⁷, J. Gregson⁸, G. Fraedrich⁹, H.H. Eckstein¹⁰, D. Calvet¹¹, R. Bulbulia¹², M.M. Brown¹³, G.J. de Borst¹⁴, J.P. Becquemin¹⁵, P.A. Ringleb¹⁶, J.L. Mas¹⁷, T.G. Brott¹⁸, L.H. Bonati¹⁹, J.P. Greving¹; on behalf of the Carotid Stenosis Trialists’ Collaboration

¹UMC Utrecht, Julius Centre for Health Sciences and Primary Care- University Medical Centre Utrecht- Utrecht University, Utrecht, The Netherlands

²UMC Utrecht, Department of Neurology and Neurosurgery- Brain Centre Rudolf Magnus- University Medical Centre Utrecht- Utrecht University, Utrecht, The Netherlands

³UKSH Campus Kiel, Clinic for Radiology and Neuroradiology, Kiel, Germany

⁴Cardiovascular Associates of the Southeast, LCc, Birmingham- AL, USA

⁵UAB School of Public Health, Department of Biostatistics, Birmingham, USA

⁶UMC Utrecht, Department of Radiology, Utrecht, The Netherlands

⁷John Radcliffe Hospital, Nuffield Department of Surgical Sciences, Oxford, United Kingdom

⁸London School for Hygiene and Tropical Medicine, Department of Medical Statistics, London, United Kingdom

⁹Medical University of Innsbruck, Department of Vascular Surgery, Innsbruck, Austria

¹⁰Klinikum rechts der Isar- Technical University Munich, Department of Vascular and Endovascular Surgery/Vascular Centre, Munich, Germany

¹¹Department of Neurology, Hô pital Sainte-Anne- Université Paris-Descartes- DHU Neurovasc Sorbonne Paris Cité- INSERM U894, Paris, France

¹²MRC Population Health Research Unit- Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health- University of Oxford, Oxford, United Kingdom

¹³University College London, Department of Brain Repair and Rehabilitation- UCL Institute of Neurology, London, United Kingdom

¹⁴University Medical Centre Utrecht- Utrecht University, Department of Vascular Surgery, Utrecht, The Netherlands

¹⁵Hopital Privé Paul D’Egine, IVPe, Champigny-sur-Marne, France

¹⁶University of Heidelberg Medical School, Department of Neurology, Heidelberg, Germany

¹⁷Hôpital Sainte-Anne- Université Paris-Descartes- DHU Neurovasc Sorbonne Paris Cité- INSERM U894, Department of Neurology, Paris, France

¹⁸Mayo Clinic, Department of Neurology, Jacksonville- FL, USA

¹⁹University Hospital Basel- University of Basel, Department of Neurology and Stroke Centre- Department of Clinical Research, Basel, Switzerland

Identification of clinical characteristics that predict risk of stroke after carotid artery stenting (CAS) or endarterectomy (CEA) could lead to personalized treatment decisions. We developed two prediction models for procedural stroke or death risk ≤30 days after CAS and CEA and one for postprocedural ipsilateral stroke risk >30 days after both procedures.

We combined individual patient-level data from 4,754 patients with symptomatic carotid stenosis (EVA-3S, SPACE, ICSS, CREST). Per-protocol logistic regression and Cox regression analyses stratified by trial were performed to study the association between predictors and outcomes. Performance was examined with c-statistics and calibration plots.

Procedural stroke or death occurred in 158 (7.2%) CAS patients and 84 (3.7%) CEA patients. Postprocedural ipsilateral stroke occurred in 130 patients (rate 6.7/1,000 person-years). Predictors of procedural risk after CAS were age, diabetes, atrial fibrillation, modified Rankin Scale (mRS) score, previous hemispheric event, serum creatinine, left-sided procedure, and ipsilateral severe carotid stenosis. Predictors of procedural risk after CEA were hypertension, mRS score, and serum creatinine. Hypertension, coronary heart disease, heart failure, and contralateral severe carotid stenosis/occlusion predicted postprocedural risk. The c-statistic was higher for the procedural CAS model (0.68; 95% CI 0.63-0.73) compared with the procedural CEA (0.61; 0.54-0.69) and postprocedural model (0.61; 0.54-0.68). Risk difference between individuals with and without risk factors was largest for the procedural CAS model (range 3% to 18%).

We developed prediction models for risk of stroke after carotid revascularisation. External validation is required before these models can be used to support treatment decisions in individual patients.

N/A

RANDOMIZED STUDY COMPARING THE TYPE OF CAROTID STENT AND CEREBRAL PROTECTION DURING CAROTID ARTERY STENTING IN PATIENTS WITH HIGH-RISK PLAQUE

L. Caputi¹, P. Montorsi², S. Galli², P. Ravagnani², G. Teruzzi², A. Annoni³, G. Calligaris², F. Fabbiocchi², D. Trabattoni², S. De Martini², L. Grancini², G. Pontone³, D. Andreini³, S. Troiano², G. Migliore² and A. Bartorelli²

¹Fondazione IRCCS Istituto Neurologico C. Besta, Cerebrovascular Department, Milan, Italy

²Centro Cardiologico Monzino, Cardiology, Milan, Italy

³Centro Cardiologico Monzino, Radiology, Milan, Italy

The role of the stent type during CAS is unclear. New double-mesh carotid stents might reduce embolic complication.

The comparison of the safety and efficacy of the carotid stent Roadsaver (RS) vs. the Carotid Wallstent (CW) during carotid artery stenting (CAS) in patients with lipid-rich plaques. Both stents were randomly tested with FilterWire (FW) and MO.MA cerebral protection.

104 consecutive patients with CA stenosis and lipid-rich plaque were randomized to CAS with CW vs. RS and MO.MA vs. FW in 4 groups: FW+RS (group1, n = 27), FW+CW (group2, n = 25), MO.MA+RS (group3, n = 27) and MO.MA+CW (group4, n = 25). Primary end-point: number of MES by TCD in the following steps: 1) target vessel access with diagnostic catheter; 2) target vessel incannulation with guide/MO.MA; 3) lesion wiring; 4) lesion predilation; 5) lesion stent crossing; 6) stent deployment; 7) stent dilation, 8) device retrieval/deflation. Secondary end-points: in-hospital and 30-day Major Adverse Cardiac and Cerebrovascular events (MACCES), acute and long-term stent and target vessel ECA patency by Doppler US at 1, 30 and 180 days after CAS.

Patient’s and lesion characteristics were similar. MO.MA vs FW significantly reduced MES (p < 0.0001) during steps 2–3-5-6-7. During RS, compared to CW group, MES were significantly lower (step 6 to 8; p = 0.031). MO.MA+RS performed significantly better than MO.MA+CW (p = 0.043). MACCES occurred in 4 patients (1 major, 3 minor events, p = NS between groups). Rare In-stent restenosis (1.06% of cases). No stent thrombosis occurred.

In patients with high-risk lipid-rich plaque undergoing CAS, MO.MA+RS seems to be a promising tool in limiting embolic risks.

ClinicalTrials.gov Identifier: NCT02915328

TIMING OF PROCEDURAL COMPLICATIONS AMONG ASYMPTOMATIC PATIENTS UNDERGOING CAROTID ENDARTERECTOMY: ANALYSIS OF VA, ACAS, ACST-1 AND GALA TRIALS

M. Poorthuis^1,2, D. Morris¹, H. Pan¹, S. Lewis³, P. Rothwell⁴, R. Bulbulia^1,5, G. de Borst², A. Halliday⁶; on behalf of the Carotid Stenosis Trialists Collaboration

¹Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, Oxford, United Kingdom

²University Medical Centre Utrecht, Department of Vascular Surgery, Utrecht, The Netherlands

³Edinburgh Clinical Trials Unit, Usher Institute of Population Health Sciences and Informatics, Edinburgh, United Kingdom

⁴Centre for Prevention of Stroke and Dementia, Nuffield Department of Clinical Neurosciences, Oxford, United Kingdom

⁵Medical Research Council Population Health Research Unit, Nuffield Department of Population Health, Oxford, United Kingdom

⁶John Radcliffe Hospital, Nuffield Department of Surgical Sciences, Oxford, United Kingdom

Effectiveness of carotid endarterectomy (CEA) for stroke prevention depends on low procedural risks. We aimed to assess frequency and timing of procedural complications after CEA, which may clarify underlying mechanisms and help inform safe discharge policies.

Individual patient data (N = 8,752) were obtained from four large trials (VA, ACAS, ACST-1, and GALA; 1983–2007). Patients undergoing CEA for asymptomatic carotid artery stenosis (N = 4,462) were included. Timing of procedural death and stroke was stratified into day 0, days 1–3, days 4–30.

Procedural complications occurred in 138 (3.1%) patients. Results are reported in the Table.

Half of the procedural deaths and strokes occur on the day of operation and one third after day 3 when most patients will have been discharged. This analysis of procedural complications may help in planning strategies to reduce procedural complications after CEA.

N/A

STENTING FOR SYMPTOMATIC VERTEBRAL ARTERY STENOSIS: POOLED INDIVIDUAL PATIENT DATA ANALYSIS OF VIST, VAST AND SAMMPRIS

E. Harshfield¹, H. Markus¹, A. Compter², W. Kuker³, L.J. Kappelle², A. Clifton⁴, H.B. van der Worp², P. Rothwell³, A. Algra^2,5; Vertebral Stenosis Trialist’s Collaboration

¹University of Cambridge, Department of Clinical Neurosciences, Cambridge, United Kingdom

²University Medical Centre Utrecht, Department of Neurology, Utrecht, The Netherlands

³Oxford University Hospitals NHS Foundation Trust, Department of Neuroradiology, Oxford, United Kingdom

⁴St. Georges University Hospitals NHS Foundation Trust, Department of Neuroradiology, London, United Kingdom

⁵University Medical Centre Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands

Symptomatic vertebral artery stenosis has been associated with high risk of recurrent stroke, particularly in the first few weeks after symptoms, with higher risks for intracranial (IC) than extracranial (EC) stenosis. Vertebral artery stenosis can be treated with stenting with good technical results, but a higher periprocedural risk has been reported for IC than EC stenosis. We performed an individual patient pooled analysis of RCTs comparing vertebral stenting with medical treatment for symptomatic vertebral stenosis.

Data were included from VIST (N = 179, EC 148, IC 31), VAST (N = 115, EC 96, IC 19) and SAMMPRIS (IC 60, no EC). Cox regression was performed stratified by trial. The primary outcome was any fatal or nonfatal stroke during follow-up. A further analysis was performed for patients randomised within 14 days of last symptoms.

The hazard ratio (HR) for any stroke in the stenting arm compared with the medical arm was 0.81 (95% CI 0.45-1.44). For EC alone it was 0.63 (0.27-1.46), and for IC alone 1.06 (0.46-2.42). For patients randomised within 14 days of last stroke/TIA the HRs were: all 0.65 (0.31-1.39), EC only 0.56 (0.17-1.87), IC only 0.72 (0.27-1.90).

This pooled analysis of available data from completed vertebral artery stenting trials did not show a statistically significant benefit for either medical or interventional therapy. There was no evidence of even a trend of benefit of either strategy for IC stenosis. Stenting for EC stenosis might be beneficial, but further larger trials are required to determine whether there is benefit in this subgroup.

SAMMPRIS:NCT00576693;

VAST:ISRCTN29597900;

VIST:ISRCTN95212240.

RARE GENETIC VARIANTS IN PATIENTS WITH CERVICAL ARTERY DISSECTION

C. Traenka¹, M. Kloss², T. Strom³, P. Lyrer¹, C. Grond-Ginsbach² and S. Engelter¹

¹University Hospital Basel, Neurology and Stroke Center, Basel, Switzerland

²University Hospital Heidelberg, Department of Neurology, Heidelberg, Germany

³Institute of Human Genetics, Helmholtz Zentrum München, Munich, Germany

To identify pathogenic genetic variants associated with cervical artery dissection (CeAD).

Patients with spontaneous CeAD and either family history of CeAD (CeAD-f) or recurrent CeAD (CeAD-r) (≥3 months after initial CeAD) were identified in databases of the stroke centers Basel (CH) and Heidelberg (D) and analyzed by whole exome sequencing. First, we analyzed variants in a panel of 30 candidate genes associated with arterial dissection (any site) or aneurysm according the OMIM database. Second, we performed a genome-wide search for pathogenic variants regarding other vascular phenotypes possibly related to CeAD. We selected all gene variants with allele frequency < 0.05 in the in-house database of the Helmholtz Center Munich that were predicted as pathogenic by SIFT, Polyphen-2 or ClinVar affecting the same gene in at least two different pedigrees. Pathogenic genetic variants from the arterial dissection/aneurysm panel were considered CeAD-causing. Pathogenic genetic variants associated with other vascular phenotypes and variants of unknown significance in genes from the arterial dissection/aneurysm panel were considered as suggestive.

We included 43 patients (CeAD-f: n = 28, 17 pedigrees; CeAD-r: n = 15). Five pedigrees with CeAD-f carried CeAD-causing variants in COL3A1, COL4A1, COL4A3, COL4A4, COL5A1, COL5A2 and FBN1, no CeAD-causing variants were found in 15 patients with CeAD-r. Suggestive variants in ABCC6, COL3A1, COL5A2, COL5A2, MEF2A, and RNF213 were detected in three pedigrees with CeAD-f and six patients with CeAD-r.

Suggestive or CeAD-causing variants were detected in 14/32 affected families. CeAD-causing variants were more likely to be identified in patients with CeAD-f than in those with CeAD-r.

N/A

Scientific Communications 21- Endovascular Stroke Treatment – Management and Complications

POOLED-ANALYSIS OF GENERAL ANESTHESIA VERSUS CONSCIOUS SEDATION OR LOCAL ANESTHESIA IN STROKE ENDOVASCULAR THROMBECTOMY

D. Campbell¹, P.A. Barber²

¹Auckland City Hospital, Department of Anesthesia and Perioperative Medicine, Auckland, New Zealand

²University of Auckland, Medicine, Auckland, New Zealand

In studies of stroke endovascular thrombectomy, non-randomized post-hoc analyses have reported worse outcome after general anesthesia (GA) than local anesthesia or conscious sedation (LA/CS). Several studies that randomized thrombectomy patients to GA or LA/CS failed to show any difference in primary outcomes, but had sufficient efficacy signals to warrant further investigation.

A systematic review was performed to identify all studies that randomized thrombectomy patients to GA or LA/CS. A pooled analysis of the identified studies was then carried out to examine the rates of successful recanalization (TICI score of 2b or 3), and good functional recovery (modified Rankin Scale score 0–2) at 3 months.

Three single-centre studies (SIESTA, ANSTROKE and GOLIATH) that randomized 368 thrombectomy patients to GA (n = 183) or LA/CS (n = 185) were identified. In GA patients, a mean 6.6-minute door to groin puncture time was offset by an 11.8-minute shorter procedure time, resulting in a 6.1-minute shorter door to reperfusion time. Thrombectomy with GA was associated with a higher rate of successful recanalization (OR 1.94, 95% CI 1.13-3.34, heterogeneity chi2 = 0.44, p = 0.02) and good functional recovery at 3 months (OR 1.83, 95% CI 1.18-2.84, heterogeneity chi2 = 2.36, p = 0.007). For every 7.4 patients receiving GA, one more achieved a good functional outcome compared with those with CS/LA.

This small pooled analysis suggests that GA during thrombectomy does not harm patients. Indeed, there is a signal of benefit with a potentially large clinical effect size. It is important that the use of GA or CS/LA remains an active area of investigation.

N/A

GENERAL ANAESTHESIA VERSUS LOCAL ANAESTHESIA WITH OR WITHOUT CONSCIOUS SEDATION DURING ENDOVASCULAR THERAPY FOR ACUTE ISCHAEMIC STROKE: A MATCHED ANALYSIS FROM THE SWISS STROKE REGISTRY

B. Wagner¹ and L.H. Bonati¹

¹University of Basel, Neurology, Basel, Switzerland

Endovascular therapy (EVT) for stroke with large vessel occlusion reduces disability. Whether general anaesthesia (GA) during EVT is superior to local anaesthesia with or without conscious sedation (LA ± CS) remains controversial.

We included consecutive patients from the Swiss Stroke Registry receiving EVT between January 2014 and June 2017 at 8 Swiss Stroke Centres. The primary outcome was disability on the modified Rankin scale after 3 months, analysed with ordered logistic regression. Secondary outcomes included symptomatic intracranial haemorrhage (sICH), recurrent stroke, and death within 3 months. We did unadjusted comparisons, as well as coarsened exact matching (CEM) and propensity score matching (PSM) for relevant patient characteristics.

1468 of 1588 patients (92%) had follow-up at 3 months and were included (GA: n = 970, 64%; LA ± CS: n = 498, 36%). Patientsreceiving GA were older (mean age 72.3 versus 71.2 years) and had more severe strokes (mean NIHSS 15.7 versus 11.5) than patients treated with LA ± CS. GA was associated with higher disability at 3 months than LA ± CS (OR 1.8, 95% CI 1.4-2.1). The effect was essentially the same with matched comparisons (CEM: n = 396, OR 1.9, 1.1-10; PSM: n = 546, OR 1.7, 1.2-2.5). Mortality was also increased with GA. Symptomatic ICH did not differ consistently across models, nor did recurrent stroke.

This largest analysis on the impact of anaesthesia type on functional outcome in EVT to date showsthat GA is associated with greater disability and mortality after 3 months than LA ± CS. This effect appears to be independent of differences in patient characteristics.

N/A

IMPACT OF BLOOD PRESSURE DURING MECHANICAL THROMBECTOMY ACCORDING TO THE COLLATERAL STATUS: AN ASTER SUBSTUDY

B. Maier¹, M. Kyheng², J. Labreuche³, P. Moisiman⁴, F. Puccinelli⁵, C. Dargazanli⁶, G. Taylor⁷, R. Riem⁸, R. Bourcier⁹, J.P. Desilles¹⁰, R. Fahed¹, H. Redjem¹, S. Smajda¹, G. Ciccio¹, S. Escalard¹, R. Blanc¹, M. Piotin¹, B. Lapergue¹¹, M. Mazighi¹; On behalf of the ASTER Investigators

¹Rothschild Fondation, Interventional Neuroradiology, Paris, France

²CHU de Lille, EA 2694 – Santé publique: é pidémiologie et qualité des soins- F-59000, Lille, France

³CHU Lille, EA 2694 – Santé publique: é pidémiologie et qualité des soins- F-59000, Lille, France

⁴Bern University Hospital, Diagnostic and Interventional Neuroradiology Departement, Bern, Switzerland

⁵Lausanne Hospital, Diagnostic and Interventional Neuroradiology, Lausanne, Switzerland

⁶Montpellier Hospital, Interventional Neuroradiology, Montpellier, France

⁷Fondation Rothschild, Intensive Care Unit, Paris, France

⁸Nantes Hospital, Anesthesiology, Nantes, France

⁹Nantes Hospital, Interventional Neuroradiology, Nantes, France

¹⁰Fondation Rothschild, Interventional Neuroradiology, Paris, France

¹¹Foch Hospital, Neurology, Suresne, France

Guidelines regarding blood pressure (BP) management during mechanical thrombectomy (MT) for anterior ischemic strokes are lacking since optimal BP targets are a matter of debate. To better describe BP during MT, we sought to investigate the relationship between BP during MT with functional outcomes according to the collateral status (CS).

We performed a post-hoc analysis of the ASTER Trial (Contact Aspiration versus Stent Retriever for Successful Recanalization). BP was measured non-invasively during MT and CS was assessed at the beginning of MT (ASITN/SIR scale) by an independent core-lab. The primary outcome was favorable outcome defined as a shift in the direction of lower score on mRS at 3 months.

159 patients were included, 99 of whom had a poor CS and 60 a good CS. Systolic, diastolic BP and mean arterial pressure (MAP) variability, defined by the coefficient of variation, was negatively associated with favorable outcome regardless of CS: per 10-unit increase, adjusted OR = 0.45, 95% CI = 0.21-0.98; 0.37, 95% CI = 0.19-0.73; 0.37, 95% CI = 0.17-0.79, respectively. According to CS, per-procedural time with MAP < 90 mmHg was negatively associated with favorable outcome in patients with poor CS but not in patients with good CS (phet = 0.06): adjusted OR = 0.88, 95% CI = 0.76-1.01 and adjusted OR = 1.10, 95% CI = 0.91-1.34 for poor CS and good CS respectively.

BP variation during MT is associated with worse functional outcome regardless of CS but the hypotension time with MAP < 90 mmHg is associated with worse outcome for patients with poor CS but not with good CS. CS seems to be an interesting parameter to guide BP management during MT.

NA

THE EFFECT OF PROCEDURAL HEMODYNAMIC CHANGES DURING ANESTHETIC MANAGEMENT ON FUNCTIONAL OUTCOME AFTER ENDOVASCULAR TREATMENT FOR ISCHEMIC STROKE: RESULTS FROM THE MR CLEAN REGISTRY

N. Samuels¹, R.A. van de Graaf², C.A.L. van den Berg³, I. Eralp⁴, K.M. Treurniet⁵, B.J. Emmer⁵, R.V. Immink⁶, C.B.L.M. Majoie⁵, W.H. van Zwam⁷, B.A.A.M. van Hasselt⁸, J. Mühling⁹, H.F. Lingsma¹⁰, D.W.J. Dippel¹¹, A.C.G.M. van Es³ and A. van der Lugt³

¹Erasmus MC, Radiology and Nuclear Medicine- Neurology- Public Health, Rotterdam, The Netherlands

²Erasmus MC, Radiology and Nuclear Medicine- Neurology, Rotterdam, The Netherlands

³Erasmus MC, Radiology and Nuclear Medicine, Rotterdam, The Netherlands

⁴Erasmus MC, Anesthesiology, Rotterdam, The Netherlands

⁵Amsterdam UMC- location AMC, Radiology and Nuclear Medicine, Amsterdam, The Netherlands

⁶Amsterdam UMC- location AMC, Anesthesiology, Amsterdam, The Netherlands

⁷Maastricht UMC, Radiology and Nuclear Medicine, Maastricht, The Netherlands

⁸Isala, Radiology and Nuclear Medicine, Zwolle, The Netherlands

⁹Radboud UMC, Anesthesiology, Nijmegen, The Netherlands

¹⁰Erasmus MC, Public Health, Rotterdam, The Netherlands

¹¹Erasmus MC, Neurology, Rotterdam, The Netherlands

Conscious sedation (CS) during endovascular treatment (EVT) may be associated with worse functional outcomes compared to local anesthesia (LA). We evaluated whether the effect of anesthetic management on functional outcome could be explained by procedural hemodynamic changes.

Patients treated with EVT for acute ischemic stroke in the MR CLEAN Registry between March 2014 and November 2017 in centers with CS or LA as standard protocol and blood pressure data available were analyzed. We evaluated whether the effect of anesthetic management on functional outcome (90-days modified Rankin Scale) changed after adjustment for duration (area under the threshold (AUT), 10%-drop from baseline in mmHg/minute) or depth (lowest mean arterial pressure from baseline in mmHg (delta-LMAP)) of hypotensive episodes.

In 230 patients treated with EVT, hypotensive episodes were longer and deeper under CS (n = 139) compared to LA (n = 91) (median AUT 1.99 [IQR 0–9.38] vs. 0 [IQR 0–0.39], p < 0.01 and median delta-LMAP 24.67 [IQR 9.70-40.33] vs. 2.33 [IQR 0–16.33], p < 0.01). CS was associated with worse functional outcome compared to LA (adjusted common odds ratio (acOR) 0.33, 95% CI 0.15-0.76). This association did not change after adjustment for AUT (acOR 0.41, 95% CI 0.15-1.11) or delta-LMAP (acOR 0.32, 95% CI 0.13-0.79).

CS was associated with poorer functional outcome and more hemodynamic changes than LA. However, hemodynamic changes did not explain the association between CS and poor outcome.

N/A

BLOOD PRESSURE COURSE AND ITS ASSOCIATION WITH OUTCOME AFTER TREATMENT WITH ENDOVASCULAR THROMBECTOMY IN PATIENTS WITH ACUTE OCCLUSIVE STROKE

M. Matusevicius¹, C. Cooray¹, M. Bottai², M. Mazya¹, G. Tsivgoulis³, A.P. Nunes⁴, T. Moreira¹, J. Ollikainen⁵, R. Tassi⁶, D. Strbian⁷, S. Holmin^1,8 and N. Ahmed^1,9

¹Karolinska Institute, Clinical Neuroscience, Stockholm, Sweden

²Karolinska Institute, Institute of Environmental Medicine, Stockholm, Sweden

³National and Kapodistrian University of Athens School of Medicine- Athens, Second Department of Neurology in National & Kapodistrian, Athens, Greece

⁴Centro Hospitalar Lisboa Central, Stroke Unit, Lisbon, Portugal

⁵Tampere University, Neurosciences and Rehabilitation, Tampere, Finland

⁶AOU Senese, Stroke Unit, Sienna, Italy

⁷Helsinki University Central Hospital, Neurology, Helsinki, Finland

⁸Karolinska University Hospital, Neuroradiology, Stockholm, Sweden

⁹Karolinska University Hospital, Neurology, Stockholm, Sweden

Optimal level for systolic blood pressure (SBP) during initial 24 hours after endovascular thrombectomy (EVT) in acute ischemic stroke is unknown; we aimed to investigate this.

We included EVT-treated patients registered 2014–2017 in the SITS International Thrombectomy Registry. Mean SBP during initial 24 hours post-EVT was analyzed as a continuous variable and in intervals (90-120, 120–140, 140–160, 160–180, and >180 mmHg). For each SBP interval, the remaining study population was used as reference. Analyses were performed separately per recanalization grade: successful (mTICI≥2b) or unsuccessful (mTICI < 2b). Primary outcome was 3-month modified Rankin Scale (mRS) score 0–2. Secondary outcomes were SICH, and 3-month death and shift in mRS score distribution. Results were adjusted for known confounders such as age and baseline NIHSS.

In patients with successful recanalization (n = 2920), higher SBP as a continuous variable was associated with unfavorable outcomes. In analysis of intervals, SBP 160–180 was associated with unfavorable outcome (mRS 0–2: adjusted odds ratio [aOR] 0.45, 95% CI 0.27-0.71; SICH: aOR 3.51, 1.51-7.57) and SBP 120–140 was associated with favorable outcomes (mRS 0–2: aOR 1.28, 1.02-1.60; death: aOR 0.47, 0.34-0.65). mRS shift analysis mirrored this pattern. In patients with unsuccessful recanalization (n = 711), higher SBP as a continuous variable was only associated with more SICH, while no SBP interval showed any significant association with the outcomes.

Mean 24-hour SBP 120–140 was associated with favorable outcomes in EVT with successful recanalization, whereas SBP 160–180 was associated with unfavorable outcomes. In patients with unsuccessful recanalization, higher SBP was only associated with more SICH.

N/A

CEREBRAL EDEMA IN PATIENTS PRESENTING WITH LARGE HEMISPHERIC ISCHEMIC STROKE UNDERGOING ENDOVASCULAR THROMBECTOMY VERSUS MEDICAL THERAPY: A META-ANALYSIS OF INDIVIDUAL PATIENT DATA

F. Ng¹, W.T. Kimberly², G. Sharma¹, S. Brown³, M. Goyal⁴, C.B.L.M. Majoie⁵, T. Jovin⁶, M. Hill⁷, K. Muir⁸, J. Saver⁹, F. Guillemin¹⁰, A. Demchuk⁷, P. White¹¹, D. Dippel¹², A. Davalos¹³, S. Bracard¹⁴, P. Mitchell¹⁵, M. Wald¹⁶, K. Sheth¹⁷ and B. Campbell¹

¹Royal Melboune Hospital, Department of Medicine and Neurology- Melbourne Brain Centre, Parkville, Australia

²Harvard Medical School, Division of Neurocritical Care, Boston, USA

³Altair Biostatistics, Altair Biostatistics-, St Louis Park, USA

⁴University of Calgary, Department of Radiology, Calgary, Canada

⁵Academic Medical Center- Amsterdam, Department of Radiology and Nuclear Medicine, Amsterdam, The Netherlands

⁶University of Pittsburgh Medical Center, Stroke Institute- Department of Neurology, Pittsburgh, USA

⁷University of Calgary, Department of Clinical Neurosciences, Calgary, Canada

⁸Queen Elizabeth University Hospital, Institute of Neuroscience and Psychology, Glasgow, United Kingdom

⁹University of California Los Angeles, Department of Neurology and Comprehensive Stroke Center- David Geffen School of Medicine, Los Angeles, USA

¹⁰University of Lorraine and University Hospital of Nancy, Clinical Investigation Centre— Clinical Epidemiology, Nancy, France

¹¹Newcastle University, Institute of Neuroscience, Newcastle upon Tyne, United Kingdom

¹²Erasmus MC University Medical Center, Department of Neurology, Rotterdam, The Netherlands

¹³Hospital Germans Trias i Pujol- Universitat Autònoma de Barcelona, Department of Neuroscience, Barcelona, Spain

¹⁴University of Lorraine and University Hospital of Nancy, Department of Diagnostic and Interventional Neuroradiology, Nancy, France

¹⁵Royal Melbourne Hospital, Department of Radiology, Parkville, Australia

¹⁶Biogen, Clinical Research, Cambridge- MA, USA

¹⁷Yale New Haven Hospital, Department of Neurology, New Haven, USA

Cerebral edema is a devastating complication of large hemispheric infarction (LHI). We investigated the impact of cerebral edema on functional outcome; and the effects of endovascular thrombectomy (EVT) and reperfusion on cerebral edema in LHI patients.

Pooled individual patient data from seven randomised, global EVT trials (HERMES collaboration) were included if there was evidence of LHI on pre-treatment imaging, defined as diffusion-MRI or CT-perfusion core lesion volume 80–300mL or CT-ASPECTS 0–5. Reperfusion was defined as angiographic modified-TICI 2b/3 score. The primary outcomes were maximum midline shift (MLS) on follow-up imaging 1–14 days after onset (analysed using multivariable linear regression) and functional improvement (≥1 point on 90-day modified Rankin Scale, ordinal logistic regression).

Among 1764 patients, 179 patients had pre-treatment LHI. Median MLS was 5.0mm (IQR 2.8,8.6) in thrombectomy and 5.5mm (IQR 3.1,8.1) in control patients (p = 0.72). MLS correlated with follow-up infarct volume (Spearman’s rho = 0.57, p < 0.001) and was associated with lower likelihood of functional improvement (common odds ratio per mm MLS = 0.83[95% CI 0.77,0.89], p < 0.001) when adjusting for age, NIHSS, glucose, CT/MR, time to follow-up imaging and treatment group. Thrombectomy (beta = 0.10[95% CI -1.15,1.35], p = 0.87) and reperfusion (beta = -0.66[95% CI -2.96,1.64], p = 0.58) were not significantly associated with MLS when adjusting for age, NIHSS, glucose, CT/MR and time to follow-up imaging. In a subgroup analysis of patients presenting with core volume>130mL or CT-ASPECTS≤3 (n = 76), EVT was associated with greater MLS (9.3 ± 5.0 vs 6.7 ± 4.5, unadjusted p = 0.03, adjusted p = 0.07).

In LHI patients, cerebral edema is associated with worse outcome. EVT and reperfusion do not appear to reduce MLS accumulation.

N/A

EFFECT OF RECANALISATION ON DEVELOPMENT OF CEREBRAL OEDEMA IN ACUTE ISCHAEMIC STROKE TREATED WITH IVT: DATA FROM SITS-ISTR

M. Thorén¹, A. Dixit², Z. Gdovinová³, L. Klecka⁴, D. Toni⁵, A. Vilionskis⁶, N. Wahlgren⁷ and N. Ahmed¹

¹Karolinska institutet Department of Clinical Neuroscience, Karolinska University Hospital Department of Neurology/Tema neuro, Stockholm, Sweden

²University of Newcastle upon Tyne, Newcastle upon Tyne NHS Foundation Trust, Newcastle upon Tyne, United Kingdom

³P.J. Safarik University Košice, Department of Neurology Faculty of Medicine, Košice, Slovak Republic

⁴Municipal hospital of Ostrava, Departement of Neurology, Ostrava, Czech Republic

⁵University La Sapienza Rome, Unità di Trattamento Neurovascolare, Rome, Italy

⁶Institute of Clinic Neurology Vilnius University, Republican Vilnius University Hospital Department of Neurology, Vilnius, Lithuania

⁷Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden

Extensive infarction and expanding cerebral oedema (COED) due to a middle cerebral artery occlusion confers 70% mortality unless treated surgically. Reperfusion may cause blood brain barrier disruption and promote cerebral oedema and secondary parenchymal haemorrhage (PH). We aimed to investigate the association of recanalisation with development of early COED and PH after recanalisation therapy.

From SITS-International Stroke Treatment Registry, we selected patients with artery occlusion at baseline, denoted by Hyperdense Artery Sign or CT/MRI angiographic occlusion. We inferred recanalisation from resolution of radiological signs of occlusion at 24 hours. Primary outcome was COED and secondary outcome was PH type 1 or 2 detected on 24h scans. Logistic regression with adjustment for baseline variables was used.

22184 patients fulfilled the inclusion criteria (n = 18318 received IVT, n = 3071 received IVT + thrombectomy, n = 795 received thrombectomy alone). Recanalisation was inferred in 64.1%. Median age was 71 vs. 71 years and NIHSS 15 vs. 16 in recanalised vs. non-recanalised patients respectively. COED was less (29.4% vs. 44.6%, p < 0.05) and PH was more (8.92% vs. 6.52%, p < 0.05) frequent in recanalised than non-recanalised patients. Multivariable logistic regression showed that recanalisation was associated with a lower risk for COED (OR 0.53, 95% CI 0.49-0.56) and a higher risk for PH (OR 1.46, 95% CI 1.29-1.64).

In patients with ischaemic stroke, we observed a reduced risk of COED and increased risk of PH in recanalised patients compared to non-recanalised patients at 24h.

N/A

QUANTIFICATION OF 24-HOUR DIFFUSION-WEIGHTED IMAGING LESION REVERSAL AFTER THROMBECTOMY

R. Meier¹, P. Lux¹, R. McKinley¹, S. Jung², U. Fischer², J. Gralla¹, R. Wiest¹ and J. Kaesmacher¹

¹University Hospital Inselspital, Support Center for Advanced Neuroimaging- University Institute of Diagnostic and Interventional Neuroradiology, Bern, Switzerland

²University Hospital Inselspital, Department of Neurology, Bern, Switzerland

The purpose of this study was to quantify the extent of diffusion-weighted imaging (DWI) lesion reversal in ischemic stroke within 24 hours after thrombectomy.

Patients who underwent thrombectomy were retrospectively included. A total of 120 patients were included in the study, 5 patients were excluded due to poor image quality and another 5 due to absence of an ischemic core. Acute and 24-hour DWI data were co-registered. The ischemic core, 24-hour DWI lesion, and the affected gray/white matter regions were annotated manually. DWI reversal was estimated through voxel-wise analysis.

Of the 110 patients, 90 were successfully reperfused (> = TICI 2b, group I) and exhibited a significantly lower 24-hour DWI lesion volume (median 9.5mL, IQR 2.7–21.9mL) than the remaining 20 patients with unsuccessful reperfusion (<TICI 2b, group II; median 23.9mL, IQR 9.1–45.6mL, p < 0.01). Patients of group I showed larger DWI reversal than those of group II (median 3.1mL versus 1.9mL, IQR 1.4–10.5mL versus 0.9–7mL, p = 0.19) with a larger fraction of reversed tissue being white matter (median 50.9% versus 44.9%, IQR 39.5–63.6% versus 29.1–54.2%, p = 0.15). Twenty-four patients (24/90, 26.7%) showed absolute and relative reversal of >10 mL and >10%, respectively. Patients with successful reperfusion within 4.5 hours of symptom onset (n = 52) showed significantly larger reversal (median 5.2mL, IQR 2.1–13.8mL) than patients treated beyond 4.5 hours (n = 29, median 1.4mL, IQR 0.6–2.9mL, p < 0.01).

Considerable 24-hour DWI reversal occurs in about one quarter of thrombectomy patients with successful reperfusion, preferably within 4.5 hours from onset.

N/A

Scientific Communications 22-Risk Factors and Prognosis

THE IMPACT OF TRANSIENT NEUROLOGICAL ATTACKS ON LIFE EXPECTANCY AND YEARS LIVED WITH STROKE AND CORONARY HEART DISEASE

A. Heshmatollah¹, K. van der Willik², P.J. Koudstaal³, M.A. Ikram² and M.K. Ikram¹

¹Erasmus MC University Medical Center, Epidemiology and Neurology, Rotterdam, The Netherlands

²Erasmus MC University Medical Center, Epidemiology, Rotterdam, The Netherlands

³Erasmus MC University Medical Center, Neurology, Rotterdam, The Netherlands

Transient neurological attacks (TNA) are associated with an increased risk of stroke and coronary heart disease (CHD). However, their impact on the remaining life expectancy and years lived with and without stroke or stroke and CHD combined (ASCVD; atherosclerotic cardiovascular disease) is unknown.

Between 1990 and 2016, 1,202 stroke-free and CHD-free participants from the prospective population-based Rotterdam Study aged ≥ 65 years suffered a first-ever TNA. These persons were matched to participants without incident TNA using age, sex, and birth year. To calculate the remaining life expectancy with and without stroke/ASCVD, multi-state life tables were constructed using three transitional states: TNA/no TNA to stroke/ASCVD, TNA/no TNA to death, and stroke/ASCVD to death.

The overall life expectancy between participants with and without incident TNA was comparable (18.4 vs. 18.2 years at age 65). However, participants with incident TNA spent twice as many remaining life years with disease. At age 65, persons with incident TNA spent on average 2.0 years with stroke and 3.2 years with ASCVD compared to 1.1 and 2.1 years for participants of the same age without incident TNA. Similar results were seen across all ages and for both men and women (Figure 1).

TNA does not shorten the overall life expectancy but has a large impact on the loss of stroke-free and ASCVD-free life years. More focus on secondary prevention in TNA patients could potentially decrease these years lived in disability.

N/A

INCREASED LONG-TERM RISK OF INTRACEREBRAL HAEMORRHAGE AFTER TRANSIENT NEUROLOGICAL ATTACKS: A POPULATION-BASED STUDY

M.A. Tuna¹, P.M. Rothwell¹; on behalf of the Oxford Vascular Study

¹Oxford University, Nuffield Department of Clinical Neuroscience, Oxford, United Kingdom

There is a substantial evidence-base for management of definite (NINDS-criteria) TIA, but the two other main categories of transient neurological events of likely cerebrovascular aetiology have usually been excluded from trials: transient isolated mono-symptomatic syndromes (TIMS; e.g. isolated diplopia, isolated vertigo, isolated bilateral visual loss, isolated dysarthria etc) that do not fit the NINDS-criteria for TIA; and non-focal transient neurological attacks (TNA), sometimes associated with cerebral amyloid angiopathy (amyloid spells). Both types of events have an uncertain prognosis and there is great variation in management in practice, particularly in recommendations for life-long antiplatelet treatment. We compared the risks of intracerebral haemorrhage and ischaemic stroke in TNAs and TIMS.

All patients who sought medical attention after sudden onset transient neurological symptoms were ascertained prospectively in a population of 92,728 in Oxfordshire, UK from 2002–18. Events were classified at baseline as NINDS-TIA, TIMS and TNA. Stroke risk was determined by face-to-face follow-up to 10-years.

Among 2142 consecutive patients (mean age = 72.2), 984 had NINDS-TIA, 520 had TIMS, 638 had TNA. Use of antithrombotics was similar after TNA and TIMS (75.8% vs 81.0%; p = 0.47), but the 10-year risk of ICH was higher in TNAs than in TIMS (17-events vs 2-events; 3.9%, 95% CI 1.9-5.9 vs 0.5%, 0.09-1.1, p = 0.002), whereas the 10-year risk of ischaemic stroke was similar (87 vs 57; 18.9%, 15–22.8 vs 14.7%, 10.8-18.6, p = 0.17).

TNAs have an increased long-term risk of ICH compared with TIMS. The balance of risk and benefit from life-long antiplatelet treatment might therefore differ.

N/A

DIFFERENTIAL RISK FACTOR PROFILE AMONG YOUNG STROKE PATIENTS OF DIFFERENT ETHNICITIES: THE GLOBAL OUTCOME AFTER STROKE AT YOUNG AGE (GOAL)-INITIATIVE

M. Jacob¹, M.S. Ekker¹, M. van Dongen¹, K. Aarnio², A. Annamalai³, A. Arauz⁴, M. Arnold⁵, M. Barboza⁶, M. Bolognese⁷, R. Brouns⁸, B. Chuluun⁹, E. Chuluunbaatar¹⁰, B. Dagvajantsan⁹, S. Debette¹¹, A. Don¹², C. Enzinger¹³, E. Ekizoglu¹⁴, S. Fandler¹⁵, F. Fazekas¹⁵, A. Fromm¹⁶, T. Gattringer¹⁵, G. Gulli¹⁷, M. Hoffmann¹⁸, T. Hora¹⁹, C. Jern²⁰, K. Jood²¹, M. Kamouchi²², Y. Kim²³, T. Kitazono²⁴, S. Kittner²⁵, T. Kleinig²⁶, K. Klijn¹, J. Körv²⁷, T. Lee²⁸, D. Leys²⁹, N. Maaijwee⁷, N. Martinez-Majander², J. Marto³⁰, M. Mehndiratta³¹, V. Mifsud³², V. Montanaro¹⁹, M. Owolabi³³, V. Patel³⁴, M. Phillips³⁵, B. Piechowski-Iozwiak³², A. Pikula³⁶, J. Ruíz-Sandoval³⁷, B. von Sarnowski³⁸, F. Schreuder¹, R.H. Swartz⁴⁰, K.S. Tan⁴¹, D. Tanne⁴², T. Tatlisumak², V. Thijs²⁶, A.M. Tuladhar¹, M. Viana-Baptista³⁰, R. Vibo²⁷, T.Y. Wu⁴³, N. Yeşilot¹⁴, U. Waje-Andreassen¹⁶, A. Pezzini⁴⁴, J. Putaala², F.E. de Leeuw¹; GOAL-team

¹Radboud University Medical Centre- Donders Institute for Brain- Cognition and Behaviour, Department of Neurology, Nijmegen, The Netherlands

²Helsinki University Hospital, Department of Neurology, Helsinki, Finland

³North Tees and Hartlepool, NHS Foundation trust, Stockton on Tees, United Kingdom

⁴National Institute of Neurology and Neurosurgery of Mexico, Manuel Velasco Suárez, Mexico city, Mexico

⁵Inselspital University Hospital- University of Bern, Department of Neurology, Bern, Switzerland

⁶Hospital Dr. Rafael A. Calderon Guardia, Neurosciences Department, San Jose, Costa Rica

⁷Lucerne Cantonal Hospital, Neurocenter- Department of Neurology and Neurorehabilitation, Luzern, Switzerland

⁸Vrije Universiteit Brussel, The Faculty of Medicine and Pharmacy, Jette, Belgium

⁹Mongolian National Univeristy of Medical Sciences, International School of Traditional Medicine, Uluunbaatar, Mongolia

¹⁰Taipei Medical University, Depatrment of Global Health and Development, Taipei, Taiwan R.O.C

¹¹Inserm, Bordeaux Population Health Research Centre, Bordeaux, France

¹²Tel Aviv University, School of Medicine- Sackler Faculty of Medicine, Tel Aviv, Israel

¹³Medical University of Graz, Department of Neurology- Department of Radiology- division of neuroradiology, Graz, Austria

¹⁴Istanbul University- Faculty of Medicine, Department of Neurology, Istanbul, Turkey

¹⁵Medical University of Graz, Department of Neurology, Graz, Austria

¹⁶Haukeland University Hospital, Centre for Neurovascular Diseases- Department of Neurology, Bergen, Norway

¹⁷Ashford and St Peter’s Hospitals, Department of Stroke Medicine, Chertsey, United Kingdom

¹⁸University of Central Florida, Stroke Center- Orlando VA Medical Center, Orlando, USA

¹⁹SARAH Hospital of Rehabilitation, Department of Neurology, Brasilia, Brazil

²⁰University of Gothenburg, Insitute of Biomedicine- the Sahlgrenska Academy, Gothenburg, Sweden

²¹University of Gothenburg, Institute of Neuroscience and physiology- the Sahlgrenska Academy, Gothenburg, Sweden

²²Kyushu University, Department of Health Care Administration and Management- Center for Cohort Studies- Graduate School of Medical Sciences, Fukuola, Japan

²³Hanyang University, Department of Neurology- College of Medicine, Seoul, Republic of Korea

²⁴Kyushu University, Department of Medicine and Clinical Science- Graduate School of Medical Sciences- Center for Cohort Studies, Fukuola, Japan

²⁵University of Maryland School of Medicine, Department of Neurology- Veterans Affairs Maryland Health Care System, Baltimore, USA

²⁶Royal Adelaide Hospital, Department of Neurology, Adelaide, Australia

²⁷University of Tartu, Department of Neurology and Neurosurgery, Tartu, Estonia

²⁸Chang Gung Memorial Hospital- Linkou Medical Center- and College of Medicine- Chang Gung University, Department of Neurology, Taoyuan, Taiwan R.O.C

²⁹University of Lille, Department of Neurology, Lille, France

³⁰Universidade Nova de Lisboa, Department of Neurology- Hospital Egas Moniz- Centro Hospitalar Lisboa Ocidental and CEDOC- Nova Medical School, Lisbon, Portugal

³¹G.B. Pant Hospital, Department of Neurology, New Delhi, India

³²Cleveland Clinic Abu Dhabi, Neurological Institute, Abu Dhabi, United Arab Emirates

³³University of Ibadan, Department of Medicine- College of Medicine, Ibadan, Nigeria

³⁴Inkosi Albert Luthuli Central Hospital, Department of Neurology- Nelson R. Mandela School of Medicine, Mayville, South Africa

³⁵Waikato Hospital, Department of Neurology, Hamilton, New Zealand

³⁶University of Toronto, Department of Medicine Neurology- University Health Network, Toronto, Canada

³⁷Hospital Civil de Guadalajara Fray Antonio Alcalde, Department of Neurology, Guadalajara, Mexico

³⁸University Medicine, Department of Neurology, Greifswald, Germany

³⁹Hospital Civil de Guadalajara Fray Antonio Alcalde, Department of Neurology, Guadalajara, Mexico

⁴⁰Sunnybrook Health Sciences Centre, Department of Medicine Neurology, Toronto, Canada

⁴¹University of Malaya, Department of Medicine, Kuala Lumpur, Malaysia

⁴²Tel Aviv University, Department of Epidemiology and Preventive Medicine, School of Public Health, Sackler Faculty of Medicine, Tel-Hashomer, Israel

⁴³Christchurch Hospital, Department of Neurology, Christchurch, New-Zealand

⁴⁴University of Brescia, Department of Clinical and Experimental Sciences, Neurology Clinic, Brescia, Italy

The incidence of young stroke is increasing worldwide, with still an unknown remaining cause in 30–40% of all cases. Exploring ethnic differences in risk factors may increase our understanding of the contribution of risk factors in young stroke patients and give rise to new preventive, individually tailored strategies.

Patients were participants of the GOAL-study, a worldwide multicenter individual patient data meta-analysis based on existing data from young stroke cohorts. All patients aged 18–50 years with ischemic stroke were included. Outcome was distribution of vascular risk factors, as defined according to the 2014 guidelines of the American Stroke Association, stratified for ethnicity. Logistic regression was used to adjust for age and sex.

A total of 10,036 patients (males 56.0%) were analyzed. Mean-age was 39.5 ± 7.7 years. The most prevalent risk factors varied by ethnicity with hypertension being most prevalent among Asians (56.6%); diabetes among Black Africans (20.6%) and dyslipidemia and obesity among Caucasians (40.1% and 15.8% respectively) (Figure 1). When compared to Caucasians, hypertension was significantly more prevalent in Asians (odds ratio (OR) 2.7, 95% CI 2.3-3.2, p < 0.0001) and diabetes more in Black Africans (OR 2.9, 95% CI 2.3-3.7, p < 0.0001). Dyslipidemia was significantly more found in Caucasians, compared to Blacks (OR 1.8, 95% CI 1.1-2.9, p < 0.0001), Hispanics (OR 4.2, 95% CI 3.6-4.9, p < 0.0001) and Asians (OR 1.6, 95% CI 1.4-1.9, p < 0.0001).

The prevalence of conventional risk factors differ significantly among young stroke patients of different ethnicities. These differences could be used to explore differences in etiology and to tailor preventive strategies.

N/A

THE ROLE OF HAEMATOLOGICAL TRAITS IN STROKE AND ITS SUBTYPES

E. Harshfield¹, M. Sims^2,3, M. Traylor¹, W. Astle^4,5, W. Ouwehand^2,4,5,6,7 and H. Markus¹

¹University of Cambridge, Department of Clinical Neurosciences, Cambridge, United Kingdom

²University of Cambridge, Department of Haematology, Cambridge, United Kingdom

³Oxford University Hospitals NHS Foundation Trust, Oxford Haemophilia and Thrombosis Centre, Oxford, United Kingdom

⁴University of Cambridge, Department of Public Health and Primary Care, Cambridge, United Kingdom

⁵University of Cambridge, NHS Blood and Transplant, Cambridge, United Kingdom

⁶University of Cambridge, BHF Cambridge Centre of Cardiovascular Research Excellence, Cambridge, United Kingdom

⁷Wellcome Trust Sanger Institute, Department of Human Genetics, Hinxton, United Kingdom

Thrombosis and platelet activation play a central role in stroke pathogenesis, and antiplatelet and anticoagulant therapies are central to stroke prevention. However, whether such haematological traits contribute equally to all ischaemic stroke subtypes is uncertain. Furthermore, identification of associations with new traits may offer novel treatment opportunities. We used genetic data to ascertain causal relationships between a wide range of haematological traits and ischaemic stroke and its subtypes.

We obtained summary statistics from 25 published GWAS of haematological traits involving 350,000 individuals, and genetic associations with stroke from MEGASTROKE (N = 520,000). Using two-sample Mendelian randomisation we analysed associations of genetically elevated levels of 36 blood cell traits and 48 haemostasis traits with risk of ischaemic (IS), cardioembolic (CES), large-artery (LAS), and small-vessel stroke (SVS).

Several factors on the intrinsic clotting pathway were significantly associated (P < 1.111x10⁻⁴) with CES and LAS, but not with SVS (e.g. reduced FVIII activity with IS/CES/LAS; raised FVIII antigen with IS/CES; increased FXI activity with IS/CES). Elevated plateletcrit was associated with IS and CES.

These results implicate the intrinsic pathway in pathogenesis of CES and possibly LAS, but not SAS. Plateletcrit may also have a direct causal effect on ischaemic stroke risk. The lack of association with SAS suggests thrombosis may be less important for this stroke subtype. Further investigation of these traits may lead to better understanding of the role of thrombosis in specific stroke subtypes and identification of targets for novel preventative drugs. (Funding: EU Horizon 2020 grant agreement No 667375 – CoSTREAM.)

N/A

INNATE IMMUNITY AND THE RISK OF STROKE: A PROSPECTIVE COHORT STUDY

L. Fani¹, K.D. van der Willik¹, M.J.G. Leening¹, M. Kavousi¹, M.A. Ikram¹ and M.K. Ikram¹

¹Erasmus MC, Epidemiology, Rotterdam, The Netherlands

The inflammatory hypothesis of atherothrombosis has gained support by trials targeting the innate immunity pathway of cardiovascular disease. Yet, the role of innate immunity for developing stroke remains unknown. We determined the association of components of immunity with development of stroke in the general population.

Between 2002–2015, we measured granulocytes and platelets (innate immunity) and lymphocytes (adaptive immunity) and their derived ratios (granulocyte-to-lymphocyte ratio [GLR], platelet-to-lymphocyte ratio [PLR], and systemic immune-inflammation index [SII]), reflecting an imbalance towards innate immunity. These blood cell counts were measured three times between 2002–2015 in stroke-free and coronary heart disease-free participants of the population-based Rotterdam Study and followed-up for stroke (until 2015). Joint models were used that enabled us to study the longitudinal evolution of the various markers during the pre-clinical phase.

Of 7,734 participants (median age 64.3 years, 67.2% women), with a median follow-up of 8.1 years, 378 participants developed a stroke. Doubling of granulocyte and platelet counts were associated with an increased stroke risk (HR [95% CI]: 1.67 [1.13– 2.48] and 1.32 [0.91–1.92], respectively), whereas lymphocytes were associated with a decreased stroke risk (0.89 [0.66–1.20]). Doubling of the derived ratios GLR, PLR, and SII were all associated with an increased stroke risk (HR [95% CI]: 1.35 [1.03–1.75], 1.21 [0.95–1.54], and 1.20 [0.98–1.47], respectively).

An increase of innate immunity markers GLR, PLR, and SII are associated with an increased risk of stroke in the general population, suggesting that the innate immunity pathway plays an important role in the pathogenesis of stroke.

N/A

PRO-BNP: A NOVEL PREDICTOR OF STROKE RISK AFTER TRANSIENT ISCHEMIC ATTACK

E. Rodríguez Castro¹, P. Hervella², I. Ló pez-Dequidt¹, S. Arias-Rivas¹, M. Santamaría-Cadavid¹, I. Ló pez-Loureiro², A. da Silva-Candal², M. Pérez-Mato², T. Sobrino², F. Campos², J. Castillo², M. Rodríguez-Yáñez¹ and R. Iglesias-Rey²

¹Hospital Clínico Universitario de Santiago, Stroke Unit- Neurology Department, Santiago de Compostela, Spain

²Health Research Institute of Santiago de Compostela IDIS- Hospital Clínico Universitario, Clinical Neurosciences Research Laboratory, Santiago de Compostela, Spain

Elevated levels of B-type natriuretic peptide (BNP) and its prohormone (pro-BNP) can predict an increased risk of cardiovascular events and ischemic stroke. The limited reliability to predict the risk of stroke after a transient ischemic attack (TIA) justifies the objective of our study to determine the role of pro-BNP in patients with TIAs.

From our prospective stroke registry, we performed a retrospective study in all patients with the diagnosis of TIA admitted to the Stroke Unit of our Hospital between January 2008 and March 2018. Pro-BNP was determined in the first hours after TIA. The endpoint was the development of stroke during the follow-up.

381 patients were included. Mean time of follow-up was 36.8 ± 16.4 months. 224 patients were hospitalized due to a stroke during the follow-up, and 157 were not. Pro-BNP serum levels were higher in patients who suffered a stroke compared to those who did not (p < 0.001). We also found greater levels of this marker the earlier the stroke happened (p = 0.024). A cut-off point of 800 pg/mL of pro-BNP predicted a stroke with a sensitivity of 64% and a specificity of 79% (p < 0.001), and was independently associated with higher risk of stroke after a TIA (OR: 6.65, p < 0.001). This association persisted for different etiopathogenic TIA groups (cardioembolic: OR 26.12, p < 0.001; undetermined: OR 4.87, p = 0.006; atherothrombotic: OR 1.67, p = 0.044).

The early determination of pro-BNP is a simple and very useful alternative to predict the prognosis after TIA regardless of the etiopathogenesis of the TIA.

N/A

APPLICATION OF RISK PREDICTION MODELS TO IMPROVE EFFICACY OF SCREENING TO DETECT ASYMPTOMATIC CAROTID STENOSIS

M. Poorthuis^1,2, S. Massa¹, P. Sherliker^1,3, R. Clack¹, R. Clarke¹, A. Halliday⁴, G. de Borst², R. Bulbulia^1,3 and S. Lewington^1,3

¹Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, Oxford, United Kingdom

²University Medical Center Utrecht, Department of Vascular Surgery, Utrecht, The Netherlands

³MRC Population Health Research Unit, Nuffield Department of Population Health, Oxford, United Kingdom

⁴John Radcliffe Hospital, Nuffield Department of Surgical Sciences, Oxford, United Kingdom

Prediction models for the detection of asymptomatic carotid stenosis (ACS) might allow more effective screening. We aim to identify existing prediction models for individualized risk estimation of the prevalence of ACS and externally validate them in a screening population.

We conducted a systematic search of PubMed and EMBASE for prediction models of the prevalence of ACS ≥50%. We externally validated them using data of a subset of 595,943 individuals attending screening clinics for carotid screening in the US and UK. We assessed predictive performance using c-statistics and calibration plots.

Six risk prediction models were identified. Three used moderate and three used severe ACS as predicted outcome. Included predictors were age, sex, smoking, hypertension, hypercholesterolemia, diabetes, (cerebro) vascular disease, height, measured blood pressure, and blood lipids. In the external validation cohort, 7,021 (1.18%) participants had moderate or severe (50-99%) ACS and 1,192 (0.21%) had severe (70-99%) stenosis. The c-statistics ranged only from 0.67 to 0.75 for moderate ACS, and from 0.66 to 0.78 for severe ACS. Comparing the lowest to the highest decile of the predicted risk in the best model, the observed prevalence ranged from 0.1% to 4.1% for moderate ACS, and from 0.01% to 1.1% for severe ACS.

Current prediction models can reliably identify a subgroup of patients in whom screening might be worthwhile (number needed to scan≈25). However, clinical applicability is hampered by the low prevalence of ACS. Future models should include better predictors or should be developed in populations at higher risk for ACS.

N/A

GENETICALLY DETERMINED LEVELS OF THE HUMAN PLASMA PROTEOME AND RISK OF STROKE SUBTYPES: A MENDELIAN RANDOMISATION STUDY

L. Chen¹, J. Peters¹, B. Prins¹, A.S. Butterworth¹, M. Traylor², E. Yonova-Doing¹, P. Surendran¹, H.S. Markus² and J.M.M. Howson¹

¹University of Cambridge, MRC/BHF- Cardiovascular Epidemiology Unit, Cambridge, United Kingdom

²University of Cambridge, Stroke Research Group- Department of Clinical Neurosciences, Cambridge, United Kingdom

Stroke has very few treatments or established prevention strategies. Although conventional risk factors, such as hypertension, account for some stroke risk, much remains unexplained. Many blood biomarkers have been associated with stroke but their aetiological role is unclear. In this study, we investigated the causal association of plasma proteins with stroke subtypes through Mendelian randomisation (MR).

Genome Wide Association Studies (GWAS) were performed for each of 241 plasma proteins (measured using O-link assays), in up to 4,996 blood donors from the INTERVAL study. Conditionally independent variants associated with each protein were used as instrumental variables and their associations with stroke subtypes were obtained from MEGASTROKE (>40,000 cases). We used Inverse-variance-weighted (IVW) MR and evaluated model assumptions using MR-Egger and MR-PRESSO. Proteins were considered to be causally associated with stroke subtypes if P < 2x10⁻⁴ (Bonferroni-adjusted for 241 proteins).

We identified 16 proteins significantly associated with at least one stroke subtype (s) (ten for any ischemic, five for large artery, eight for cardioembolic stroke). Nine of these proteins were causally associated with >1 stroke subtype, while seven proteins were associated with a specific stroke subtype. For example, we found that genetic predisposition to lower Tissue Factor Pathway Inhibitor (TFPI), an anti-coagulation protein, was associated with higher risk of any ischemic stroke (OR[95% CI]:1.09[1.07-1.12], P = 9.92x10-14). None of the proteins were causally associated with small vessel stroke.

By harnessing GWAS results from proteins and stroke subtypes, we identify proteins that likely to have causal roles in stroke, which may offer potential novel future therapeutic targets.

N/A

META-ANALYSIS OF INTRACRANIAL AND EXTRACRANIAL LARGE ARTERY ATHEROSCLEROTIC STROKE

B. Gaynor¹, J.F. Carrera², H. Xu¹, T. Jaworek¹, N. Cullell^3,4, E.G. Holliday⁵, I. Fernández-Cadenas^3,4, B.D. Mitchell¹, B.B. Worrall^2,6

¹University of Maryland School of Medicine, Division of Endocrinology- Diabetes- and Nutrition, Baltimore, USA

²University of Virginia, Department of Neurology, Charlottesville, USA

³Hospital Mútua de Terrassa, Stroke Pharmacogenomics and Genetics. Fundació Docència i Recerca MútuaTerrassa, Terrassa, Spain

⁴Sant Pau Hospital, Stroke Pharmacogenomics and Genetics. Sant Pau Institute of Research, Barcelona, Spain

⁵University of Newcastle, Centre for Clinical Epidemiology and Biostatistics- School of Medicine and Public Health, Newcastle, Australia

⁶University of Virginia, Department of Public Health Science, Charlottesville, USA

Most genome-wide association (GWA) stroke-associated loci are subtype-specific. Given recognized differences in epidemiology and underlying vessel structure, we hypothesize genetic heterogenity within the major stroke subtype, large artery atherosclerotic (LAA) stroke, and investigated those with intra- and extra-cranial disease in the NINDS Stroke Genetics Network (SiGN).

The SiGN network is an international collaboration consisting of ∼17,000 ischemic stroke cases with Causative Classification of Stroke (CCS) phenotypic data. We analyzed European-ancestry case with CCS ‘Evident’ or ‘Probable’ LAA (n = 3,428). We used four subphenotypes: LAAa-any extracranial stenosis (n = 1,270), LAAb-any intracranial stenosis (n = 618), LAAc-only extracranial stenosis (n = 1,096), and LAAd-only intracranial stenosis (n = 444)). We performed GWAS across 10 analysis strata using SNPTEST and meta-analyzed the summary results using GWAMA. SNPs showing association in the meta-analysis were then tested in an independent sample from a Spanish network.

Of the 6 established LAA stroke loci, 1 was found to be associated with intracranial only (ENDRA), 3 with extracranial only (HDAC9, TM4SF4-TM4SF1, 9p21), 1 with both (TSPAN2), and 1 with neither (LINC01492).

Table

We identified 2 potentially novel loci on our meta-analysis, including loci mapping to TNS1 (Tensin1) and RELN (Reelin).

Both the novel findings and the look up of previously established LAA stroke loci demonstrated patterns suggesting some genetic associations are subtype specific.

N/A

FATTY ACID BINDING PROTEIN EXPRESSION IN CLOTS RETRIEVED BY MECHANICAL THROMBECTOMY FROM PATIENTS WITH ACUTE ISCHAEMIC STROKE

O.M. Mereuta¹, S. Fitzgerald¹, A. Douglas¹, R. Rossi¹, A.M. Silva Santos^1,2, A. Pandit³, J. Thornton^4,5, A. O’Hare⁵, S. Power⁵, P. Brennan⁵, A. Rentzos⁶, T. Tatlisumak⁶, T. Gunnarsson⁶, M. Davidson⁷, A. Brederlau⁷, A. Allardt⁷, W. Brinjikji⁸, D.F. Kallmes⁸ and K.M. Doyle¹

¹National University of Ireland Galway, CÚRAM–Centre for Research in Medical Devices- Physiology Department, Galway, Ireland

²Centro Universitário Unievangélica, School of Medicine, Aná polis-Goiás, Brazil

³National University of Ireland Galway, CÚRAM–Centre for Research in Medical Devices, Galway, Ireland

⁴Royal College of Surgeons in Ireland, Beaumont Hospital, Dublin, Ireland

⁵Beaumont Hospital, Department of Radiology, Dublin, Ireland

⁶Sahlgrenska University Hospital, Department of Interventional and Diagnostic Neuroradiology, Gothenburg, Sweden

⁷Sahlgrenska University Hospital, Department of Neurology, Gothenburg, Sweden

⁸Mayo Clinic, Department of Radiology, Rochester- Minnesota, USA

Lipid accumulation and inflammation are considered hallmarks of the unstable atherosclerotic plaque. In particular, adipocyte fatty acid binding protein (FABP4) expression within the plaque is associated with its progression and vulnerability. The aim of this study was to investigate the histopathology of thrombi collected from acute ischaemic stroke patients who underwent mechanical thrombectomy focusing on the presence of FABP4 and other atherosclerotic plaque components, including collagen and dystrophic calcification.

250 mechanically extracted thrombi were collected from three partner hospitals: Beaumont Hospital (Dublin, Ireland), Sahlgrenska University Hospital (Gothenburg, Sweden) and Mayo Clinic (Rochester, Minnesota, USA). Clots were immediately formalin-fixed and embedded in paraffin. 3-µm thickness serial sections were cut and stained with Martius Scarlett Blue to identify erythrocytes, fibrin, white blood cells and platelets/other. Masson’s and von Kossa stains were performed to identify collagen and calcification, respectively. The expression of FABP4 was assessed by immunofluorescence.

FABP4 expression by adipocytes was confirmed in eight of 250 clots (3.2%). Adipocytes represented the main component in one ‘white’ clot. Collagen was associated in two cases whereas von Kossa staining showed calcification in one case.

The expression of potentially atherogenic components in these clots suggests that they may have originated from large vessel occlusions. Further studies are required to confirm the atherosclerotic aetiology. Our findings have implications for neurointervention and therapeutic strategies to reduce atherosclerotic plaque progression and stroke recurrence.

Acknowledgements: Science Foundation Ireland (Grant Number 13/RC/2073) and Cerenovus.

N/A

EARLY OXIDATIVE STRESS BIOMARKERS OF ISCHEMIC PENUMBRA AND INFARCT GROWTH IN ACUTE STROKE

S. Lorenzano¹, N.S. Rost¹, M. Khan², H. Li³, F.O. Lima¹, A. Chutinet¹, R.E. Green¹, T.K. Thankachan¹, K. Arai⁴, A.T. Som⁴, L.D.D. Pham⁴, O. Wu⁵, G.J. Harris³, E.H. Lo⁴, J.B. Blumberg⁶, P.E. Milbury⁷, S.K. Feske⁸ and K.L. Furie²

¹Massachusetts General Hospital- Harvard Medical School, Department of Neurology, Boston, USA

²Rhode Island Hospital- Alpert Medical School of Brown University, Department of Neurology, Providence, USA

³Massachusetts General Hospital- Harvard Medical School, Department of Radiology, Boston, USA

⁴Massachusetts General Hospital- Harvard Medical School, Departments of Neurology and Radiology- Neuroprotection Research Laboratory- Neuroscience Center, Boston, USA

⁵Massachusetts General Hospital- Harvard Medical School, Athinoula A. Martinos Center for Biomedical Imaging, Boston, USA

⁶Tufts University, Antioxidant Research Laboratory- Jean Mayer USDA Human Nutrition Research Center on Aging, Boston, USA

⁷Tufts University, Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy, Boston, USA

⁸Brigham and Women’s Hospital- Harvard Medical School, Department of Neurology, Boston, USA

Oxidative stress is an early response to cerebral ischemia. We sought to evaluate whether hyperacute plasma concentrations of oxidative stress biomarkers predict ischemic penumbra and infarct growth (IG) in acute ischemic stroke (AIS) patients.

We measured plasma levels of oxidative stress biomarkers such as F2-isoprostanes (F2-isoP), total and perchloric acid Oxygen Absorbance Capacity (ORAC_TOT and ORAC_PCA), urinary levels of 8-oxo-7,8-dihydro-2’-deoxyguoanosine, and inflammatory and tissue-damage biomarkers (high-sensitivity CRP, matrix metalloproteinases 2 and 9), in a prospective study of AIS patients presenting within 9 hours of symptom onset. Baseline DWI- and PWI-weighted imaging volume and final infarct volume were analyzed using semi-automated method. Mismatch was defined as baseline mean transit time volume minus DWI volume.A percent mismatch cut-off of >20% was considered clinically significant.A stricter definition of mismatch excluding too small DWI and PWI lesions was also used.IG volume was defined as the difference between final infarct volume and baseline DWI volume.

Of 220 subjects (41% females; mean[SD] age 69.4[14.7];median[IQR] NIHSS 8[4-14];55.5% treated with IV t-PA), 153 had mismatch >20% and 170 had IG. In multivariate analyses, baseline F2-isoP(OR 2.44, 95% CI 1.19-4.98;P = 0.014) and ORACPCA(OR 4.18, 95% CI 1.41-12.41;P = 0.010) emerged as independent molecular predictors of >20% mismatch and the stricter mismatch definition, respectively.ORACTOT, as a marker of total plasma antioxidant capacity, significantly predicted mismatch salvage volume (B 44.12, 95% CI 11.07-77.18;P = 0.010). Baseline F2-isoP was independent predictor of IG occurrence (OR 2.57, 95% CI 1.37–4.83;P = 0.007) and IG volume (B, 0.38, 95% CI 0.04–0.72;P = 0.03).

If validated, these findings may add to our understanding of the role of oxidative stress in cerebral tissue fate during acute ischemia.

N/A

SERUM NEUROFILAMENT LIGHT CHAIN PREDICTS BRAIN VOLUME LOSS IN PATIENTS WITH ATRIAL FIBRILLATION

A. Polymeris¹, T. Sinnecker^1,2, P. Benkert³, N. Rodondi^4,5, A. Mueller⁶, J. Beer⁷, G. Moschovitis⁸, D. Hayoz⁹, J. Novak¹⁰, J. Schlaepfer¹¹, M. di Valentino¹², U. Fischer¹³, A. Monsch¹⁴, J. Wuerfel², M. Schwenkglenks¹⁵, M. Kuehne¹⁶, S. Osswald¹⁶, D. Conen^16,17, J. Kuhle¹, L. Bonati¹; for the SWISS-AF Investigators

¹University Hospital Basel and University of Basel, Department of Neurology-Stroke Center and Department of Clinical Research, Basel, Switzerland

²Medical Image Analysis Center Basel AG and University of Basel, Department of Biomedical Engineering, Basel, Switzerland

³University Hospital Basel and University of Basel, Clinical Trial Unit-Department of Clinical Research-, Basel, Switzerland

⁴University of Bern, Institute of Primary Health Care BIHAM, Bern, Switzerland

⁵Inselspital-Bern University Hospital and University of Bern, Department of General Internal Medicine, Bern, Switzerland

⁶Triemli Hospital Zurich, Department of Cardiology, Zurich, Switzerland

⁷Cantonal Hospital Baden and University Hospital Zürich, Department of Medicine and and Molecular Cardiology, Zurich, Switzerland

⁸Civic Hospital Lugano, Department of Cardiology-, Lugano, Switzerland

⁹HFR-Hô pital Cantonal Fribourg, Department of Internal Medicine, Fribourg, Switzerland

¹⁰Public Hospital Solothurn, Department of Cardiology, Solothurn, Switzerland

¹¹University Hospital Lausanne, Department of Cardiology, Lausanne, Switzerland

¹²Ospedale San Giovanni, Department of Cardiology, Bellinzona, Switzerland

¹³Inselspital-Bern University Hospital and University of Bern, Department of Neurology, Bern, Switzerland

¹⁴University Center for Medicine of Aging-Felix Platter Hospital and University of Basel, Memory Clinic, Basel, Switzerland

¹⁵University of Zurich, Epidemiology-Biostatistics and Prevention Institute, Zurich, Switzerland

¹⁶University Hospital Basel and Cardiovascular Research Institute Basel, Cardiology Division-Department of Medicine, Basel, Switzerland

¹⁷McMaster University, Population Health Research Institute, Hamilton, Canada

There is emerging evidence that atrial fibrillation (AF) is associated with cognitive dysfunction, increased risk for dementia and reduced brain volume independent of stroke, but the underlying mechanisms of these associations remain unclear. Here, we investigated the association of serum neurofilament light chain (sNfL), a neuroaxonal injury biomarker, with brain atrophy in AF patients.

Explorative analysis from the prospective observational Swiss-AF cohort study, which recruited AF patients aged ≥ 45 years (NCT02105844). sNfL concentrations were measured in duplicate at baseline using a single molecule array (SIMOA) assay. Brain MRI was obtained at baseline and after two years using a standardized protocol that included a 3D T1-weighted MPRAGE sequence, on which Structural Image Evaluation using Normalization of Atrophy (SIENA) with optimized parameters for brain extraction was applied to calculate the two-year percentage whole brain volume change (PBVC).

Pilot data were available in 245 Swiss-AF patients (median age 73, 74% male). In a simple linear regression model, baseline sNfL was significantly associated with two-year PBVC, with a 0.09% whole brain volume decrease per 10 pg/ml sNfL increase (95%-CI [0.05 – 0.13], p < .001). This association remained significant (0.07% volume decrease per 10 pg/ml sNfL increase, 95%-CI [0.02 – 0.12], p = 0.005) after adjustment for normalized baseline brain volume, age, history of stroke and other vascular risk factors.

In AF patients, baseline sNfL was predictive of brain atrophy at two years independent of stroke history and other clinical factors. This association might reflect a chronic neurodegenerative process in AF.

N/A

THE ROLE OF TRIMETHYLAMINE-N-OXIDE IN PATIENTS WITH ACUTE ISCHEMIC STROKE

S. Mundiyanapurath¹, C. Schneider², J. Hauke³ and J. Okun³

¹University Hospital Heidelberg, Department of Neurology, Heidelberg, Germany

²University Hospital Heidelberg, Neurology, Heidelberg, Germany

³University Hospital Heidelberg, Dietmar-Hopp Metabolic Center- Center for Child and Adolescend Medicine, Heidelberg, Germany

Trimethylamine-N-oxide (TMAO) is a biomarker of the gut microbiome and is associated with cardiovascular diseases. However, there have been conflicting reports regarding its association with ischemic stroke questioning its prognostic value. We hypothesized that TMAO values might change after stroke and aimed at analyzing the early time course of TMAO values. We also wanted to explore the differences in TMAO levels between different stroke etiologies.

In this prospective case-control study, we include patients with ischemic stroke < 24 hours after last-seen-well and control patients treated for other neurologic disorders with only few cardiovascular risk factors at a tertiary care university hospital. The TMAO-values on admission, after 48 hours and after 3 months were analyzed. Patients and controls were excluded if they were treated with antibiotics within the last 30 days and if they were treated with probiotics or chemotherapy.

The results will be ready for presentation for ESOC 2019. A preplanned interim analysis with 100 patients showed that stroke patient had significantly higher median TMAO values compared to control patients: 5.4 (3.0-8.0) vs. 2.8 (1.6-5.0) μmol/l, p = 0.002. There was a significant decrease in TMAO values in stroke patients from admission to 48 hours after the stroke (median decrease 0.7 (0.6-4.2) p = 0.004), which was not found for the control group (median decrease 0.1 (-0.9-1.3), p = 0.758). We will present data of 300 patients at the conference.

This study will help to understand the contradicting publications on TMAO and differentiate whether high TMAO-levels occur prior or after stroke.

N/A

PATIENTS WITH HIGH POST-ENDOVASCULAR TREATMENT PLASMA LEVELS OF NEUTROPHIL ACTIVATION MARKERS ARE AT HIGHER RISK OF POOR FUNCTIONAL OUTCOME, HT, AND MORTALITY

L. Di Meglio¹, J.P. Desilles^1,2, V. Ollivier¹, M. Solo Nomenjanahary¹, R. Blanc², M. Piotin², M.C. Bouton¹, M. Jandrot-Perrus¹, B. Ho-Tin-Noe¹ and M. Mazighi^1,2

¹Inserm u1148 LVTS, Inserm, Paris, France

²Rothschild Foundation Hospital, Department of Interventional Neuroradiology., Paris, France

We investigated possible relationships between plasma levels of neutrophil activation markers and clinical characteristics and outcome in thrombectomy (EVT)-treated stroke patients.

Neutrophil activation markers were measured in plasma samples from 72 acute ischemic stroke patients. For each patient, samples were drawn before EVT (baseline), 1 hour after the end of EVT, and 24 hours after EVT.

Baseline levels of neutrophil activation markers were not associated with in hospital-mortality, outcome, or hemorrhagic transformation (HT). At 1 hour after EVT, there was a statistically significant increase in plasma myeloperoxidase (MPO) and citrullinated histones (H3c) as compared to baseline. Although intravenous t-PA treatment prior to EVT did not affect baseline MPO plasma levels, it lowered the increase in plasma MPO observed at 1 hour post-EVT. In hospital-mortality was associated with higher plasma levels of H3c at 1 hour post-EVT, and with higher plasma levels of MPO, MMP9, and neutrophil elastase at 24 hours post-EVT. Patients with a favorable outcome (modified Rankin Scale (mRS)≤2) had lower MPO, MMP9, and H3c plasma levels at 1 hour post-EVT, and lower MPO, neutrophil elastase, and H3c plasma levels at 24 hour post-EVT, as compared to patients with a mRS>2. Patients evolving towards HT had higher MPO and H3c plasma levels at 1 hour post-EVT, and higher MPO and neutrophil elastase plasma levels at 24 hour post-EVT, as compared to patients without HT.

In conclusion, patients with high post-EVT plasma levels of neutrophil activation markers are at higher risk of poor functional outcome, HT, and mortality.

NCT02907736

A GENOME-WIDE INTERACTION ANALYSIS WITH HYPERTENSION IDENTIFIES NOVEL GENETIC RISK LOCI FOR CEREBRAL SMALL VESSEL DISEASE ON BRAIN MRI

M. Sargurupremraj¹, S. Seshadri^2,3, P. Mathews M⁴, M. Fornage⁵ and S. Debette^1,3,6

¹University of Bordeaux, inserm u1219 – Bordeaux Population Health, Bordeaux, France

²UT Health San Antonio, Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Diseases, San Antonio, USA

³Boston University School of Medicine, Department of Neurology, Boston, USA

⁴Imperial College London, Department of Medicine, London, United Kingdom

⁵University of Texas Health Science Center, Brown Foundation Institute of Molecular Medicine, Houston, USA

⁶Bordeaux University Hospital, Department of Neurology, Bordeaux, France

MRI-defined white matter hyperintensity (WMH) burden is a highly heritable endophenotype for cerebral small vessel disease (cSVD) and a powerful predictor of stroke and dementia risk. Hypertension (HTN) being the strongest known risk factor for WMH, we sought to enhance the discovery of genetic risk loci for WMH burden genome-wide (GW) by accounting for interaction of single nucleotide polymorphisms (SNPs) with HTN.

We jointly meta-analyzed the SNP-main effects and the interaction (SNP-by-HTN) effects (2 degree-of-freedom) for WMH burden in 23 studies from the CHARGE consortium and UK-Biobank participants, gathering 36,114 stroke-free individuals. Analyses of shared genetic variation of WMH burden with other vascular or neurological traits were carried out at the GW, regional and risk-score level. Gene regulatory information was integrated to infer relevant tissue enrichment patterns.

We identified 17 independent WMH risk SNPs satisfying GW significance (P < 5.00 × 10-08), of which 11 are novel, one locus being observed in African-Americans only. In HTN-stratified analyses some of these loci were predominantly associated with WMH in hypertensives or non-hypertensives. WMH burden exhibited significant shared genetic variation with ischemic-stroke, body-mass index, type-2 diabetes, LDL-cholesterol, and migraine-with-aura. Tissue enrichment analyses suggested strong co-localization of WMH-loci and chromatin-marks for enhancer and promoter elements in fetal-brain and neural stem cells.

We tripled the number of genetic risk loci for WMH burden, leading to novel insight into cSVD biology. Shared genetic determinants with several vascular and neurological traits were revealed. Tissue enrichment patterns suggest involvement of mechanisms that also play roles in neurodevelopment.

N/A

MULTIANCESTRY GWAS META-ANALYSIS OF PERIVASCULAR SPACE BURDEN, AN EMERGING MRI-MARKER OF SMALL VESSEL DISEASE, BY THE CHARGE CONSORTIUM

M.G. Duperron¹, M. J. Knol², J. Rafael Romero³, S. Frenzel⁴, N. Dueker⁵, E. Hofer⁶, M. Luciano⁷, C. Carrera V⁸, S. Hilal⁹, C. Li-Hslan Chen⁹, P. Delgado⁸, R. Schmidt¹⁰, T. Rundek¹¹, C. Tzourio¹, A. Teumer¹², M.A. Ikram², J. Wardlaw¹³, S. Seshadri³, H. H. H. Adams² and S. Debette¹

¹University of Bordeaux- Inserm, Bordeaux Population Health Research Center- UMR 1219, Bordeaux, France

²Erasmus MC University Medical Center, Department of Epidemiology, Rotterdam, The Netherlands

³Boston University School of Medicine, Department of Neurology, Boston, USA

⁴University Medicine Greifswald, Department of Psychiatry and Psychotherapy, Greifswald, Germany

⁵University of Miami, Miller School of Medicine, Miami, USA

⁶Clinical Division of Neurogeriatrics, Department of Neurology, Graz, Austria

⁷Centre for Cognitive Ageing and Cognitive Epidemiology, Department of Psychology, Edinburgh, United Kingdom

⁸Vall Hebron Research Institute, Neurovascular Research Lab, Barcelona, Spain

⁹National University of Singapore, Department of Pharmacology, Singapor, Singapore

¹⁰Institute for Medical Informatics- Statistics and Documentation, Medical University of Graz, Graz, Austria

¹¹University of Miami, Department of Neurology, Miami, USA

¹²University of Greifswald, Institute of Community Medicine, Greifswald, Germany

¹³University of Edinburgh, Brain Research Imaging Centre- Centre for Clinical Brain Sciences- Edinburgh Dementia Research Centre- and Centre for Cognitive Ageing and Cognitive Epidemiology, Edinburgh, United Kingdom

Dilated perivascular space (dPVS) burden is an emerging MRI-marker of cerebral small vessel disease of unknown mechanism, recently shown to be highly heritable. We conducted the first meta-analysis of multiancestry genome-wide association studies of dPVS burden to decipher its genetic underpinnings.

We included 8,959 stroke-free participants from 16 population-based cohorts and 4 ethnicities (87.4% European, 8.2% Hispanic, 2.4% Asian, 1.9% African-American). dPVS were rated in white matter (WM), basal ganglia (BG), hippocampus, and mesencephalon using semi-quantitative scales. We tested association of genetic variants with the top quartile of dPVS burden distribution in each cohort followed by multi-ancestry sample-size weighted meta-analysis. Associations reaching p < 10⁻⁷ were followed up in an independent sample (N = 2,425 Europeans). We searched for shared genetic variation with related vascular and neurological phenotypes using linkage disequilibrium-score regression.

We identified two loci associated with WM dPVS burden at p < 10-6 on chr9q31.3 (p = 9.7x10-8) and chr3p25.1 (p = 8.3x10-7). Both were replicated in the follow-up sample and reached genome-wide significance in a combined meta-analysis (p = 1.7x10-9 and p = 1.8x10-8). We found significant positive genetic correlation of BG (and nominally hippocampus) dPVS burden with any stroke, ischemic stroke, and nominally significant positive genetic correlation of BG dPVS burden with deep intra-cerebral hemorrhage (also for hippocampus dPVS), small vessel stroke, WM hyperintensities (WMH), and systolic blood pressure.

We identified 2 loci associated with dPVS burden in WM, implicated in brain development, neuronal plasticity, and tumor related pathways, and shared genetic variation of BG and hippocampus dPVS burden with stroke, WMH and blood pressure.

N/A

Scientific Communications 24-Rehabilitation Effectiveness

EVALUATING THE EFFECTIVENESS AND IMPACT OF STROKE EARLY SUPPORTED DISCHARGE IN REAL WORLD CONDITIONS

A. Byrne¹, S. Lewis², N. Chouliara¹, L. Paley³, A. Hoffman³, A. Rudd³, C. Geue⁴, J. Waring⁵, T. Robinson⁶, M. Walker¹, P. Langhorne⁷, R. Fisher¹; The SSNAP Collaboration

¹University of Nottingham, Division of Rehabilitation and Ageing, Nottingham, United Kingdom

²University of Nottingham, Division of Epidemiology and Public Health, Nottingham, United Kingdom

³King’s College London, Sentinel Stroke National Audit Programme, London, United Kingdom

⁴University of Glasgow, Health Economics and Health Technology Assessment, Glasgow, United Kingdom

⁵Nottingham University Business School, Organisational Behaviour and Human Resource Management, Nottingham, United Kingdom

⁶University of Leicester, Department of Cardiovascular Sciences, Leicester, United Kingdom

⁷University of Glasgow, Institute of Cardiovascular and Medical Sciences, Glasgow, United Kingdom

Clinical trials have shown stroke Early Supported Discharge (ESD) reduces risk of dependency and length of hospital stay of stroke survivors. This observational study investigated the effectiveness of ESD services in real world conditions and evaluated the impact of adopting an evidence based model.

Multilevel modelling was conducted using 2013–2016 historical prospective Sentinel Stroke National Audit Programme (SSNAP) clinical data and 2015 SSNAP post-acute organisational audit data. ESD models (n = 31) were scored using a 17 item ESD consensus score to measure adherence to the evidence base (Fisher et al. 2011). ESD effectiveness was measured by time to first contact within 24 hours (responsiveness; n = 6222), treatment days/total ESD days (intensity; n = 5891) and Modified Rankin Scale (n = 6222). Propensity score matching was then used to compare length of hospital stay between ESD and non-ESD patients.

A one unit increase in ESD consensus score was significantly associated with increased odds of responsiveness by 29% (95% CI 1% – 49%) and increased treatment intensity by 2% (95% CI 0.3% – 4%) but had no effect on stroke survivor outcome measured by the Modified Rankin Scale. ESD patients stayed in hospital for 25.7 hours longer than matched non-ESD patients (95% CI 24.7–26.8).

We found that adopting defined core components of ESD is associated with providing a more responsive and intensive ESD service indicating that adherence to evidence based criteria is required for services to be effective. However, we did not show that the presence of ESD led to a reduction in length of hospital stay.

N/A

THE THERAPY–RESPONSE RELATIONSHIP IS IT SIMPLY MORE THERAPY EQUALS BETTER HEALTH OUTCOMES? THE SENTINEL STROKE NATIONAL AUDIT PROGRAMME: INVESTIGATING AND EVALUATING STROKE THERAPY (SSNAPIEST)

M. Gittins¹, S. Tyson², A. Vail¹, A. Bowen³, D. Lugo-Palacios⁴, B. Bray⁵, B. Gannon⁶, L. Paley⁷; The SSNAP collaborators and particpants

¹University of Manchester, Centre for Biostatistics- Manchester Academic Health Science Centre MAHSC, Manchester, United Kingdom

²University of Manchester, Division of Nursing- Midwifery and Social Work- MAHSC, Manchester, United Kingdom

³University of Manchester, Division of Neuroscience & Experimental Psychology- MAHSC, Manchester, United Kingdom

⁴Imperial College London, Centre for Health Policy Institute of Global Health Innovation, London, United Kingdom

⁵University College London, Farr Institute of Health Informatics Research, London, United Kingdom

⁶The University of Queensland, Centre for Business and Economics of Health, Brisbane, United Kingdom

⁷Kings College London, School of Population Health & Environmental Sciences, London, United Kingdom

Stroke therapy is an important factor driving improved patient recovery, with more therapy generally considered to equal improved health outcomes. Here we investigate whether the therapy ‘dose-response’ is a simple linear relationship.

All in-patient strokes still present at three days (n = 94,905) as reported to SSNAP (July 2013–2015) were included. Robust multilevel mixed effects regression models adjusted for all available confounding factors investigated the association between amount of therapy received, defined as “average minutes/day during in-patient stay”, and the health outcome modified Rankin Scale. Physiotherapy (PT), Occupational Therapy (OT) and Speech and Language Therapy (SLT) were all investigated separately using flexible ‘natural cubic splines’ (NCS) to investigate changes in association across increasing levels of average therapy received.

The NCS approach indicated an optimal threshold point between 5–10min/day of stay. Increases in all therapies were associated with improvements in health outcome mRS prior to the threshold. Further increases in average minutes of OT and SLT per day of stay post threshold were associated with continued but gradual improvements at discharge, whereas increasing PT was associated with a gradual decline such that an approximate 40 min/day would see a return to baseline health.

Small doses of therapy/day appear to be beneficial to patient health with increases beyond this associated with diminishing returns or even detriment to disability at discharge. Our data however, contains strong limitations that urgently require further investigation in a prospective research study before confirming or refuting these findings.

N/A

ARE IN-PATIENTS WITH STROKE IN THE UK GETTING ENOUGH THERAPY? DATA FROM THE NATIONAL STROKE REGISTRY FOR ENGLAND, WALES, AND NORTHERN IRELAND

V. Mccurran¹, M. James², L. Clark³, C. Wolfe¹, A. Hoffman¹, A. Rudd¹; The SSNAP Collaboration

¹King’s College London, School of Population Health and Environmental Sciences, London, United Kingdom

²Royal Devon and Exeter NHS Foundation Trust, Royal Devon and Exeter NHS Foundation Trust, Exeter, United Kingdom

³Dorset County Hospital NHS Foundation Trust, Dorset County Hospital NHS Foundation Trust, Dorchester, United Kingdom

Occupational therapy (OT), physiotherapy (PT), and speech (SLT) are integral to stroke rehabilitation. National Clinical Guidelines for Stroke 2016, recommend a minimum of 45 minutes per day of each appropriate therapy, for a minimum of 5 days per week where it can be tolerated.

Data from 82,190 patients discharged between April 2017-March 2018 were analysed from the Sentinel Stroke National Audit Programme (SSNAP). Number of days therapy was applicable, median average number of minutes per day and percentage of days therapy received were calculated using OT as the exemplar.

Median length of stay was 7 days (IQR 3–23). The proportion of days therapy was received decreases after 4 days from 100% to 60% after 14 days. Over 90 days median minutes of therapy given on the day treated remained unchanged at about 40 minutes, as did the proportion of days on which patients were considered well enough to be treated.

Our findings indicate the intensity of therapy declines sharply after the first 5 days. This decline particularly affects the more disabled patients with longer lengths of stay, when they may be the very patients for whom maximising therapy input would make the greatest difference to outcome. Data not presented show a similar pattern for PT and SLT.

N/A

INCREASING SOCIAL, COGNITIVE AND PHYSICAL ACTIVITY ON ACUTE STROKE UNITS USING AN EXPERIENCE-BASED CO-DESIGN APPROACH: THE COLLABORATIVE REHABILITATION ENVIRONMENTS IN ACUTE STROKE (CREATE) STUDY

D. Clarke¹, K. Gombert², G. Robert³, R. Harris³, C. McKevitt⁴, S. Honey⁵, A. Macdonald⁶, G. Cloud⁷ and F. Jones²

¹University of Leeds, Academic Unit of Elderly Care and Rehabilitation, Leeds, United Kingdom

²St George’s and Kingston University of London, Allied Health- Midwifery and Social Care, London, United Kingdom

³Kings College London, Florence Nightingale Faculty of Nursing- Midwifery & Palliative Care, London, United Kingdom

⁴Kings College London, Primary Care & Public Health Sciences, London, United Kingdom

⁵University of Leeds, Institute of Health Sciences, Leeds, United Kingdom

⁶Glasgow School of Art, School of Design, Glasgow, United Kingdom

⁷Alfred Health, Stroke Services, Melbourne, Australia

Observational studies in stroke units since the 1980s consistently demonstrate patients have few opportunities for social, cognitive or physical activity outside formal therapy sessions. Boredom and apathy are commonplace impacting negatively on rehabilitation participation. The CREATE study brought together stroke survivors, caregivers, staff and other stakeholders to co-design interventions to increase inpatient activity.

Mixed-method, case comparison evaluation incorporating full or accelerated cycle Experience-Based Co-Design (EBCD) methods in four stroke units in England. Ethnographic observations (400 hours); 72 staff, 52 patient, 25 carer interviews and 173 patient reported outcome and experience measures. Behavioural mapping was completed on three occasions pre and post EBCD cycles in each site. Normalisation Process Theory was used in analysis of implementation and assimilation of co-produced interventions.

Baseline data in all sites identified high levels of inactivity. Co-produced interventions included unit level environmental changes which facilitated therapeutic group activities and greater social and cognitive engagement between patients, families and staff. Communal spaces were repurposed; engagement with community and voluntary sectors established singing, art, lunch and exercise groups. Personalising bed spaces enabled staff to understand the person who is the patient and individually tailor activities. Patients and families engaged positively with EBCD and reported substantive changes had occurred. Staff reported positive changes in their working practices and unit environments. Ethnographic observations and behavioural mapping data indicated more patient-staff interactions and increased individual and group activity.

Experienced-Based Co-Design methods facilitated development and implementation of interventions which contributed to increased social, cognitive and physical activity on acute stroke units.

N/A

TAKING CHARGE AFTER STROKE: REASONS FOR EFFECTIVENESS FOR A PERSON-CENTRED APPROACH TO LIFE BEYOND STROKE

H. Mcnaughton¹, V. Fu²; On behalf of The Taking Charge After Stroke (TaCAS) Investigators

¹Royal Derby Hospital, Stroke, DERBY, United Kingdom

²Medical Research Institute of New Zealand, Stroke/Rehabilitation research, Wellington, New Zealand

The Take Charge (TC) intervention – a ’talking therapy’ approach facilitating self-rehabilitation for people discharged to community living following stroke – improves independence and quality of life. Number needed to treat for 1 more person to be independent, 12 months after stroke is 8 – equivalent to intravenous thrombolysis within 3 hours of acute stroke. But how is it effective? Contrasting the TC intervention with current rehabilitation practice and using secondary outcomes from the Taking Charge after Stroke (TaCAS) trial we explore reasons for effectiveness.

TaCAS methods and main results are presented separately at ESOC. 400 participants were randomised within 16 weeks of stroke to one or two TC sessions or control. The TC session is designed to promote autonomy, mastery, connectedness and a sense of purpose and is a wholly personal approach to the person with stroke. Take Charge rejects all the elements of SMART (specific, measurable, achievable, realistic/relevant, timed) goal setting except for relevance to the person with stroke.

In control but not intervention groups, better outcomes were associated with higher scores on the Autonomy/mastery/purpose/connectedness (AMP-C) measure at baseline, suggesting that the TC intervention modified baseline perceptions so that more people improved. Significant improvements in Frenchay Activity Index scores at 12 months but not 6 months suggest people were ‘taking charge’ of their own lives over time. Other results and supporting literature will be discussed.

The Take Charge intervention encourages intrinsic motivation over time with a powerful effect on 12-month outcome.

Australian New Zealand Clinical Trials Registry ACTRN12615001163594

TAKING CHARGE AFTER STROKE: COST-EFFECTIVENESS ANALYSIS OF A RANDOMISED CONTROLLED TRIAL OF A PERSON-CENTRED INTERVENTION TO PROMOTE SELF-REHABILITATION

B. Te Ao¹, H. Mcnaughton², V. Fu³; On behalf of the Taking Charge after Stroke (TaCAS) Investigators

¹University of Auckland, School of Population Health, Auckland, New Zealand

²Royal Derby Hospital, Stroke, Derby, United Kingdom

³Medical Research Institute of New Zealand, Stroke/Rehabilitation research, Wellington, New Zealand

The Take Charge intervention – a brief ’talking therapy’ approach to facilitate self-rehabilitation for people discharged to community living following stroke – improves independence and quality of life. We performed a cost-effectiveness analysis as part of the Taking Charge after Stroke (TaCAS) trial.

TaCAS methods and main results are being presented separately at ESOC. 400 participants were randomly allocated to either one or two Take Charge (TC) sessions or a control intervention. One TC session costs €60. Health utilities were derived from EuroQol 5D scores, 12 months after stroke. Direct and indirect costs of care for all participants between baseline (mean 45 days after stroke) and final followup 12 months after stroke were calculated from interview, hospital records and Department of Health records.

In the control group the mean 1 year cost of care was €8,123 (95% CI 4,567 to 9,178) and mean health utility score was 0.71 (95% CI 0.67 to 0.75) compared to the TC intervention combined group where mean 1 year cost was €5,850 (95% CI 3,809 to 6,038) and mean health utility score was 0.75 (95% CI 0.73 to 0.77). Mean cost saving per participant was €2,273. Number needed to treat to gain 1 QALY is 25. Cost per quality adjusted life year (QALY) gained is (-)€56,825. Take Charge remains dominant over a wide range of cost assumptions. Most of the cost saving came from reduced rates of institutional care and hospital readmission.

The Take Charge intervention is likely cost-saving.

Australian New Zealand Clinical Trials Registry ACTRN12615001163594

EVALUATION OF AN EXTENDED STROKE REHABILITATION SERVICE (EXTRAS): COST-EFFECTIVENESS RESULTS

H. Rodgers¹, N. Bhattarai², P. McMeekin³, L. Shaw¹, R. Cant⁴, A. Drummond⁵, G. Ford⁶, A. Forster⁷, R. Francis¹, K. Hills¹, D. Howel², A.M. Laverty⁸, C. McKevitt⁹, C. Price¹, E. Stamp², E. Stevens⁹ and L. Vale²

¹Newcastle University, Stroke Research Group, Newcastle upon Tyne, United Kingdom

²Newcastle University, Institute of Health and Society, Newcastle upon Tyne, United Kingdom

³Northumbria University, Faculty of Health and Life Sciences, Newcastle upon Tyne, United Kingdom

⁴Lay Investigator, Stroke Research Group, Newcastle upon Tyne, United Kingdom

⁵Nottingham University, School of Health Sciences, Nottingham, United Kingdom

⁶Oxford University, Medical Sciences Division, Oxford, United Kingdom

⁷Leeds University, Academic Unit of Elderly Care and Rehabilitation, Leeds, United Kingdom

⁸Northumbria Healthcare NHS Foundation Trust, Stroke Northumbria, Ashington, United Kingdom

⁹King’s College London, School of Population Health and Environmental Sciences, London, United Kingdom

Stroke survivors frequently report unmet needs in the longer term but there is limited evidence to guide on-going rehabilitation.

This RCT involved 19 UK centres which provided Early Supported Discharge (ESD). Stroke patients were randomised to receive EXTRAS or usual care. EXTRAS involved five rehabilitation reviews provided by an ESD team member between one and 18 months post-ESD. Reviews were predominantly undertaken by telephone and consisted of a semi-structured assessment of rehabilitation needs with goal setting and action planning. The primary outcome was performance in extended activities of daily living (Nottingham Extended Activities of Daily Living (NEADL) scale) at 24 months. Cost-effectiveness was estimated using resource utilisation costs (adaptation Client Service Receipt Inventory) and quality adjusted life years (QALYs) derived from the EQ-5D-5L.

573 participants were randomised (EXTRAS n = 285, usual care n = 288). Mean 24 month NEADL scores were EXTRAS 40.0 and usual care 37.2: an adjusted mean difference of 1.8 (95% CI -0.7 to 4.2). The intervention group mean resource utilisation costs were lower: -£311 (95% CI -£3,292 to £2,787). Savings were predominantly in social care costs. EXTRAS provided more QALYs (0.07, 95% CI 0.01 to 0.12). There was a 90% chance of EXTRAS being cost-effective at conventional thresholds of willingness to pay (£20,000 per QALY).

EXTRAS did not improve stroke survivors’ performance in extended activities of daily living. However, due to impact on costs and QALYs, there is a high chance that EXTRAS could be considered cost-effective.

ISRCTN45203373

ECONOMIC EVALUATION OF A PHASE III INTERNATIONAL RANDOMISED CONTROLLED TRIAL OF VERY EARLY MOBILISATION AFTER STROKE (AVERT)

L. Gao¹, L. Sheppard¹, O. Wu², L. Churilov³, M. Mohebbi⁴, J. Collier³, J. Bernhardt³, F. Ellery³, H. Dewey⁵, M. Moodie¹; The AVERT Trial Collaboration Group

¹Deakin University, Deakin Health Economics, Melbourne, Australia

²University of Glasgow, Institute of Health and Wellbeing, Glasgow, United Kingdom

³The Florey Institute of Neuroscience and Mental Health, NHMRC Research Excellence in Stroke Rehabilitation and Brain Recovery, Melbourne, Australia

⁴Deakin University, Biostatistic Department, Melbourne, Australia

⁵Monash University, Eastern Health Clinical School, Melbourne, Australia

Assessing the cost-effectiveness of interventions within a Phase III randomised controlled trial (RCT). In this pre-specified economic evaluation, we examined the long-term cost of care post stroke, comparing experimental intervention (early, intensive mobilisation; VEM) to usual care (UC).

Cost-utility analyses were based on 12 -month outcomes. The analysis was conducted using health sector and societal perspectives. Unit costs were sourced from participating countries. Details on resource use were sourced from Cost Case Report Form which had been adapted to meet individual country needs. Dichotomised Modified Rankin Scale (mRS) scores (0-2 vs 3–6) and Quality Adjusted Life Years (QALYs) were used to compare the treatment effect of VEM and UC. The base case analysis was performed on an Intention-To-Treat (ITT) basis and 95% confidence intervals (CI) for cost and QALYs were estimated by bootstrapping.

VEM and UC groups were comparable in the quantity of resource use and cost of each component. There were no differences in the probability of achieving a favourable mRS outcome (0.030, 95% CI: -0.022 to 0.082), QALYs (0.013, 95% CI: -0.041 to 0.016) and cost (AUD1082, 95% CI: -$2520 to $4685 from a health sector perspective; or AUD$102, 95% CI: -$6907 to $7111, from a societal perspective).

Despite higher intensity of training, VEM and UC were associated with comparable costs, mRS outcome and QALY gains at 12 months. Compared with to UC, VEM is unlikely to be cost-effective and was shown in the primary efficacy analysis to lead to lower odds of a favourable outcomes. It is therefore not recommended.

N/A

Young Stroke Physicians and Researchers: Research Design Workshop for Studies in Development

THE PASHT TRIAL: RANDOMIZED CROSSOVER SHAM-CONTROLLED TRIAL EVALUATING SAFETY AND FEASIBILITY OF TRANSCRANIAL DIRECT CURRENT STIMULATION (TDCS) FOR PAROXYSMAL SYMPATHETIC HYPERACTIVITY POST STROKE AND TBI

F. Saleh Velez¹, C. Bonin Pinto², J.G. Ortiz¹, A. Mansour¹ and C. Lazaridis¹

¹University of Chicago Pritzker School of Medicine, Department of Neurology, Chicago, USA

²University of Sao Paulo, Neuroscience and Behavior Department- Psychology Institute, Sao Paulo, Brazil

Paroxysmal sympathetic hyperactivity (PSH) is a clinical syndrome commonly observed post stroke or Traumatic Brain Injury (TBI). A hypothesis of its pathophysiology is an abnormal lack of inhibition from CNS control centers causing an unopposed sympathetic drive. Current treatment is purely symptomatic not taking in consideration brain plasticity mechanism. tDCS has the potential to exert powerful neuromodulatory effects over thalamic modulation (cortico-thalamic pathway) increasing endogenous inhibitory systems. Therefore, tDCS could restore the connections between the descending inhibitory pathways improving PHS symptoms; nonetheless, there are no studies using tDCS in the PHS population.

Test tDCS safety and feasibility in PSH patient’s and collect initial clinical effects data as secondary aim. We hypothesize that anodal-tDCS can be safely applied in PHS patient’s and can lead to a decrease of PSH episodes when compared to sham.

Prospective cross-over, academic-driven, single center, randomized, sham- controlled, double blinded phase 1 trial evaluating safety and feasibility of a single session of active tDCS (anodal, 20min, 2mA, over M1) vs sham (72 hours washout period). 15 patients will be recruited and randomized over a one year period. The target population includes acute ischemic or hemorrhagic stroke, traumatic brain injury patient’s that meet criteria for PSH syndrome.

Primary outcome: Safety and feasibility indexed by the tdcs safety questionnaire and clinical observation of the patients. Secondary aim: evaluate clinical effects in order to provide data to support a sample size calculation for a potential second fully powered study. Secondary endpoints:Clinical outcomes measurements (HR,RR,Temp,SBP,ICP); neurophsyiologic measures (EEG) and neuro-imaging (Brain CT/MRI).

n/a

BONE MARROW MONONUCLEAR CELLS (BM-MNCS) TRANSPLANTATION IN ACUTE ISCHEMIC STROKE (IBIS TRIAL). AN ADAPTIVE DESIGN FOR A PHASE IIB-III RANDOMIZED MULTICENTER CELL THERAPY TRIAL

F. Moniche¹, I. Escudero¹, E. Zapata-Arriaza², V. Escamilla³, C. Calderon-Cabrera³, J. Martin³, A. Vega-Salvatierra⁴, F. Mancha⁴, M. Zapata⁴, R. Valverde⁵, E. Aguera⁵, M.A. Gamero⁶, S. Perez⁶, M. Moya⁷, J. Ortega², I. Gutierrez², A. Gonzalez², J. Montaner⁶; The IBIS trial Investigators

¹Hospital Virgen del Rocio, Neurology, Sevilla, Spain

²Hospital Virgen del Rocio, Interventional Neuroradiology, Sevilla, Spain

³Hospital Virgen del Rocio, Hematology, Sevilla, Spain

⁴Instituto de Biomedicina de Sevilla-IBiS, Neurovascular Lab, Seville, Spain

⁵Hospital Reina Sofia, Neurology, Cordoba, Spain

⁶Hospital Virgen Macarena, Neurology, Sevilla, Spain

⁷Hospital Puerta del Mar, Neurology, Cadiz, Spain

Adaptive trial design has been proposed as a means to increase the efficiency of randomized clinical trials. Although cell therapy with bone marrow mononuclear cells (BM-MNCs) seems to be safe in stroke patients, the efficacy of cells in humans is not known. We aim to evaluate efficacy of BM-MNCs in stroke patients with a seamless phase 2–3 trial.

IBIS trial is an ongoing multicenter prospective phase IIb, randomized, controlled (non-treated group as control), assessor-blinded, academic clinical trial of intra-arterial transplantation of autologous BM-MNC in acute ischemic stroke patients. Inclusion criteria: patients within 1–7 days from stroke onset and NIHSS score of 6–20. Seventy-six patients are planned to be randomized to BM-MNCs transplantation or control group (1:1). Injection of BM-MNCs is done with two different doses (2x10⁶/kg or 5x10⁶/kg in 1:1 proportion).

With 54 patients included, we have planned a seamless Phase 2–3 design adaptive design to select recalculate sample power when recruiting period will be almost finished, and to select the most promising cell dose. In stage 1, patients are randomly assigned to receive two different doses of BM-MNCs. A single dose will be selected for further testing at stage 2. The final analysis would be based on the combined data from the two stages. The primary outcome is the modified Rankin Scale at 180 days.

IBIS trial is designed as a confirmatory trial of efficacy of intra-arterial autologous BM-MNC in moderate-severe acute ischemic stroke patients, enhancing the likelihood of finding a true benefit.

The trial is registered as NCT02178657.

CLINICAL AND PROGNOSTIC VALUE OF LONG-TERM LESIONAL AND PERILESIONAL CHANGES IN SYMPTOMATIC LACUNAR INFARCTS

C. Loos¹, A. Mondelaers², S. Dekeyzer³, J. Schröder⁴, W. Saeys⁴, P. Vanacker^1,5,6 and L. Yperzeele^1,6

¹Neurovascular Reference Center- Antwerp University Hospital, Department of Neurology, Edegem, Belgium

²Neurovascular Reference Center- Antwerp University Hospital, Department of Radiology, Edegem, Belgium

³Antwerp University Hospital, Department of Radiology, Edegem, Belgium

⁴University of Antwerp, Department of Rehabilitation Sciences and Physiotherapy, Antwerp, Belgium

⁵az groeninge, Department of Neurology, Kortrijk, Belgium

⁶University of Antwerp, Research Group on Translational Sciences, Antwerp, Belgium

Cerebral small vessel disease (cSVD) causes lacunar strokes, cognitive impairment, gait disturbances and functional disability. Several cSVD features can be identified on MRI, including lacunar infarcts, white matter hyperintensities (WMH), perivascular spaces and cerebral microbleeds. Additionally, cavitation and perilesional changes, including WMH “caps” and “tracks”, have been described as long-term radiological changes of symptomatic lacunar infarcts. Insights into radiological evolution of cSVD features might increase knowledge of the natural disease course.

To investigate the clinical and prognostic value of long-term lesional (cavitation) and perilesional changes (WMH “caps” and “tracks”) of symptomatic lacunar infarcts in patients with cSVD.

We plan a prospective, observational, single center, cohort study in patients with a lacunar stroke syndrome and recent lacunar infarct on MRI (according to STRIVE criteria). At stroke onset, vascular risk factor assessment, 24-hour blood pressure measurements, functional outcome (mRS, gait parameters) and structural brain MRI (DWI, FLAIR, T2, T1 and T2*) will be performed. One and 3 year after stroke, reassessment, including vascular risk factor management, stroke recurrence, functional outcome (mRS, Stroke Impact Scale, gait parameters) and follow-up structural brain MRI (similar protocol) will be performed. We will assess changes in imaging characteristics of index lacunar infarcts and surrounding white matter (WMH “caps” and “tracks”) and record presence of cSVD features, using STRIVE criteria.

We will explore associations between WMH “caps”, WMH “tracks”, and different radiological and clinical outcome measurements. Furthermore, the relation between WMH “caps”, 24-hour blood pressure measurements and vascular risk factor management will be explored.

N/A

N/A

N/A

PROPRIOCEPTION AND FUNCTIONAL IMPAIRMENT IN UNILATERAL NEGLECT AFTER STROKE

G. Fisher¹, S.C. Gandevia² and D.S. Kennedy^1,2

¹University of Technology Sydney, Graduate School of Health, Sydney, Australia

²Neuroscience Research Australia, Motor Impairment Research Centre, Sydney, Australia

Unilateral neglect (UN) is a complex condition occurring after stroke characterised by the failure to report, respond, or orient to contralesional stimuli. UN is an independent predictor of poor functional recovery. The processes affected by UN are mediated by proprioception, a fundamental contributor to movement control and function in the world. Components of proprioception can be conceptualised into the categories ‘ peripheral’ and ‘central’ based on the origin of their contribution to the sense. Neither UN nor proprioception are consistently assessed in the literature or in clinical settings.

To determine if the presence of UN after stroke results in more frequent and severe deficits in peripherally mediated upper limb (UL) proprioception, and UL function compared to stroke without UN.

This is a prospective cohort study of age and gender matched stroke-affected participants with and without UN and healthy controls (n = 15 per group). UN is measured using the Catherine Bergego Scale, which assesses UN ecologically in all domains. Peripheral proprioception is assessed via UL joint position matching and reaction time using a custom handheld Inertial Motion Unit (IMU). UL function is assessed using the UL Physiological Profile, a novel and validated series of tests developed at Neuroscience Research Australia. Participants are recruited from inpatient rehabilitation units of three major Sydney hospitals, and assessed at admission to and discharge from the unit, and at six months follow-up.

N/A

The accurate, yet clinically usable IMU and the ecological assessment of UN will provide new insights into proprioceptive deficits in this population.

N/A

UNITED STATES

SECURITIES AND EXCHANGE COMMISSION

Washington,D.C. 20549

SCHEDULE 14A

ProxyStatement Pursuant to Section 14(a) of the Securities

Exchange Act of 1934

☑ Filed by theRegistrant ☐ Filed by a Party otherthan the Registrant

Check the appropriatebox:
☐	Preliminary ProxyStatement
☐	CONFIDENTIAL, FOR USE OF THECOMMISSION ONLY (AS PERMITTED BY RULE 14a-6(e)(2))
☐	Definitive ProxyStatement
☑	Definitive AdditionalMaterials
☐	Soliciting Material Pursuant toss.240.14a-12

GENERAL MOTORS COMPANY