Displaying One Session

Scientific Communications
Session Type
Scientific Communications
Room
Hall L3
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM

COMPLICATED CAROTID ARTERY PLAQUES AS A CAUSE OF CRYPTOGENIC STROKE – PRIMARY RESULTS OF THE CAROTID PLAQUE IMAGING IN ACUTE STROKE (CAPIAS) STUDY

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
02:30 PM - 02:40 PM

Abstract

Group Name

on behalf of the CAPIAS study team

Background And Aims

In up to 30% of ischemic stroke patients no definite cause can be established. The observational multicenter CAPIAS study (NCT01284933) examined the hypothesis that a substantial proportion of cryptogenic stroke (CS) cases relate to complicated (AHA-lesion-type-VI) non-stenosing carotid artery plaques.

Methods

We prospectively recruited 234 patients (age >49 years, plaque thickness >2mm; <70% (NASCET) stenosis) with acute ischemic stroke and a corresponding DWI-positive lesion in the territory of a single carotid artery. Plaque characteristics were determined qualitatively and quantitatively by high-resolution contrast-enhanced carotid MRI at 3T using dedicated surface coils. 104 patients with CS and 73 patients with cardioembolic or small vessel stroke (CES/SVS; TOAST criteria) were included in the final analysis. The pre-defined primary comparisons were i) the prevalence of complicated plaques (hemorrhage, thrombus, or fibrous cap rupture) ipsilateral vs. contralateral to the infarct in CS and ii) the prevalence of ipsilateral complicated plaques in CS vs. CES/SVS.

Results

The prevalence of complicated plaques in CS was significantly higher ipsilateral to the infarct (31%) compared to contralateral (13%, p=0.001). Moreover, the prevalence of ipsilateral complicated plaques was significantly higher in CS compared to CES/SVS (15%, p=0.021). The lipid rich necrotic core/ wall area index was larger in patients with CS compared to CES/SVS (p<0.05) while there was no difference in luminal narrowing.

Conclusions

Complicated non-stenosing ipsilateral carotid artery plaques account for a substantial proportion of CS. Diagnostic work-up in anterior circulation CS should include a short carotid MRI sequence to identify a subset of CS patients with atherothrombotic stroke.

Trial Registration Number

NCT01284933

Hide

PREDICTORS OF PROGRESSION OF PLAQUE IN CAROTID BIFURCATION

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
02:40 PM - 02:50 PM

Abstract

Group Name

The ANTIQUE Trial Investigators

Background And Aims

Carotid plaque progression belongs to factors increasing stroke risk. The aim was to identify factors influencing carotid plaque progression.

Methods

The ANTIQUE study (ClinicalTrials NCT02360137) participants completed sonographic controls during 3 years were enrolled to analysis. Duplex sonography of cervical arteries was performed in 6-month intervals with measurement of plaque width in carotids. Plaque width measurement error (ME) was set as 99th percentile of difference between 2 measurements of in 2-week interval. Stable and progressive plaques were defined as plaque width difference between initial and final measurements ˂1ME and >2ME, resp. Univariate and multivariate logistic regression analysis (LRA) was performed to identify factors (age, gender, body mass index, blood pressure, carotid plaque width, arterial hypertension, diabetes mellitus, coronary heart disease, atrial fibrillation, myocardial infarction, stroke, vascular surgery/stenting, smoking, alcohol use) influencing the plaque progression.

Results

Totally 1391 patients (466 males, age 67.2±9.2 years) were enrolled to the analysis. Stabile plaques in both carotids were detected in 332 patients. Progressive plaque in at least one carotid artery was detected in 255 patients. Higher age (66.7 vs. 69.5 years), male gender (37.7% vs. 49.4%), greater plaque width (2.61 vs. 3.12 mm), coronary heart disease (19.6% vs. 28.6%), vascular surgery/stenting in history (11.1% vs. 22.8%) and smoking (9.9% vs. 17.3%) were more frequently present in patients with progressive plaque (p˂0.05 in all cases). Multivariate LRA identified only plaque width (OR=1.850) as the independent factor influencing plaque progression.

Conclusions

Carotid plaque width (corresponding with stenosis severity) is the independent risk factor for plaque progression.

Supported by grant-MHCR17-31016A.

Trial Registration Number

ClinicalTrials Identifier: NCT02360137

Hide

PROGNOSIS POTENTIAL OF LEFT ATRIAL DYSFUNCTION EVALUATION IN ESUS PATIENTS

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
02:50 PM - 03:00 PM

Abstract

Background And Aims

Atrial dysfunction is related to the occurrence of ischemic stroke (IS) in patients with coronary heart disease without atrial fibrillation (AF). Morphological and fonctionnal analysis of left atrium may be useful to identify patients with an embolic stroke of undetermided source (ESUS) at high risk of atrial fibrillation (AF) or recurrent stroke (RS).

Methods

We conducted a monocentric study in 501 IS patients hospitalized in our stroke unit at Saint-Antoine hospital (Paris) between october 2016 and december 2017. Patients with an ESUS underwent transesophageal echography and prolonged cardiac monitoring. Decreased atrial reservoir strain (ARS) was defined as ≤ 23%. Dilatation of the left atrium was defined by a volume ≥ 34mL/m2. Follow-up is ongoing.

Results

We identified 90 patients of 501 with an ESUS (18%). Among them (mean age :61 years ; female : 39%), AF was diagnosed in 14%, significant patent foramen ovale in 12%, complex aortic atheroma in 5%, hematologic disorder in 4%, and neoplasy in 2%.RS occured in 9% of patients after 36 months. Atrial dilatation and decreased ARS were respectively reported in 33% and 32% of the patients. 30% of the patients with decreased ARS and 20 % of the patients with atrial dilatation presented AF during the follow up. 15% of the patients with decreased ARS and 7% of the patients with atrial dilatation presented a RS.

Conclusions

Morphological and functional impairment of the left atrium are frequent in ESUS and could be a marker of risk of AF or RS.

Trial Registration Number

Not applicable

Hide

ECHOCARDIOGRAPHIC MARKERS IN CRYPTOGENIC STROKE WITH INCIDENTAL PATENT FORAMEN OVALE

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
03:00 PM - 03:10 PM

Abstract

Group Name

The Mechanisms of Embolic Stroke Clarified by Transesophageal Echocardiography for Embolic Stroke of Undetermined Source/ Cryptogenic Stroke (CHALLENGE ESUS/CS) registry Collaborators

Background And Aims

This study aimed to identify the association between the Risk of Paradoxical Embolism (RoPE) score and echocardiographic markers in cryptogenic stroke (CS) with patent foramen ovale (PFO).

Methods

Multicenter observational registry of CS included 677 patients who underwent favorable echocardiography between April 2014 and December 2016. Patients with PFO were classified as having a stroke with pathogenic PFO (RoPE score of >6, n=32) and incidental PFO (RoPE score of =<6, n=264). We compared clinical characteristics, laboratory findings, echocardiographic markers, new-onset of atrial fibrillation (NOAF), recurrence and favorable outcome (modified Rankin scale score <2) at discharge between two groups. Multiple logistic regression analysis was performed to determine risk factors of the low RoPE score group.

Results

Predictors related to the low RoPE score group were higher NIHSS score (p=0.035), hyperlipidemia (p=0.003), higher log10 brain natriuretic peptide (p=0.032), left atrial (LA) enlargement (p<0.001), higher E/e’ (P=0.001), lower LA appendage flow velocity (p=0.001), not large PFO (p=0.021), and aortic arch atheroma (p=0.002). The NOAF and recurrence occurred only in patients with the low RoPE score group. (6%, p=0.232 and 4%, p=0.607) The number of patients with a favorable outcome was small in the low RoPE score group. (58% vs. 78%, p=0.035) An independent predictor of the low RoPE score group was LA enlargement (age- and sex-adjusted odds ratio, 1.14; 95 % confidence interval, 1.00-1.32; p=0.039).

Conclusions

CS patients with the RoPE score =<6 could have the risk of cardioembolism.

Hide

IMPROVING THE FAST TEST USING MACHINE LEARNING AND LASSO REGRESSION ON ROUTINELY COLLECTED PRE-HOSPITAL DATA, WITH INTERNAL AND EXTERNAL VALIDATION

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
03:10 PM - 03:20 PM

Abstract

Background And Aims

Under the Greater Manchester (GM) regional centralised acute stroke pathway, ambulance clinicians use FAST to indicate suspected stroke patients, who are sent directly to Hyper Acute Stroke Units (HASUs). An audit by GM Stroke Operational Delivery Network showed that 48% of patients diverted to HASUs were not experiencing strokes. We aimed to develop an add-on to FAST which will reduce the number of non-strokes presenting to HASUs.

Methods

We analysed all data recorded by the ambulance clinicians in the Patient Record Form (PRF) of patients presenting with a suspected stroke to Salford Royal Foundation Trust (SRFT) between 1 September 2015 and 28 February 2017. We developed models using lasso logistic regression (LLR) and random forest (RF) modelling, to predict the diagnosis of stroke. Model performance was assessed using bootstrap internal validation. Models were also developed using a subset of the PRF variables which matched those recorded by the North East Ambulance Service, to perform external validation (ongoing).

Results

A total of 4218 patients were admitted to SRFT with a suspected stroke. Using all variables, the random forest (AUC:0.91; 95%CI:0.91-0.92) outperformed the lasso logistic regression (AUC:0.74; 95%CI:0.73-0.76), corresponding to a reduction in non-strokes by 80%. Using the subset of matched variables led to the same conclusions with a small drop in the AUC (RF:0.89; 95%CI:0.88-0.89 vs. LLR:0.73; 95%CI:0.71-0.74), and a similar reduction in non-strokes.

Conclusions

Non-strokes presenting at HASUs can be reduced by up to 80% using routinely collected pre-hospital data, without greatly affecting the number of missed strokes.

Hide

HYPERSENSITIVE TROPONIN T AS A BIOMARKER FOR DIFFERENTIATING BETWEEN MIGRAINE AND TRANSIENT ISCHAEMIC ATTACK

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
03:20 PM - 03:30 PM

Abstract

Background And Aims

Migraine is the most frequent differential diagnosis of transient ischaemic attack (TIA). Differentiating between migraines and TIAs is often difficult, even in specialized TIA Clinics. The "Diagnosis of TIA" (DOT) score has been used for differentiating TIAs from other diagnoses. As primary aim, we investigated whether elevated hypersensitive troponin T (HS-TnT) levels were associated with the diagnosis of TIA vs. migraine. As secondary aim, we evaluated whether adding HS-TnT levels to the DOT score resulted in improved C-Statistic.

Methods

We retrospectively included patients with either a diagnosis of migraine or TIA who were evaluated in a specialized TIA Clinic at a Regional Stroke Centre. We estimated odds ratios (OR) and 95% confidence intervals for the association between elevated HS-TnT levels (14 ng/l, 99th-percentile) and the diagnosis of TIA vs. migraine by applying logistic regression analyses. We compared the C-statistic of the DOT score with vs. without HS-TnT levels.

Results

We included 132 TIA and 46 migraine patients. Median age was 70 years and 49% were females. Median (interquartile range) HS-TnT levels for TIA and migraine were 9.5 (5.0-17.3) and 4.5 (3.0-7.0), respectively. Elevated HS-TnT (OR 20.8; 3.25-434.1; p=0.01) was independently associated with a diagnosis of TIA. The addition of HS-TnT levels (C-Statistic 0.809) slightly improved the discriminatory performance of the DOT score (C-Statistic 0.767).

Conclusions

Accounting for the relatively small sample-size of this hypothesis-generating study, elevated HS-TnT levels were associated with a diagnosis of TIA relative to migraine. The discriminatory performance of the DOT score slightly increased after the addition of HS-TnT levels.

Trial Registration Number

Not applicable

Hide

MULTI-FUNCTION CONTRIBUTIONS OF GREY AND WHITE MATTER REGIONS INFERRED BY MULTIVARIATE GAME-THEORETICAL LESION ANALYSIS IN STROKE PATIENTS

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
03:30 PM - 03:40 PM

Abstract

Group Name

on behalf of the WAKE-UP investigators

Background And Aims

Univariate lesion symptom mapping (ULSM) is commonly applied for lesion-symptom inference in stroke but prone to false positive result. We applied a novel multivariate approach based on game theory (Multiperturbation Shapley value Analysis, MSA) to infer combined causal contributions of white and grey matter regions to brain function.

Methods

Lesion patterns of 452 acute ischemic stroke patients from the WAKE-UP trial were analysed together with four behavioural factors based on NIHSS sub-scores: language, spatial attention, and right and left motor function. We defined 295 regions based on atlases of grey and white matter regions (Brainnetome and JHU-ICBM atlas). We employed estimated MSA and bootstrap validation to obtain functional contributions of each region to the four predefined NIHSS factors.

Results

Significant contributions of brain regions related to language (factor 1) are shown below. Clinical deficits in this category were mostly related to lesions localized in the left insula, inferior frontal gyrus and thalamic radiation. For the remaining factors (not shown), deficits in spatial attention were predominantly linked to the right superior temporal gyrus and striatum as main contributors. The contralateral anterior and posterior parts of the internal capsule were identified as main contributors for motor function.factor1foresoc.png

Conclusions

In our study, we validate the application of MSA in a large dataset of stroke patients. Our findings demonstrate that main contributions to language and spatial attention arise from grey matter regions, whereas white matter tracts contribute significantly to motor function. In addition we explore functional interactions between brain regions to highlight overlapping and synergistic interactions.

Trial Registration Number

Not applicable

Hide

CARDIAC DYSFUNCTION AND FUNCTIONAL OUTCOME IN ISCHEMIC STROKE PATIENTS WITH DIABETES: PRELIMINARY RESULTS FROM THE SICFAIL (STROKE-INDUCED CARDIAC FAILURE IN MICE AND MEN) COHORT STUDY

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
03:40 PM - 03:50 PM

Abstract

Background And Aims

Patients with diabetes exhibit a higher burden of cardiac comorbidity and have worse prognosis after ischemic stroke (IS) than non-diabetics. We investigated the association of systolic dysfunction (SD), diastolic dysfunction (DD), and clinically overt heart failure (HF) with poor functional outcome (mRS>2) at one year in diabetic IS patients.

Methods

Baseline data (including medical history, clinical examination, echocardiography, and blood sampling) and one-year mRS were prospectively collected within the hospital-based single-centre SICFAIL study. We defined diabetes as self-report, intake of anti-diabetic medication or elevated HbA1c values at baseline (01/2014-02/2017). We developed an initial logistic regression model for mRS>2 including baseline characteristics, risk factors, and comorbidities and calculated the model’s accuracy using the area under curve (AUC). Additionally, we added SD, DD and HF into separate models. Odds Ratios (OR) with 95% confidence intervals were reported.

Results

Of 523 IS patients with information on echocardiography and one-year mRS, 148 (28%) had diabetes at baseline (median age=74 years, 61% male). Of them, 59 (39%) had mRS>2. In the initial model (AUC=0.86, 95% CI (0.79-0.93)), independent predictors of mRS>2 in diabetic IS patients were higher age (per year: 1.08, 1.02-1.14), NIHSS>4 (5.0, 1.79-14.0), coronary heart disease (3.68, 1.33-10.17), and elevated blood glucose (per mg/dl: 1.02, 1.01-1.03). SD, DD and HF were not statistically significant associated with mRS>2 when adding to this model.

Conclusions

Cardiac dysfunction at baseline was not independently associated with poor one-year outcome in diabetic IS patients. Ongoing follow-up will provide further insights regarding the impact of cardiac dysfunction on long-term prognosis after IS.

Trial Registration Number

DRKS00011615

Hide

RISK OF VASCULAR EVENTS AFTER A FIRST-EVER ISCHEMIC STROKE IN A POPULATION-BASED STROKE REGISTRY

Session Type
Scientific Communications
Date
07.11.2020, Saturday
Session Time
02:30 PM - 04:00 PM
Room
Hall L3
Lecture Time
03:50 PM - 04:00 PM
Presenter

Abstract

Background And Aims

To assess the prevalence and predictors of vascular events during the follow-up period among patients with a first-ever ischemic stroke (FEIS) in a population-based registry.

Methods

All the residents in the district of L’Aquila diagnosed with a FEIS in 2011-2013 were included in the registry and were followed up to 5 years. Case-fatality and vascular events were assessed.

Results

Among the 1,280 patients with first-ever stroke included in the registry, 910 (71.1%) had a FEIS; 753 (82.7%) patients (50.7% women; mean age78.5, standard deviation 11.2 years) survived more than 30 days after the event and were included in the analyses. At the 5-year follow-up, 107 (14.2%) patients had a vascular event, including 43 (5.7%) fatal events; 67 (8.9%) patients had a FEIS recurrence, 22 (2.9%) a transient ischemic attack, 15 (2.0 %) a myocardial infarction, and 3 (0.4%) a primary intracerebral hemorrhage. The rate of nonfatal stroke recurrence (59; 7.8%) was sevenfold higher than of that of fatal stroke (8; 1.1%). The Cox regression analysis, adjusted by sex, age, and vascular risk factors, showed that only diabetes mellitus (hazard ratio 2.34; 95% confidence interval, 1.29-4.27; P=0.005) was an independent predictor of vascular events at 5 years.

Conclusions

The rates of recurrent stroke within 5 years from the FEIS is more than fourfold higher that of cardiac events. The results of our study reinforce the need for more effective targeted secondary stroke prevention probably achievable with a careful evaluation of stroke etiology and strict management of cardiovascular risk factors.

Trial Registration Number

Not applicable

Hide