Elizabeth Pulcine
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Elizabeth Pulcine has a total of 89 co-authors. Here are the top co-authors by number of shared publications:
- Nomazulu Dlamini
Department of Paediatrics, The Hospital for Sick Children, University of Toronto, Canada
5 shared publications - Prakash Muthusami
The Hospital for Sick Children, and University of Toronto
4 shared publications - Adam Andrew Dmytriw
Nuffield Department of Surgical Sciences, Medical Sciences Division, University of Oxford, Oxford, United Kingdom
3 shared publications - Kartik Dev Bhatia
Department of Medical Imaging, Children’s Hospital at Westmead, New South Wales, Australia (K.D.B.).
3 shared publications - Carmen Parra-Farinas
Division of Neuroradiology and Division of Neurointervention, Hospital for Sick Children, University of Toronto, ON, Canada
2 shared publications - Christine K. Fox
Department of Neurology and Weill Institute for Neuroscience University of California San Francisco San Francisco California USA
2 shared publications
Latest Publications
- 14 July 2026
- Published by Wolters Kluwer Health in Neurology
- Vol. 107 (1) , e218149
Thrombectomy for Pediatric Large Vessel Occlusion Stroke With Mild Presenting Symptoms
Background and ObjectivesNeurologic outcomes after pediatric large vessel occlusion (LVO) stroke are poor. In the absence of a pediatric randomized clinical trial, cohort and registry studies have ...
- 14 July 2026
- Published by Wolters Kluwer Health in Neurology
- Vol. 107 (1) , e218149
Thrombectomy for Pediatric Large Vessel Occlusion Stroke With Mild Presenting Symptoms
Background and ObjectivesNeurologic outcomes after pediatric large vessel occlusion (LVO) stroke are poor. In the absence of a pediatric randomized clinical trial, cohort and registry studies have ...
- Open Access
- 25 May 2026
- Published by Wolters Kluwer Health in Critical Care Explorations
- Vol. 8 (6) , e1418
Cerebral Venous Sinus Thrombosis in Pediatric Critical Care
IMPORTANCE: Pediatric cerebral venous sinus thrombosis (CVST) is being increasingly recognized and can pose substantial risks of morbidity and mortality. Data on the epidemiology, management, and outcomes of CVST in the PICU remain limited. OBJECTIVES: To describe the clinical characteristics, management, and outcomes of critically ill children with CVST during their admission to the PICU. DESIGN, SETTING, AND PARTICIPANTS: We conducted a retrospective observational cohort study in a quaternary PICU in Toronto, Canada, between 2018 and 2023. Patients 18 years old and younger with acute primary CVST (CVST being the primary indication for ICU admission) and secondary CVST (diagnosis during an admission for an alternative diagnosis) were included in this study. MAIN OUTCOMES AND MEASURES: The primary outcome was in-hospital mortality. Descriptive statistics were used to describe characteristics and outcomes. RESULTS: Thirty patients were admitted with a diagnosis of CVST: 19 (63%) primary, 11 (37%) secondary. Fourteen (47%) had an associated cerebral infarct, and nine (30%) had an associated intracranial hemorrhage. The most common condition associated with secondary CVST was a brain disease requiring neurosurgical intervention (5/11). Five (17%) children with CVST died in this study, of which four had a primary CVST. Children residing in neighborhoods with increased marginalization were disproportionally represented in this cohort. CONCLUSIONS AND RELEVANCE: Primary CVST is more common than secondary and is associated with significant mortality. The disproportionate impact on marginalized children emphasizes the need for heightened awareness and determination of factors associated with this finding.
- 25 May 2026
- Published by Wolters Kluwer Health in Critical Care Explorations
- Vol. 8 (6) , e1418
Open AccessCerebral Venous Sinus Thrombosis in Pediatric Critical Care
IMPORTANCE: Pediatric cerebral venous sinus thrombosis (CVST) is being increasingly recognized and can pose substantial risks of morbidity and mortality. Data on the epidemiology, management, and outcomes of CVST in the PICU remain limited. OBJECTIVES: To describe the clinical characteristics, management, and outcomes of critically ill children with CVST during their admission to the PICU. DESIGN, SETTING, AND PARTICIPANTS: We conducted a retrospective observational cohort study in a quaternary PICU in Toronto, Canada, between 2018 and 2023. Patients 18 years old and younger with acute primary CVST (CVST being the primary indication for ICU admission) and secondary CVST (diagnosis during an admission for an alternative diagnosis) were included in this study. MAIN OUTCOMES AND MEASURES: The primary outcome was in-hospital mortality. Descriptive statistics were used to describe characteristics and outcomes. RESULTS: Thirty patients were admitted with a diagnosis of CVST: 19 (63%) primary, 11 (37%) secondary. Fourteen (47%) had an associated cerebral infarct, and nine (30%) had an associated intracranial hemorrhage. The most common condition associated with secondary CVST was a brain disease requiring neurosurgical intervention (5/11). Five (17%) children with CVST died in this study, of which four had a primary CVST. Children residing in neighborhoods with increased marginalization were disproportionally represented in this cohort. CONCLUSIONS AND RELEVANCE: Primary CVST is more common than secondary and is associated with significant mortality. The disproportionate impact on marginalized children emphasizes the need for heightened awareness and determination of factors associated with this finding.
- Open Access
NeoCVST Score: Classification of Brain Injury Secondary to Neonatal Cerebral Venous Sinus Thrombosis
Background: Neonatal cerebral venous sinus thrombosis (CVST) is associated with intracranial hemorrhage (ICH) and ischemic lesions. There is no scale to characterize the spectrum of brain injury secondary to neonatal CVST. Objective: To develop the Neonatal CVST Hemorrhage Score (NeoCVST Score) to characterize ICH and brain injury in neonates with CVST. Methods: This was a retrospective cohort study of neonates with CVST diagnosed using brain MRI/MRV. The NeoCVST Score was developed using the study cohort, integrating elements from previous hemorrhage classification systems and expert consensus. Logistic regression examined associations between NeoCVST score and neurodevelopmental outcomes (Pediatric Stroke Outcome Measure). Interrater reliability was assessed with intraclass correlation coefficient. Results: The study included 100 neonates (77% term and 23% preterm) with CVST. Thrombosis of multiple venous sinuses was present in 62%. ICH was present in 63%. Supratentorial hemorrhage was present in 57% and included germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) grades 1-2 (22%), GMH-IVH grade 3 (15%), parenchymal (43%) and thalamic (18%) hemorrhage. Infratentorial hemorrhage was present in 19% and included cerebellar (18%) and brainstem (4%) hemorrhage. Extra-axial hemorrhage was present in 32% and included epidural (2%), subdural (26%) and subarachnoid hemorrhage (6%). Ischemic brain injury was present in 67% and included lesions in the medullary vein distribution (13%), white matter (54%), basal ganglia (17%) and thalamus (25%). Neurodevelopmental outcomes included 40% with normal outcomes and 60% with neurodevelopmental impairments. NeoCVST total score (OR=1.1, P=0.02) and subscores for thalamic hemorrhage (OR=1.9, P=0.04), thalamic ischemia (OR=2.2, P=0.005) and bilateral thalamic ischemia (OR=2.8, P=0.01) were predictors of adverse neurodevelopmental outcome. Inter-rater reliability showed moderate-good agreement between reviewers with an intraclass correlation coefficient of 0.71. Conclusions: The NeoCVST Score is a simple clinical tool to characterize ICH and brain injury secondary to neonatal CVST. Increasing NeoCVST total score and subscores for thalamic hemorrhage and ischemia were associated with worse neurodevelopmental outcomes.
Open AccessNeoCVST Score: Classification of Brain Injury Secondary to Neonatal Cerebral Venous Sinus Thrombosis
Background: Neonatal cerebral venous sinus thrombosis (CVST) is associated with intracranial hemorrhage (ICH) and ischemic lesions. There is no scale to characterize the spectrum of brain injury secondary to neonatal CVST. Objective: To develop the Neonatal CVST Hemorrhage Score (NeoCVST Score) to characterize ICH and brain injury in neonates with CVST. Methods: This was a retrospective cohort study of neonates with CVST diagnosed using brain MRI/MRV. The NeoCVST Score was developed using the study cohort, integrating elements from previous hemorrhage classification systems and expert consensus. Logistic regression examined associations between NeoCVST score and neurodevelopmental outcomes (Pediatric Stroke Outcome Measure). Interrater reliability was assessed with intraclass correlation coefficient. Results: The study included 100 neonates (77% term and 23% preterm) with CVST. Thrombosis of multiple venous sinuses was present in 62%. ICH was present in 63%. Supratentorial hemorrhage was present in 57% and included germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) grades 1-2 (22%), GMH-IVH grade 3 (15%), parenchymal (43%) and thalamic (18%) hemorrhage. Infratentorial hemorrhage was present in 19% and included cerebellar (18%) and brainstem (4%) hemorrhage. Extra-axial hemorrhage was present in 32% and included epidural (2%), subdural (26%) and subarachnoid hemorrhage (6%). Ischemic brain injury was present in 67% and included lesions in the medullary vein distribution (13%), white matter (54%), basal ganglia (17%) and thalamus (25%). Neurodevelopmental outcomes included 40% with normal outcomes and 60% with neurodevelopmental impairments. NeoCVST total score (OR=1.1, P=0.02) and subscores for thalamic hemorrhage (OR=1.9, P=0.04), thalamic ischemia (OR=2.2, P=0.005) and bilateral thalamic ischemia (OR=2.8, P=0.01) were predictors of adverse neurodevelopmental outcome. Inter-rater reliability showed moderate-good agreement between reviewers with an intraclass correlation coefficient of 0.71. Conclusions: The NeoCVST Score is a simple clinical tool to characterize ICH and brain injury secondary to neonatal CVST. Increasing NeoCVST total score and subscores for thalamic hemorrhage and ischemia were associated with worse neurodevelopmental outcomes.
- Open Access
Feasibility of Non-Sedate Magnetic Resonance Imaging for Children with Cerebral Palsy: Tolerance and Structural Analysis Considerations
Background/Objectives: Non-sedate magnetic resonance imaging (MRI) can be challenging for young children with neuromotor disabilities, often resulting in motion-degraded images that complicate interpretation in the context of underlying neuropathology. This study aimed to characterize tolerance factors and barriers related to awake MRI of the pediatric brain and to examine additional considerations in analyzing structural scans affected by motion and pathology. Methods: 10 children (mean age 5y9m; 5 girls; GMFCS level IV) with cerebral palsy (CP) underwent non-sedate 3T MRI of the brain. Tolerance factors and challenges were documented. MRI quality and automated structural preprocessing with Freesurfer (FS) v.8.0 were reviewed by a pediatric neuroradiologist and neurologist. To assess the impact of motion, automated basal ganglia segmentation was compared with manual segmentation. Segmentation accuracy was characterized using Dice Coefficient (D). Results: Five participants (50%) tolerated non-sedate structural MRI, although two of them were unable to remain still. Factors affecting MRI tolerance included sensitivity to scanner noise (n = 4), hyperkinetic movement (n = 2), difficulty with positioning/padding (n = 4), fear of clinical environment (n = 2) or confined scanner interior (n = 2), and earbud discomfort (n = 3). Automated structural preprocessing with FS yielded discrepancies in gray-white matter boundaries in motion-degraded scans, necessitating manual correction. Automated segmentation of motion-compromised scans closely agreed with manual delineation of the caudate (D ≥ 0.85) and putamen (D ≥ 0.78), while the pallidum was least reproducible (D = 0.58). Conclusions: Tailored acquisition and processing strategies are necessary to support non-sedate MRI in children with CP, preserve downstream neuroimaging analyses, and promote inclusion of underrepresented populations in research.
Open AccessFeasibility of Non-Sedate Magnetic Resonance Imaging for Children with Cerebral Palsy: Tolerance and Structural Analysis Considerations
Background/Objectives: Non-sedate magnetic resonance imaging (MRI) can be challenging for young children with neuromotor disabilities, often resulting in motion-degraded images that complicate interpretation in the context of underlying neuropathology. This study aimed to characterize tolerance factors and barriers related to awake MRI of the pediatric brain and to examine additional considerations in analyzing structural scans affected by motion and pathology. Methods: 10 children (mean age 5y9m; 5 girls; GMFCS level IV) with cerebral palsy (CP) underwent non-sedate 3T MRI of the brain. Tolerance factors and challenges were documented. MRI quality and automated structural preprocessing with Freesurfer (FS) v.8.0 were reviewed by a pediatric neuroradiologist and neurologist. To assess the impact of motion, automated basal ganglia segmentation was compared with manual segmentation. Segmentation accuracy was characterized using Dice Coefficient (D). Results: Five participants (50%) tolerated non-sedate structural MRI, although two of them were unable to remain still. Factors affecting MRI tolerance included sensitivity to scanner noise (n = 4), hyperkinetic movement (n = 2), difficulty with positioning/padding (n = 4), fear of clinical environment (n = 2) or confined scanner interior (n = 2), and earbud discomfort (n = 3). Automated structural preprocessing with FS yielded discrepancies in gray-white matter boundaries in motion-degraded scans, necessitating manual correction. Automated segmentation of motion-compromised scans closely agreed with manual delineation of the caudate (D ≥ 0.85) and putamen (D ≥ 0.78), while the pallidum was least reproducible (D = 0.58). Conclusions: Tailored acquisition and processing strategies are necessary to support non-sedate MRI in children with CP, preserve downstream neuroimaging analyses, and promote inclusion of underrepresented populations in research.
- 1 September 2025
- Published by Wolters Kluwer Health in Stroke
- Vol. 56 (9) , 2440-2449
Thrombectomy Versus Medical Management for Pediatric Arterial Ischemic Stroke With Large Baseline Infarct
BACKGROUND: Multiple recent randomized trials have demonstrated the benefit of thrombectomy over medical management alone in adult patients with large vessel occlusion (LVO) stroke and large infarct on baseline imaging. Cohort studies have also identified improved functional outcomes in pediatric patients who received thrombectomy. However, the role of thrombectomy in pediatric stroke with large baseline infarct remains uncertain. METHODS: A case-control study was undertaken using pooled data from 3 cohort studies of pediatric stroke (Save ChildS [Safety and Outcome of Endovascular Recenalization in Childhood Stroke Study], Save ChildS Pro [Safety and Outcome of Endovascular Recenalization in Childhood Stroke Prospective Registry], Pediatric LVO Stroke Study). Pediatric patients of age 1 to 18 years with acute anterior circulation LVO stroke presenting within 24 hours since last seen well with an Alberta Stroke Program Early CT Score of 0 to 5 on CT or magnetic resonance imaging were included. Isolated M2 occlusion or focal cerebral arteriopathy–inflammatory subtype cases were excluded. Thrombectomy-treated patients were compared with patients who received medical management alone. The primary clinical outcome was the pediatric modified Rankin Scale score at 90 days, compared between groups using ordinal logistic regression. RESULTS: Of 56 pediatric patients with anterior circulation LVO and low Alberta Stroke Program Early CT Score presenting between January 1, 2000 and August 31, 2023 from 45 centers across Europe, North America, and Australia, 40 patients were eligible for inclusion (female: n=14, 35.0%; mean age, 9.1 years; range, 1.5–17; SD, 5.27). Thrombectomy-treated patients (n=24) had significantly better pediatric modified Rankin Scale scores at 90 days than medical management alone patients (n=16; odds ratio, 3.68 [95% CI, 1.11–12.21]; P =0.034). There was no significant difference between groups in the rate of symptomatic intracranial hemorrhage ( P =0.806). CONCLUSIONS: In this multicentre case-control study, pediatric patients (age, 1–18 years) with anterior circulation LVO stroke and low Alberta Stroke Program Early CT Score who received thrombectomy had significantly better functional outcomes than those treated with medical management alone. Exclusion of pediatric patients from thrombectomy based on low Alberta Stroke Program Early CT Score alone may not be appropriate.
- 1 September 2025
- Published by Wolters Kluwer Health in Stroke
- Vol. 56 (9) , 2440-2449
Thrombectomy Versus Medical Management for Pediatric Arterial Ischemic Stroke With Large Baseline Infarct
BACKGROUND: Multiple recent randomized trials have demonstrated the benefit of thrombectomy over medical management alone in adult patients with large vessel occlusion (LVO) stroke and large infarct on baseline imaging. Cohort studies have also identified improved functional outcomes in pediatric patients who received thrombectomy. However, the role of thrombectomy in pediatric stroke with large baseline infarct remains uncertain. METHODS: A case-control study was undertaken using pooled data from 3 cohort studies of pediatric stroke (Save ChildS [Safety and Outcome of Endovascular Recenalization in Childhood Stroke Study], Save ChildS Pro [Safety and Outcome of Endovascular Recenalization in Childhood Stroke Prospective Registry], Pediatric LVO Stroke Study). Pediatric patients of age 1 to 18 years with acute anterior circulation LVO stroke presenting within 24 hours since last seen well with an Alberta Stroke Program Early CT Score of 0 to 5 on CT or magnetic resonance imaging were included. Isolated M2 occlusion or focal cerebral arteriopathy–inflammatory subtype cases were excluded. Thrombectomy-treated patients were compared with patients who received medical management alone. The primary clinical outcome was the pediatric modified Rankin Scale score at 90 days, compared between groups using ordinal logistic regression. RESULTS: Of 56 pediatric patients with anterior circulation LVO and low Alberta Stroke Program Early CT Score presenting between January 1, 2000 and August 31, 2023 from 45 centers across Europe, North America, and Australia, 40 patients were eligible for inclusion (female: n=14, 35.0%; mean age, 9.1 years; range, 1.5–17; SD, 5.27). Thrombectomy-treated patients (n=24) had significantly better pediatric modified Rankin Scale scores at 90 days than medical management alone patients (n=16; odds ratio, 3.68 [95% CI, 1.11–12.21]; P =0.034). There was no significant difference between groups in the rate of symptomatic intracranial hemorrhage ( P =0.806). CONCLUSIONS: In this multicentre case-control study, pediatric patients (age, 1–18 years) with anterior circulation LVO stroke and low Alberta Stroke Program Early CT Score who received thrombectomy had significantly better functional outcomes than those treated with medical management alone. Exclusion of pediatric patients from thrombectomy based on low Alberta Stroke Program Early CT Score alone may not be appropriate.
- 14 January 2025
- Published by Wolters Kluwer Health in Neurology
- Vol. 104 (1) , e210203
Teaching NeuroImage: Stroke-Like Migraine Attacks After Radiation Therapy Syndrome
- 14 January 2025
- Published by Wolters Kluwer Health in Neurology
- Vol. 104 (1) , e210203
Teaching NeuroImage: Stroke-Like Migraine Attacks After Radiation Therapy Syndrome
- Open Access
- 28 March 2024
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 45 (5) , 588-591
Mechanical Thrombectomy for Pediatric Arterial Ischemic Stroke from Acute M2 Occlusion
SUMMARY: Pediatric large-vessel occlusion has a poor natural history. Recent retrospective studies have demonstrated the potential benefits, feasibility, and safety profile of mechanical thrombectomy in children. However, the role of thrombectomy in pediatric M2 occlusions remains uncertain. In this clinical report, we present a multicenter series of 6 pediatric patients with acute M2 occlusion (female = 1, male = 5; age range, 0.9–16.0 years, mean = 9.2). All 6 patients having undergone thrombectomy had excellent clinical outcomes (pediatric mRS = 0–1) at 3 months and final available follow-up (median, 12 months; range, 3–72 months). Factors relevant to treatment decision-making in pediatric M2 occlusions are discussed, including the important role of multidisciplinary team discussions during acute management.
- 28 March 2024
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 45 (5) , 588-591
Open AccessMechanical Thrombectomy for Pediatric Arterial Ischemic Stroke from Acute M2 Occlusion
SUMMARY: Pediatric large-vessel occlusion has a poor natural history. Recent retrospective studies have demonstrated the potential benefits, feasibility, and safety profile of mechanical thrombectomy in children. However, the role of thrombectomy in pediatric M2 occlusions remains uncertain. In this clinical report, we present a multicenter series of 6 pediatric patients with acute M2 occlusion (female = 1, male = 5; age range, 0.9–16.0 years, mean = 9.2). All 6 patients having undergone thrombectomy had excellent clinical outcomes (pediatric mRS = 0–1) at 3 months and final available follow-up (median, 12 months; range, 3–72 months). Factors relevant to treatment decision-making in pediatric M2 occlusions are discussed, including the important role of multidisciplinary team discussions during acute management.
- 1 July 2022
- Published by Wolters Kluwer Health in Stroke
- Vol. 53 (7) , 2241-2248
Cerebral Venous Sinus Thrombosis in Preterm Infants
Background: Neonatal cerebral venous sinus thrombosis (CVST) can lead to brain injury and neurodevelopmental impairments. Previous studies of neonatal CVST have focused on term infants, and studies of preterm infants are lacking. In this study, we examined the clinical and radiological features, treatment and outcome of CVST in preterm infants. Methods: This was a retrospective, consecutive cohort study of preterm infants (gestational age <37 weeks) with radiologically confirmed CVST. All magnetic resonance imaging/MRV and CT/CTV scans were re-reviewed to study thrombus characteristics and pattern of brain injury. Outcome was assessed by the validated pediatric stroke outcome measure at the most recent clinic visit. Results: Twenty-six preterm infants with CVST were studied. Of these, 65% were moderate-late preterm (32–37 weeks), 27% very preterm (28–32 weeks), and 8% extreme preterm (<28 weeks). Most (73%) were symptomatic at presentation with seizures or abnormal exam. Transverse (85%) and superior sagittal (42%) sinuses were common sites of thrombosis. Parenchymal brain injury was predominantly periventricular (35%) and deep white matter (31%) in location. Intraventricular hemorrhage occurred in 46%. Most infants (69%) were treated with anticoagulation. No treated infant (including eleven with pretreatment hemorrhage) had new or worsening post-treatment hemorrhage. Outcomes ranged from no deficits (50%), mild-moderate (25%), and severe (25%) impairment. Conclusions: In our sample of preterm infants with CVST, more than one-quarter were asymptomatic. White matter brain lesions predominated and one-half had neurological deficits at follow-up. Anticoagulation of preterm CVST in this small cohort appeared to be safe. Larger studies of preterm CVST are needed.
- 1 July 2022
- Published by Wolters Kluwer Health in Stroke
- Vol. 53 (7) , 2241-2248
Cerebral Venous Sinus Thrombosis in Preterm Infants
Background: Neonatal cerebral venous sinus thrombosis (CVST) can lead to brain injury and neurodevelopmental impairments. Previous studies of neonatal CVST have focused on term infants, and studies of preterm infants are lacking. In this study, we examined the clinical and radiological features, treatment and outcome of CVST in preterm infants. Methods: This was a retrospective, consecutive cohort study of preterm infants (gestational age <37 weeks) with radiologically confirmed CVST. All magnetic resonance imaging/MRV and CT/CTV scans were re-reviewed to study thrombus characteristics and pattern of brain injury. Outcome was assessed by the validated pediatric stroke outcome measure at the most recent clinic visit. Results: Twenty-six preterm infants with CVST were studied. Of these, 65% were moderate-late preterm (32–37 weeks), 27% very preterm (28–32 weeks), and 8% extreme preterm (<28 weeks). Most (73%) were symptomatic at presentation with seizures or abnormal exam. Transverse (85%) and superior sagittal (42%) sinuses were common sites of thrombosis. Parenchymal brain injury was predominantly periventricular (35%) and deep white matter (31%) in location. Intraventricular hemorrhage occurred in 46%. Most infants (69%) were treated with anticoagulation. No treated infant (including eleven with pretreatment hemorrhage) had new or worsening post-treatment hemorrhage. Outcomes ranged from no deficits (50%), mild-moderate (25%), and severe (25%) impairment. Conclusions: In our sample of preterm infants with CVST, more than one-quarter were asymptomatic. White matter brain lesions predominated and one-half had neurological deficits at follow-up. Anticoagulation of preterm CVST in this small cohort appeared to be safe. Larger studies of preterm CVST are needed.
- 21 June 2021
- Published by Elsevier in Journal of Thrombosis and Haemostasis
- Vol. 19 (10) , 2428-2439
Hemorrhagic transformation and stroke recurrence in children with cardiac disease receiving antithrombotic therapy for secondary stroke prevention
- 21 June 2021
- Published by Elsevier in Journal of Thrombosis and Haemostasis
- Vol. 19 (10) , 2428-2439
Hemorrhagic transformation and stroke recurrence in children with cardiac disease receiving antithrombotic therapy for secondary stroke prevention