Random expert sampling for deep learning segmentation of acute ischemic stroke on non-contrast CT.

BACKGROUND: Outlining acutely infarcted tissue on non-contrast CT is a challenging task for which human inter-reader agreement is limited. We explored two different methods for training a supervised deep learning algorithm: one that used a segmentation defined by majority vote among experts and another that trained randomly on separate individual expert segmentations. METHODS: The data set consisted of 260 non-contrast CT studies in 233 patients with acute ischemic stroke recruited from the multicenter DEFUSE 3 (Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3) trial. Additional external validation was performed using 33 patients with matched stroke onset times from the University Hospital Lausanne. A benchmark U-Net was trained on the reference annotations of three experienced neuroradiologists to segment ischemic brain tissue using majority vote and random expert sampling training schemes. The median of volume, overlap, and distance segmentation metrics were determined for agreement in lesion segmentations between (1) three experts, (2) the majority model and each expert, and (3) the random model and each expert. The two sided Wilcoxon signed rank test was used to compare performances (1) to 2) and (1) to (3). We further compared volumes with the 24 hour follow-up diffusion weighted imaging (DWI, final infarct core) and correlations with clinical outcome (modified Rankin Scale (mRS) at 90 days) with the Spearman method. RESULTS: The random model outperformed the inter-expert agreement ((1) to (2)) and the majority model ((1) to (3)) (dice 0.51±0.04 vs 0.36±0.05 (P<0.0001) vs 0.45±0.05 (P<0.0001)). The random model predicted volume correlated with clinical outcome (0.19, P<0.05), whereas the median expert volume and majority model volume did not. There was no significant difference when comparing the volume correlations between random model, median expert volume, and majority model to 24 hour follow-up DWI volume (P>0.05, n=51). CONCLUSION: The random model for ischemic injury delineation on non-contrast CT surpassed the inter-expert agreement ((1) to (2)) and the performance of the majority model ((1) to (3)). We showed that the random model volumetric measures of the model were consistent with 24 hour follow-up DWI.

Acute ischaemic stroke in active cancer versus non-cancer patients: stroke characteristics, mechanisms and clinical outcomes.

BACKGROUND AND PURPOSE: Demographics, clinical characteristics, stroke mechanisms and long-term outcomes were compared between acute ischaemic stroke (AIS) patients with active cancer (AC) versus non-cancer patients. METHODS: Using data from 2003 to 2021 in the Acute STroke Registry and Analysis of Lausanne, a retrospective cohort study was performed comparing patients with AC, including previously known and newly diagnosed cancers, with non-cancer patients. Patients with inactive cancer were excluded. Outcomes were the modified Rankin Scale (mRS) score at 3 months, death and cerebrovascular recurrences at 12 months before and after propensity score matching. RESULTS: Amongst 6686 patients with AIS, 1065 (15.9%) had a history of cancer. After excluding 700 (10.4%) patients with inactive cancer, there were 365 (5.5%) patients with AC and 5621 (84%) non-cancer AIS patients. Amongst AC patients, 154 (42.2%) strokes were classified as cancer related. In multivariable analysis, patients with AC were older (adjusted odds ratio [aOR] 1.02, 95% confidence interval [CI] 1.00-1.03), had fewer vascular risk factors and were 48% less likely to receive reperfusion therapies (aOR 0.52, 95% CI 0.35-0.76). Three-month mRS scores were not different in AC patients (aOR 2.18, 95% CI 0.96-5.00). At 12 months, death (adjusted hazard ratio 1.91, 95% CI 1.50-2.43) and risk of cerebrovascular recurrence (sub-distribution hazard ratio 1.68, 95% CI 1.22-2.31) before and after propensity score matching were higher in AC patients. CONCLUSIONS: In a large institutional registry spanning nearly two decades, AIS patients with AC had less past cerebrovascular disease but a higher 1-year risk of subsequent death and cerebrovascular recurrence compared to non-cancer patients. Antithrombotic medications at discharge may reduce this risk in AC patients.

Ischemic Stroke in Cancer: Mechanisms, Biomarkers, and Implications for Treatment.

Ischemic stroke is an important cause of morbidity and mortality in cancer patients. The underlying mechanisms linking cancer and stroke are not completely understood. Long-standing and more recent evidence suggests that cancer-associated prothrombotic states, along with treatment-related vascular toxicity, such as with chemotherapy and immunotherapy, contribute to an increased risk of ischemic stroke in cancer patients. Novel biomarkers, including coagulation, platelet and endothelial markers, cell-free DNA, and extracellular vesicles are being investigated for their potential to improve risk stratification and patient selection for clinical trials and to help guide personalized antithrombotic strategies. Treatment of cancer-related stroke poses unique challenges, including the need to balance the risk of recurrent stroke and other thromboembolic events with that of bleeding associated with antithrombotic therapy. In addition, how and when to restart cancer treatment after stroke remains unclear. In this review, we summarize current knowledge on the mechanisms underlying ischemic stroke in cancer, propose an etiological classification system unique to cancer-related stroke to help guide patient characterization, provide an overview of promising biomarkers and their clinical utility, and discuss the current state of evidence-based management strategies for cancer-related stroke. Ultimately, a personalized approach to stroke prevention and treatment is required in cancer patients, considering both the underlying cancer biology and the individual patient's risk profile.

Undiagnosed major risk factors in acute ischaemic stroke patients: frequency, profile, stroke mechanisms and outcome.

BACKGROUND AND PURPOSE: There is scarce clinical information about the clinical profile of patients with acute ischaemic stroke with previously undiagnosed major vascular risk factors (UMRFs). METHODS: This was a retrospective analysis of data from the Acute Stroke Registry and Analysis of Lausanne registry between 2003 and 2018 with univariate and multivariate logistic regression analyses comparing clinical profiles of patients with UMRFs to patients with at least one previously diagnosed MRF (DMRF). RESULTS: In all, 4354 patients (median age 70 years [interquartile range 15.2], 44.7% female) were included after excluding 763 (14.9%) for lack of consent and three for missing information. Amongst 1125 (25.8%) UMRF patients, 69.7% (n = 784) had at least one newly diagnosed MRF and the others none. The newly detected MRFs were dyslipidaemia (61.4%), hypertension (23.7%), atrial fibrillation (10.2%), diabetes mellitus (5.2%), ejection fraction <35% (2.0%) and coronary disease (1.0%). Comparing UMRF patients to DMRF patients, multivariate analysis showed a positive association with lower age, non-Caucasian ethnicity, contraceptive use (<55 years old), smoking (≥55 years old) and patent-foramen-ovale-related stroke mechanism. A negative association was found with pre-stroke antiplatelet use and higher body mass index. Functional outcome did not differ. Cerebrovascular recurrences were similar between groups. CONCLUSIONS: In this large single-centre cohort, 69.7% of patients with acute ischaemic stroke and UMRF were newly diagnosed with at least one new MRF, the most common being dyslipidaemia, hypertension or atrial fibrillation. Patients of the UMRF group were younger, more often smokers and on contraceptives, and had more patent-foramen-ovale-related strokes.

Incidence and outcome of perforations during medium vessel occlusion compared with large vessel occlusion thrombectomy.

BACKGROUND: Vessel perforation during thrombectomy is a severe complication and is hypothesized to be more frequent during medium vessel occlusion (MeVO) thrombectomy. The aim of this study was to compare the incidence and outcome of patients with perforation during MeVO and large vessel occlusion (LVO) thrombectomy and to report on the procedural steps that led to perforation. METHODS: In this multicenter retrospective cohort study, data of consecutive patients with vessel perforation during thrombectomy between January 1, 2015 and September 30, 2022 were collected. The primary outcomes were independent functional outcome (ie, modified Rankin Scale 0-2) and all-cause mortality at 90 days. Binomial test, chi-squared test and t-test for unpaired samples were used for statistical analysis. RESULTS: During 25 769 thrombectomies (5124 MeVO, 20 645 LVO) in 25 stroke centers, perforation occurred in 335 patients (1.3%; mean age 72 years, 62% female). Perforation occurred more often in MeVO thrombectomy (2.4%) than in LVO thrombectomy (1.0%, p<0.001). More MeVO than LVO patients with perforation achieved functional independence at 3 months (25.7% vs 10.9%, p=0.001). All-cause mortality did not differ between groups (overall 51.6%). Navigation beyond the occlusion and retraction of stent retriever/aspiration catheter were the two most common procedural steps that led to perforation. CONCLUSIONS: In our cohort, perforation was approximately twice as frequent in MeVO than in LVO thrombectomy. Efforts to optimize the procedure may focus on navigation beyond the occlusion site and retraction of stent retriever/aspiration catheter. Further research is necessary in order to identify thrombectomy candidates at high risk of intraprocedural perforation and to provide data on the effectiveness of endovascular countermeasures.

IV Thrombolysis With or Without Endovascular Treatment for Suspected Ischemic Stroke in Patients With Intracranial Tumors.

BACKGROUND AND OBJECTIVES: IV thrombolysis (IVT) for suspected ischemic stroke in patients with intracranial neoplasms is off-label. However, data on risks of intracranial hemorrhage (ICH) are scarce. METHODS: In a multicenter registry-based analysis within the European Thrombolysis in Ischemic Stroke Patients (TRISP) collaboration, we assessed frequencies of symptomatic and fatal ICH after IVT for suspected ischemic stroke in patients with intracranial tumors by descriptive statistics and analyzed associations with clinical and imaging characteristics by binary logistic regression. Definition of symptomatic ICH was based on the clinical criteria of the European Cooperative Acute Stroke-II trial including hemorrhage at any site in cranial imaging and concurrent clinical deterioration. RESULTS: Screening data of 21,289 patients from 14 centers, we identified 105 patients receiving IVT; among them were 29 patients (28%) with additional endovascular treatment, with suspected, that is, imaging-based, or histologically confirmed diagnosis of intracranial tumors. Among 104 patients with CT or MRI after IVT available, symptomatic and fatal ICH were observed in 9 and 4 patients (9% and 4%, respectively). Among 82 patients with suspected or confirmed meningioma, symptomatic and fatal ICH occurred in 6 and 3 patients (7% and 4%), respectively. In 18 patients with intra-axial suspected or confirmed primary or secondary brain tumors, there was 1 symptomatic nonfatal ICH (6%). Of 4 patients with tumors of the pituitary region, 2 patients (50%) had symptomatic ICH including 1 fatal ICH (25%). Tumor size was not associated with the occurrence of symptomatic ICH (odds ratio 2.8, 95% CI 0.3-24.8, p = 0.34). DISCUSSION: In our dataset from routine clinical care, we provide insights on the safety of IVT for suspected ischemic stroke in patients with intracranial tumors, a population that is commonly withheld thrombolysis in clinical practice and prospective trials. Except for a potential high risk of symptomatic ICH after IVT in patients with tumors of the pituitary region, frequencies of symptomatic ICH in patients with intracranial tumors in our cohort seem to be in the upper range of rates observed in previous studies within the TRISP cooperation. These results may guide individual treatment decisions in patients with acute stroke and intracranial tumors with potential benefit of IVT.

Interaction between intravenous thrombolysis and clinical outcome between slow and fast progressors undergoing mechanical thrombectomy: a post-hoc analysis of the SWIFT-DIRECT trial.

BACKGROUND: In proximal occlusions, the effect of reperfusion therapies may differ between slow or fast progressors. We investigated the effect of intravenous thrombolysis (IVT) (with alteplase) plus mechanical thrombectomy (MT) versus thrombectomy alone among slow versus fast stroke progressors. METHODS: The SWIFT-DIRECT trial data were analyzed: 408 patients randomized to IVT+MT or MT alone. Infarct growth speed was defined by the number of points of decay in the initial Alberta Stroke Program Early CT Score (ASPECTS) divided by the onset-to-imaging time. The primary endpoint was 3-month functional independence (modified Rankin scale 0-2). In the primary analysis, the study population was dichotomized into slow and fast progressors using median infarct growth velocity. Secondary analysis was also conducted using quartiles of ASPECTS decay. RESULTS: We included 376 patients: 191 IVT+MT, 185 MT alone; median age 73 years (IQR 65-81); median initial National Institutes of Health Stroke Scale (NIHSS) 17 (IQR 13-20). The median infarct growth velocity was 1.2 points/hour. Overall, we did not observe a significant interaction between the infarct growth speed and the allocation to either randomization group on the odds of favourable outcome (P=0.68). In the IVT+MT group, odds of any intracranial hemorrhage (ICH) were significantly lower in slow progressors (22.8% vs 36.4%; OR 0.52, 95% CI 0.27 to 0.98) and higher among fast progressors (49.4% vs 26.8%; OR 2.62, 95% CI 1.42 to 4.82) (P value for interaction <0.001). Similar results were observed in secondary analyses. CONCLUSION: In this SWIFT-DIRECT subanalysis, we did not find evidence for a significant interaction of the velocity of infarct growth on the odds of favourable outcome according to treatment by MT alone or combined IVT+MT. However, prior IVT was associated with significantly reduced occurrence of any ICH among slow progressors whereas this was increased in fast progressors.

Clinical and Demographic Characteristics, Mechanisms, and Outcomes in Patients With Acute Ischemic Stroke and Newly Diagnosed or Known Active Cancer.

BACKGROUND AND OBJECTIVES: Patients with a new diagnosis of cancer carry an increased risk of acute ischemic stroke (AIS), and this risk varies depending on age, cancer type, stage, and time from diagnosis. Whether patients with AIS with a new diagnosis of neoplasm represent a distinct subset from those with a previously known active malignancy remains unclear. We aimed to estimate the rate of stroke in patients with newly diagnosed cancer (NC) and previously known active cancer (KC) and to compare the demographic and clinical features, stroke mechanisms, and long-term outcomes between groups. METHODS: Using 2003-2021 data from the Acute STroke Registry and Analysis of Lausanne registry, we compared patients with KC with patients with NC (cancer identified during AIS hospitalization or within the following 12 months). Patients with inactive and no history of cancer were excluded. Outcomes were the modified Rankin scale (mRS) score at 3 months and mortality and recurrent stroke at 12 months. We used multivariable regression analyses to compare outcomes between groups while adjusting for important prognostic variables. RESULTS: Among 6,686 patients with AIS, 362 (5.4%) had active cancer (AC), including 102 (1.5%) with NC. Gastrointestinal and genitourinary cancers were the most frequent cancer types. Among all patients with AC, 152 (42.5%) AISs were classified as cancer related, with nearly half of these cases attributed to hypercoagulability. In multivariable analysis, patients with NC had less prestroke disability (adjusted odds ratio [aOR] 0.62, 95% CI 0.44-0.86) and fewer prior stroke/transient ischemic attack events (aOR 0.43, 95% CI 0.21-0.88) than patients with KC. Three-month mRS scores were similar between cancer groups (aOR 1.27, 95% CI 0.65-2.49) and were predominantly driven by the presence of newly diagnosed brain metastases (aOR 7.22, 95% CI 1.49-43.17) and metastatic cancer (aOR 2.19, 95% CI 1.22-3.97). At 12 months, mortality risk was higher in patients with NC vs patients with KC (hazard ratio [HR] 2.11, 95% CI 1.38-3.21), while recurrent stroke risk was similar between groups (adjusted HR 1.27, 95% CI 0.67-2.43). DISCUSSION: In a comprehensive institutional registry spanning nearly 2 decades, 5.4% of patients with AIS had AC, a quarter of which were diagnosed during or within 12 months after the index stroke hospitalization. Patients with NC had less disability and prior cerebrovascular disease, but a higher 1-year risk of subsequent death than patients with KC.

[Neurology: what's new in 2022]. / Neurologie : ce qui a changé en 2022.

The year 2022 was marked by the development of numerous new treatments for refractory myasthenia gravis. The link between epilepsy and cerebrovascular disorder was studied and lamotrigine discovered to be the optimal treatment choice for epilepsy secondary to stroke to prevent mortality on patient of 45 years and older. New randomized study finally demonstrated the utility of thrombectomy in selected patients with basilar artery occlusion. The causal relationship between Epstein-Barr infection and multiple sclerosis has been proved thanks to a large cohort study. A new possibility of subcutaneous continuous levodopa administration gave promising result. Finally, numerous studies confirmed the efficacy and excellent tolerability of anti-CGRP antibodies.

L'année 2022 a été marquée par l'arrivée de nombreux traitements pour la myasthénie réfractaire. Le lien entre l'épilepsie et le risque cérébro-vasculaire a été bien étudié, démontrant que la lamotrigine semble être le meilleur traitement pour prévenir la mortalité chez les patients de 45 ans et plus. De nouvelles études ont enfin pu établir l'utilité de la thrombectomie dans les occlusions basilaires. Le lien entre le virus d'Epstein-Barr et la sclérose en plaques a pu être prouvé à la suite d'une importante étude de cohorte. Une nouvelle technique d'administration sous-cutanée de la lévodopa semble prometteuse. Enfin, de nombreuses études confirment l'efficacité et l'excellente tolérance des anticorps anti-CGRP (Calcitonine Gene Related Protein).

Goals of care changes after acute ischaemic stroke: decision frequency and predictors.

OBJECTIVES: Little is known about the factors leading to a change in goals of care (CGC) in patients with an acute ischaemic stroke (AIS). Our aim was to analyse the proportion and outcome of such patients and identify medical predictors of a CGC during acute hospitalisation. METHODS: We retrospectively reviewed all patients who had an AIS over a 13-year period from the prospectively constructed Acute Stroke Registry and Analysis of Lausanne. We compared patients with a CGC during the acute hospital phase to all other patients and identified associated clinical and radiological variables using logistic regression analysis. RESULTS: A CGC decision was taken in 440/4264 (10.3%) consecutive patients who had an AIS. The most powerful acute phase predictors of a CGC were transit through the intensive care unit, older age, pre-existing disability, higher stroke severity and initial decreased level of consciousness. Adding subacute phase variables, we also identified active oncological disease, fever and poor recanalisation as predictors. 76.6% of the CGC patients died in the stroke unit and 1.0% of other patients, and 30.5% of patients with a CGC received a palliative care consultation. At 12 months, 93.6% of patients with CGC had died, compared with 10.1% of non-CGC patients. CONCLUSIONS: Over three-quarters of AIS patients with CGC died in hospital, but less than a third received a palliative care consultation. The identified clinical and radiological predictors of a CGC may allow physicians to initiate timely the decision-making process for a possible CGC.

Rare stroke mechanisms in 4154 consecutive patients: causes, predictors, treatment, and outcomes.

BACKGROUND: Rare mechanisms of stroke (RMS) in acute ischemic stroke (AIS) have rarely been studied applying a systematic approach. Our aim was to define the frequency, etiologies, predictors, and outcomes of RMS in a consecutive series of AIS. METHODS: Data from consecutive patients from 2003 to 2016 were derived from the Acute STroke Registry and Analysis of Lausanne (ASTRAL). Frequency of subcategories of RMS was calculated. In a case-control design, RMS were compared to strokes of all other mechanisms. Outcome was assessed with 3-month Rankin-shift and 12-month mortality and recurrence rates. RESULTS: Out of 4154 AISs, 222 (5.3%) were found to have a RMS (42.0% female, median age 66 years). The most frequent RMS etiologies were medical interventions (25.6%), active oncological disease (22.5%), and vasculitis (11.7%). In multivariate analysis, RMS patients were younger, had more preceding and bilateral strokes, and a higher admission temperature. They were associated with less traditional risk factors and more systemic disease (such as AIDS, coagulopathy, and cancer). RMS also had more early ischemic changes on plain CT, less revascularization treatments, and more symptomatic hemorrhagic transformations. They presented significantly higher 3-month disability (Rankin-shift-ORadj 1.74), 12-month recurrence (ORadj 1.99), and mortality rates (ORadj 2.41). CONCLUSIONS: RMS occurred in 5.3% of a large population of consecutive AISs and are most frequently related to medical interventions, cancer, and vasculitis. RMS patients have less traditional risk factors but more systemic comorbidities, hemorrhagic transformations, recurrences, and a worse long-term outcome. Identification of RMS has direct implications for early treatment and long-term outcome.

[Neurology : what's new in 2021]. / Neurologie.

In 2021, we assisted to the publication of new diagnostic criteria, classifications, and guidelines (CIDP, brain tumors, auto-immune encephalitis). Several studies helped to define the pharmacological management of focal and generalized epileptic seizures and epilepsy in pregnant women. The availability of biomarkers and the approval of immunotherapies are modifying the landscape of dementia management. Endovascular interventions without previous thrombolysis seems to be effective in anterior circulation acute ischemic stroke (AIS) and severe posterior circulation AIS. Neurologic complications of Sars-CoV-2 infection were further studied, as well as the efficacy of vaccines in immunosuppressed patients. New molecules and techniques show promising results for the treatment of migraine and cluster headache.

L'année 2021 a été marquée par la publication des nouveaux critères diagnostiques, classifications et guidelines (polyradiculonévrite inflammatoire démyélinisante chronique, tumeurs cérébrales, encéphalites autoimmunes). L'attitude thérapeutique dans les épilepsies focales ou généralisées et l'épilepsie chez la femme enceinte a été mieux définie. Les marqueurs biologiques et les immunothérapies modifient le paysage de la prise en charge des démences. Le traitement endovasculaire des AVC de la circulation antérieure semble efficace indépendamment d'une thrombolyse préalable, ainsi qu'en cas d'AVC sévère de la circulation postérieure. Les complications neurologiques du SARS-CoV-2 ont été éclaircies et l'efficacité des vaccins étudiée chez les patients immunosupprimés. Plusieurs nouvelles molécules et techniques montrent des résultats prometteurs pour les migraines et céphalées en grappe.

Intima-Media Thickness and Pulsatility Index of Common Carotid Arteries in Acute Ischaemic Stroke Patients with Diabetes Mellitus.

Ultrasonographic parameters such as the common carotid artery (CCA) pulsatility index (PI) and CCA intima-media thickness (IMT) have been associated with an increased mortality and risk of recurrent stroke, respectively. We hypothesized that these ultrasonographic parameters may be useful for monitoring diabetic patients after an acute stroke. We analysed retrospective data of consecutive acute ischaemic stroke patients from the ASTRAL registry who underwent pre-cerebral ultrasonographic evaluation within 7 days of symptom onset. We compared clinical, demographic, radiological and ultrasonographic parameters in diabetic versus non-diabetic patients (univariable and multivariable analyses) and the association of these parameters with CCA PI and CCA IMT. We analysed 1507 carotid duplex ultrasound examinations from patients with a median age of 74 years. Cardiovascular co-morbidities, including hypertension, hypercholesterolemia, obstructive sleep apnoea syndrome, higher body-mass index (BMI) and peripheral artery disease, were associated with diabetes mellitus (DM). Diabetics were more often under antiplatelet therapy and had atrial fibrillation at admission. Diabetic patients showed an increased CCA PI and IMT in line with more atherosclerotic changes on acute CTA compared to non-diabetic patients. Taking IMT as the dependent variable in a second analysis, DM, higher age, hypertension, smoking and CCA PI were associated with higher IMT. Taking CCA PI as the dependent variable in a third analysis, DM, higher age and higher NIHSS at admission were associated with higher CCA PI values. Increased IMT was also associated with higher PI. We show that CCA PI and IMT are higher in diabetic patients in the first week after an initial stroke.

Extracellular Vesicle Surface Markers as a Diagnostic Tool in Transient Ischemic Attacks.

Background and Purpose: Extracellular vesicles (EVs) are promising biomarkers for cerebral ischemic diseases, but not systematically tested in patients with transient ischemic attacks (TIAs). We aimed at (1) investigating the profile of EV-surface antigens in patients with symptoms suspicious for TIA; (2) developing and validating a predictive model for TIA diagnosis based on a specific EV-surface antigen profile. Methods: We analyzed 40 subjects with symptoms suspicious for TIA and 20 healthy controls from a training cohort. An independent cohort of 28 subjects served as external validation. Patients were stratified according to likelihood of having a real ischemic event using the Precise Diagnostic Score, defined as: unlikely (score 0­1), possible-probable (score 2­3), or very likely (score 4­8). Serum vesicles were quantified by nanoparticle tracking analysis and EV-surface antigen profile characterized by multiplex flow cytometry. Results: EV concentration increased in patients with very likely or possible-probable TIA (P<0.05) compared with controls. Nanoparticle concentration was directly correlated with the Precise Diagnostic score (R=0.712; P<0.001). After EV immuno-capturing, CD8, CD2, CD62P, melanoma-associated chondroitin sulfate proteoglycan, CD42a, CD44, CD326, CD142, CD31, and CD14 were identified as discriminants between groups. Receiver operating characteristic curve analysis confirmed a reliable diagnostic performance for each of these markers taken individually and for a compound marker derived from their linear combinations (area under the curve, 0.851). Finally, a random forest model combining the expression levels of selected markers achieved an accuracy of 96% and 78.9% for discriminating patients with a very likely TIA, in the training and external validation cohort, respectively. Conclusions: The EV-surface antigen profile appears to be different in patients with transient symptoms adjudicated to be very likely caused by brain ischemia compared with patients whose symptoms were less likely to due to brain ischemia. We propose an algorithm based on an EV-surface-antigen specific signature that might aid in the recognition of TIA.

Automatic segmentation, feature extraction and comparison of healthy and stroke cerebral vasculature.

Accurate segmentation of cerebral vasculature and a quantitative assessment of its morphology is critical to various diagnostic and therapeutic purposes and is pertinent to studying brain health and disease. However, this is still a challenging task due to the complexity of the vascular imaging data. We propose an automated method for cerebral vascular segmentation without the need of any manual intervention as well as a method to skeletonize the binary segmented map to extract vascular geometric features and characterize vessel structure. We combine a Hessian-based probabilistic vessel-enhancing filtering with an active-contour-based technique to segment magnetic resonance and computed tomography angiograms (MRA and CTA) and subsequently extract the vessel centerlines and diameters to calculate the geometrical properties of the vasculature. Our method was validated using a 3D phantom of the Circle-of-Willis region, demonstrating 84% mean Dice similarity coefficient (DSC) and 85% mean Pearson's correlation coefficient (PCC) with minimal modified Hausdorff distance (MHD) error (3 surface pixels at most), and showed superior performance compared to existing segmentation algorithms upon quantitative comparison using DSC, PCC and MHD. We subsequently applied our algorithm to a dataset of 40 subjects, including 1) MRA scans of healthy subjects (n = 10, age = 30 ± 9), 2) MRA scans of stroke patients (n = 10, age = 51 ± 15), 3) CTA scans of healthy subjects (n = 10, age = 62 ± 12), and 4) CTA scans of stroke patients (n = 10, age = 68 ± 11), and obtained a quantitative comparison between the stroke and normal vasculature for both imaging modalities. The vascular network in stroke patients compared to age-adjusted healthy subjects was found to have a significantly (p < 0.05) higher tortuosity (3.24 ± 0.88 rad/cm vs. 7.17 ± 1.61 rad/cm for MRA, and 4.36 ± 1.32 rad/cm vs. 7.80 ± 0.92 rad/cm for CTA), higher fractal dimension (1.36 ± 0.28 vs. 1.71 ± 0.14 for MRA, and 1.56 ± 0.05 vs. 1.69 ± 0.20 for CTA), lower total length (3.46 ± 0.99 m vs. 2.20 ± 0.67 m for CTA), lower total volume (61.80 ± 18.79 ml vs. 34.43 ± 22.9 ml for CTA), lower average diameter (2.4 ± 0.21 mm vs. 2.18 ± 0.07 mm for CTA), and lower average branch length (4.81 ± 1.97 mm vs. 8.68 ± 2.03 mm for MRA), respectively. We additionally studied the change in vascular features with respect to aging and imaging modality. While we observed differences between features as a result of aging, statistical analysis did not show any significant differences, whereas we found that the number of branches were significantly different (p < 0.05) between the two imaging modalities (201 ± 73 for MRA vs. 189 ± 69 for CTA). Our segmentation and feature extraction algorithm can be applied on any imaging modality and can be used in the future to automatically obtain the 3D segmented vasculature for diagnosis and treatment planning as well as to study morphological changes due to stroke and other cerebrovascular diseases (CVD) in the clinic.

A score to predict one-year risk of recurrence after acute ischemic stroke.

BACKGROUND: An acute ischemic stroke carries a substantial risk of further recurrences. We aimed at developing and validating a prognostic tool to predict one-year stroke recurrence after acute ischemic stroke. METHODS: An integer score was derived by Cox regression analysis on a hospital-referred cohort of 3246 acute ischemic stroke patients from Switzerland, and tested for external validity in three similar independent cohorts from Athens (n = 2495), Milan (n = 1279), and Helsinki (n = 714) by means of calibration and discrimination. RESULTS: In the derivation cohort, the recurrence rate was 7% (n = 228/3246). We developed a nine-point score comprising: previous stroke or transient ischemic attack (1-point), stroke mechanism (small vessel disease and unknown mechanism: 0-points; rare stroke mechanism: 3-points; other mechanisms: 1-point), pre-stroke antiplatelets (1-point), active malignancy (2-points), chronic cerebrovascular lesions on imaging (1-point) and absence of early ischemic changes on first imaging (1-point). In the derivation cohort, the one-year risk of re-stroke was 3.0% (95%CI 1.9-4.1) in 932 (29%) patients with a score 0-1, 7.2% (6.1-8.3) in 2038 (63%) with a score 2-4, and 19.2% (14.6-23.9) in 276 (8%) with a score ≥ 5. The score calibrated well in the Athens (recurrences = 208/2495), but not in the Helsinki (recurrences = 15/714) or Milan (recurrences = 65/1279) cohorts. The AUC was 0.67 in the derivation cohort, and 0.56, 0.70, and 0.63 in the Athens, Helsinki, and Milan cohorts, respectively. CONCLUSION: We developed a score to predict one-year stroke recurrence risk in patients with acute ischemic stroke. Since the score was not completely validated when applied to external datasets where it displayed poor to fair calibration and discrimination, additional efforts are required to ameliorate our accuracy for predicting stroke recurrence, by better refining this prognostic tool or developing new ones. Clinical and radiological markers of established cerebrovascular disease and stroke etiology were better predictors than the usual demographic vascular risk factors.

Acute ischaemic stroke in the absence of established vascular risk factors: Patient characteristics, stroke mechanism and long-term outcome.

BACKGROUND AND PURPOSE: Some acute ischaemic stroke (AIS) patients do not display established vascular risk factors (EVRFs). The aim was to assess their clinical characteristics, stroke subtype etiological classification and long-term outcome. METHODS: All consecutive AIS patients from the Acute Stroke Registry of Lausanne (2003-2018) were retrospectively analyzed with complete assessment of the following EVRFs: hypertension, diabetes, major cardioembolic sources, dyslipidemia, smoking, obesity, alcohol abuse, previous stroke/transient ischaemic attack and depression/psychosis. Patients without EVRFs were compared to patients with one or more EVRFs using appropriate statistical models. RESULTS: Of 4889 included patients, 103 (2.1%) had no EVRFs. In multiple regression analysis, patients without EVRFs were significantly younger (odds ratio [OR] 0.13; 95% confidence interval [CI] 0.08-0.20) and had more multiterritorial strokes (OR 3.38; 95% CI 1.26-9.05). Strokes were more often related to patent foramen ovale (PFO) (OR 3.02; 95% CI 1.44-6.32) and less to atherosclerosis, cardioembolism or small vessel disease. In patients <55 years old, PFO (OR 2.76; 95% CI 1.50-5.08) and contraceptive use in females (OR 2.75; 95% CI 1.40-5.41) were more frequent, whereas sleep apnea syndrome (OR 0.09; 95% CI 0.01-0.63) was less. In patients ≥55 years, female sex (OR 2.84; 95% CI 1.43-5.65) and active cancer (OR 3.27; 95% CI 1.34-7.94) were more prevalent. At 12 months, patients without EVRFs had worse adjusted functional outcome (Rankin shift ORadj 0.63; 95% CI 0.42-0.95) and higher rate of recurrence and death (adjusted hazard ratio 2.11; 95% CI 1.19-3.74). CONCLUSIONS: In a consecutive cohort of AIS patients, only 2% showed no EVRFs. PFO and contraceptive use exhibited a strong association with the absence of EVRFs in younger patients and female sex and active cancer in elderly patients. Our findings highlight the importance of searching for previously unknown risk factors and/or unusual stroke mechanisms in patients without EVRFs.

Prior Stroke in PFO Patients Is Associated With Both PFO-Related and -Unrelated Factors.

Background and Purpose: To identify factors associated with prior stroke at presentation in patients with cryptogenic stroke (CS) and patent foramen ovale (PFO). Methods: We studied cross-sectional data from the International PFO Consortium Study (NCT00859885). Patients with first-ever stroke and those with prior stroke at baseline were analyzed for an association with PFO-related (right-to-left shunt at rest, atrial septal aneurysm, deep venous thrombosis, pulmonary embolism, and Valsalva maneuver) and PFO-unrelated factors (age, gender, BMI, hypertension, diabetes mellitus, hypercholesterolemia, smoking, migraine, coronary artery disease, aortic plaque). A multivariable analysis was used to adjust effect estimation for confounding, e.g., owing to the age-dependent definition of study groups in this cross-sectional study design. Results: We identified 635 patients with first-ever and 53 patients with prior stroke. Age, BMI, hypertension, diabetes mellitus, hypercholesterolemia, coronary artery disease, and right-to-left shunt (RLS) at rest were significantly associated with prior stroke. Using a pre-specified multivariable logistic regression model, age (Odds Ratio 1.06), BMI (OR 1.06), hypercholesterolemia (OR 1.90) and RLS at rest (OR 1.88) were strongly associated with prior stroke.Based on these factors, we developed a nomogram to illustrate the strength of the relation of individual factors to prior stroke. Conclusion: In patients with CS and PFO, the likelihood of prior stroke is associated with both, PFO-related and PFO-unrelated factors.

Eligibility for late endovascular treatment using DAWN, DEFUSE-3, and more liberal selection criteria in a stroke center.

BACKGROUND AND PURPOSE: The real-life application of DAWN and DEFUSE-3 trials has been poorly investigated. We aimed to identify the proportion of patients with acute ischemic stroke (AIS) eligible for late endovascular treatment (EVT) in our stroke center based on trial and more liberal selection criteria. METHODS: All consecutive patients in our stroke registry (2003-2017) admitted within 5-23 hours of last proof of good health were selected if they had complete clinical and radiological datasets. We calculated the proportion of patients eligible for late EVT according to trial (DAWN and/or DEFUSE-3) and more liberal clinical/imaging mismatch criteria (including lower admission National Institutes of Health Stroke Scale score and Alberta Stroke Program Early CT Score for core estimation). RESULTS: Of 1705 patients with AIS admitted to our comprehensive stroke center in the late time window, we identified 925 patients with complete clinical and radiological data. Among them, the proportions of late EVT eligibility were 2.5% (n=23) with DAWN, 5.1% (n=47) with DEFUSE-3, and 11.1% (n=103) with more liberal criteria. Considering late-arriving patients with large vessel occlusion (n=221), the percentages of eligible patients were 10.4%, 21.3%, and 46.6%, respectively. A favorable outcome was observed at comparable rates in treated patients selected by trial or liberal criteria (67% vs 58%, p=0.49). CONCLUSIONS: In a long-term stroke registry, the proportion of late EVT eligibility varied greatly according to selection criteria and referral pattern. Among late-arriving patients referred to our comprehensive stroke center, we found 5.6% eligible according to trial (DAWN/DEFUSE-3) and 11.1% according to liberal criteria. These data indicate that late EVT could be offered to a larger population of patients if more liberal criteria are applied.