A Clinical and Imaging Fused Deep Learning Model Matches Expert Clinician Prediction of 90-Day Stroke Outcomes.

BACKGROUND AND PURPOSE: Predicting long-term clinical outcome in acute ischemic stroke is beneficial for prognosis, clinical trial design, resource management, and patient expectations. This study used a deep learning-based predictive model (DLPD) to predict 90-day mRS outcomes and compared its predictions with those made by physicians. MATERIALS AND METHODS: A previously developed DLPD that incorporated DWI and clinical data from the acute period was used to predict 90-day mRS outcomes in 80 consecutive patients with acute ischemic stroke from a single-center registry. We assessed the predictions of the model alongside those of 5 physicians (2 stroke neurologists and 3 neuroradiologists provided with the same imaging and clinical information). The primary analysis was the agreement between the ordinal mRS predictions of the model or physician and the ground truth using the Gwet Agreement Coefficient. We also evaluated the ability to identify unfavorable outcomes (mRS >2) using the area under the curve, sensitivity, and specificity. Noninferiority analyses were undertaken using limits of 0.1 for the Gwet Agreement Coefficient and 0.05 for the area under the curve analysis. The accuracy of prediction was also assessed using the mean absolute error for prediction, percentage of predictions ±1 categories away from the ground truth (±1 accuracy [ACC]), and percentage of exact predictions (ACC). RESULTS: To predict the specific mRS score, the DLPD yielded a Gwet Agreement Coefficient score of 0.79 (95% CI, 0.71-0.86), surpassing the physicians' score of 0.76 (95% CI, 0.67-0.84), and was noninferior to the readers (P < .001). For identifying unfavorable outcome, the model achieved an area under the curve of 0.81 (95% CI, 0.72-0.89), again noninferior to the readers' area under the curve of 0.79 (95% CI, 0.69-0.87) (P < .005). The mean absolute error, ±1ACC, and ACC were 0.89, 81%, and 36% for the DLPD. CONCLUSIONS: A deep learning method using acute clinical and imaging data for long-term functional outcome prediction in patients with acute ischemic stroke, the DLPD, was noninferior to that of clinical readers.

Anatomical predictors of need for decompressive craniectomy after stroke using voxel-based lesion symptom mapping.

BACKGROUND AND PURPOSE: Malignant cerebral edema (MCE) secondary to ischemic stroke is a highly morbid condition. Decompressive craniectomy (DC) is the only treatment for MCE that has been shown to reduce mortality. We examined whether early infarction and/or hypoperfusion in specific topographic regions was predictive of the need for later DC. METHODS: A retrospective database of patients evaluated for large vessel occlusion (LVO) stroke at Stanford between 2010 and 2019 was used. Thirty patients with LVO and baseline perfusion MRI who underwent DC were evaluated. Propensity matching based on age, lesion size, and recanalization status was performed on the remaining cohort. Baseline masks of apparent diffusion coefficient (ADC) + Tmax  >6 seconds lesions were generated using automated perfusion software. Voxel-based lesion symptom maping was used to perform logistic regression at each voxel to generate statistical maps of lesion location associated with DC. Hemispheres were combined to increase statistical power. RESULTS: Sixty patients were analyzed. After adjusting for age, lesion size, and recanalization status as covariates, scattered cortical regions, predominately within the temporal and frontal lobe, were mildly to moderately predictive of the need for DC (z-scores: 2.4-6.74, p < .01). CONCLUSIONS: Scattered temporal and frontal lobe regions on baseline diffusion and perfusion MRI were found to be mildly to moderately predictive of the need for subsequent DC in patients with LVO stroke.

Visual review of acute stroke neuroimaging prior to transfer acceptance increases likelihood of endovascular therapy.

OBJECTIVES: Demand for thrombectomy, and interhospital transfer to comprehensive stroke centers (CSCs), for acute stroke is increasing. There is an urgent need to identify patients most likely to benefit from transfer. We evaluated whether CSC providers' review of neuroimaging prior to transfer acceptance improved patient selection for thrombectomy and correlated with higher rates of treatment. MATERIALS AND METHODS: A retrospective database of all patients transferred to Stanford's CSC for thrombectomy between 2015-2019 was used. Pre-acceptance images, when available for visual review, were reviewed by the CSC stroke team via virtual PACS, RAPID software, or LifeImage platforms. RESULTS: 525 patients met inclusion criteria. 147 (28%) had neuroimaging available for review prior to transfer. Of those who did not recanalize en route, 267 (50.8%) underwent thrombectomy. Patients with imaging available for review prior to acceptance were significantly more likely to receive thrombectomy (68% vs 54%, RR 1.26; p=0.006, 95% CI 1.09-1.48). Patient images that were reviewed via RAPID were CT-based perfusion studies; these were more likely to receive thrombectomy (70% vs 54%, RR 1.30; p=0.01, 1.09-1.56). Patients who received EVT were more likely to have had pre-transfer vessel imaging, regardless of availability for visual review (76% vs 59%, RR 1.44; p<0.001, 1.18-1.76). CONCLUSIONS: Patients with concern for acute stroke transferred for consideration of thrombectomy who had neuroimaging visually reviewed prior to transfer acceptance and did not recanalize by time of arrival were significantly more likely to undergo thrombectomy. Additional prospective studies are needed to confirm our findings.

Association of Endovascular Thrombectomy vs Medical Management With Functional and Safety Outcomes in Patients Treated Beyond 24 Hours of Last Known Well: The SELECT Late Study.

Importance: The role of endovascular thrombectomy is uncertain for patients presenting beyond 24 hours of the time they were last known well. Objective: To evaluate functional and safety outcomes for endovascular thrombectomy (EVT) vs medical management in patients with large-vessel occlusion beyond 24 hours of last known well. Design, Setting, and Participants: This retrospective observational cohort study enrolled patients between July 2012 and December 2021 at 17 centers across the United States, Spain, Australia, and New Zealand. Eligible patients had occlusions in the internal carotid artery or middle cerebral artery (M1 or M2 segment) and were treated with EVT or medical management beyond 24 hours of last known well. Interventions: Endovascular thrombectomy or medical management (control). Main Outcomes and Measures: Primary outcome was functional independence (modified Rankin Scale score 0-2). Mortality and symptomatic intracranial hemorrhage (sICH) were safety outcomes. Propensity score (PS)-weighted multivariable logistic regression analyses were adjusted for prespecified clinical characteristics, perfusion parameters, and/or Alberta Stroke Program Early CT Score (ASPECTS) and were repeated in subsequent 1:1 PS-matched cohorts. Results: Of 301 patients (median [IQR] age, 69 years [59-81]; 149 female), 185 patients (61%) received EVT and 116 (39%) received medical management. In adjusted analyses, EVT was associated with better functional independence (38% vs control, 10%; inverse probability treatment weighting adjusted odds ratio [IPTW aOR], 4.56; 95% CI, 2.28-9.09; P < .001) despite increased odds of sICH (10.1% for EVT vs 1.7% for control; IPTW aOR, 10.65; 95% CI, 2.19-51.69; P = .003). This association persisted after PS-based matching on (1) clinical characteristics and ASPECTS (EVT, 35%, vs control, 19%; aOR, 3.14; 95% CI, 1.02-9.72; P = .047); (2) clinical characteristics and perfusion parameters (EVT, 35%, vs control, 17%; aOR, 4.17; 95% CI, 1.15-15.17; P = .03); and (3) clinical characteristics, ASPECTS, and perfusion parameters (EVT, 45%, vs control, 21%; aOR, 4.39; 95% CI, 1.04-18.53; P = .04). Patients receiving EVT had lower odds of mortality (26%) compared with those in the control group (41%; IPTW aOR, 0.49; 95% CI, 0.27-0.89; P = .02). Conclusions and Relevance: In this study of treatment beyond 24 hours of last known well, EVT was associated with higher odds of functional independence compared with medical management, with consistent results obtained in PS-matched subpopulations and patients with presence of mismatch, despite increased odds of sICH. Our findings support EVT feasibility in selected patients beyond 24 hours. Prospective studies are warranted for confirmation.

High Incidence and Unique Features of Cerebral Venous Sinus Thrombosis in Hospitalized Patients With COVID-19 Infection.

BACKGROUND: Cerebral venous sinus thrombosis (CVST) secondary to vaccine-induced thrombotic thrombocytopenia is an extremely rare side effect of adenovirus-based COVID-19 vaccines. CVST incidence associated with COVID-19 itself has not been widely reported. We report the incidence of CVST in patients hospitalized with COVID-19 during the first year of the pandemic. METHODS: We analyzed de-identified electronic medical records of a retrospective cohort of patients admitted with COVID-19 to >200 hospitals between March 2020 and March 2021. We used International Classification of Diseases, Tenth Revision codes and natural language processing extracts to identify patients with a new CVST diagnosis during COVID-19 hospitalization. The primary outcome was CVST incidence in hospitalized, COVID-19-positive patients. Secondary outcomes included CVST incidence and mortality. Incidence rates were calculated using the DerSimonian-Laird estimator method. RESULTS: Ninety-one thousand seven hundred twenty-seven patients were evaluated; 22 had new CVST diagnoses by electronic medical record review. CVST incidence in the hospitalized COVID-19 cohort was 231 per 1 000 000 person-years (95% CI, 152.1-350.8). Females<50 had the highest incidence overall (males <50: 378.4 [142-1008.2]; females<50: 796.5 [428.6-1480.4]). In patients ≥50 years old, males had a higher estimated CVST incidence (males≥50: 130.5 [54.3-313.6]; females≥50: 88.8 [28.6-275.2]). Older patients (45.5% of patients ≥50 versus 0% of <50 years of age, P=0.012) and males (44.4% of males versus 7.7% of females, P=0.023) were more likely to die in hospital. CONCLUSIONS: CVST incidence in COVID-19-positive hospitalized patients is high. Advanced age and male gender were associated with likelihood of death in hospital; further studies are required to confirm these findings.

Perfusion Imaging and Clinical Outcome in Acute Minor Stroke With Large Vessel Occlusion.

BACKGROUND: Whether bridging therapy (intravenous thrombolysis [IVT] followed by mechanical thrombectomy) is superior to IVT alone in minor stroke with large vessel occlusion is unknown. Perfusion imaging may identify subsets of large vessel occlusion-related minor stroke patients with distinct response to bridging therapy. METHODS: We conducted a multicenter international observational study of consecutive IVT-treated patients with minor stroke (National Institutes of Health Stroke Scale score ≤5) who had an anterior circulation large vessel occlusion and perfusion imaging performed before IVT, with a subset undergoing immediate thrombectomy. Propensity score with inverse probability of treatment weighting was used to account for baseline between-groups differences. The primary outcome was 3-month modified Rankin Scale score 0 to 1. We searched for an interaction between treatment group and mismatch volume (critical hypoperfusion-core volume). RESULTS: Overall, 569 patients were included (172 and 397 in the bridging therapy and IVT groups, respectively). After propensity-score weighting, the distribution of baseline variables was similar across the 2 groups. In the entire population, bridging was associated with lower odds of achieving modified Rankin Scale score 0 to 1: odds ratio, 0.73 [95% CI, 0.55-0.96]; P=0.03. However, mismatch volume modified the effect of bridging on clinical outcome (Pinteraction=0.04 for continuous mismatch volume); bridging was associated with worse outcome in patients with, but not in those without, mismatch volume <40 mL (odds ratio, [95% CI] for modified Rankin Scale score 0-1: 0.48 [0.33-0.71] versus 1.14 [0.76-1.71], respectively). Bridging was associated with higher incidence of symptomatic intracranial hemorrhage in the entire population, but this effect was present in the small mismatch subset only (Pinteraction=0.002). CONCLUSIONS: In our population of large vessel occlusion-related minor stroke patients, bridging therapy was associated with lower rates of good outcome as compared with IVT alone. However, mismatch volume was a strong modifier of the effect of bridging therapy over IVT alone, notably with worse outcome with bridging therapy in patients with mismatch volume ≤40 mL. Randomized trials should consider adding perfusion imaging for patient selection.

Thrombectomy versus Medical Management in Mild Strokes due to Large Vessel Occlusion: Exploratory Analysis from the EXTEND-IA Trials and a Pooled International Cohort.

OBJECTIVE: This study was undertaken to evaluate functional and safety outcomes for endovascular thrombectomy (EVT) versus medical management (MM) in patients with large vessel occlusion (LVO) and mild neurological deficits, stratified by perfusion imaging mismatch. METHODS: The pooled cohort consisted of patients with National Institutes of Health Stroke Scale (NIHSS) < 6 and internal carotid artery (ICA), M1, or M2 occlusions from the Extending the Time for Thrombolysis in Emergecy Neurological Deficits - Intra-Arterial (EXTEND-IA) Trial,  Tenecteplase vs Alteplase before Endovascular Thrombectomy in Ischemic Stroke (EXTEND-IA TNK) trials Part I/II and prospective data from 15 EVT centers from October 2010 to April 2020. RAPID software estimated ischemic core and mismatch. Patients receiving primary EVT (EVTpri ) were compared to those who received primary MM (MMpri ), including those who deteriorated and received rescue EVT, in overall and propensity score (PS)-matched cohorts. Patients were stratified by target mismatch (mismatch ratio ≥ 1.8 and mismatch volume ≥ 15ml). Primary outcome was functional independence (90-day modified Rankin Scale = 0-2). Secondary outcomes included safety (symptomatic intracerebral hemorrhage [sICH], neurological worsening, and mortality). RESULTS: Of 540 patients, 286 (53%) received EVTpri and demonstrated larger critically hypoperfused tissue (Tmax > 6 seconds) volumes (median [IQR]: 64 [26-96] ml vs MMpri : 40 [14-76] ml, p < 0.001) and higher presentation NIHSS (median [IQR]: 4 [2-5] vs MMpri : 3 [2-4], p < 0.001). Functional independence was similar (EVTpri : 77.4% vs MMpri : 75.6%, adjusted odds ratio [aOR] = 1.29, 95% confidence interval [CI] = 0.82-2.03, p = 0.27). EVT had worse safety regarding sICH (EVTpri : 16.3% vs MMpri : 1.3%, p < 0.001) and neurological worsening (EVTpri : 19.6% vs MMpri : 6.7%, p < 0.001). In 414 subjects (76.7%) with target mismatch, EVT was associated with improved functional independence (EVTpri : 77.4% vs MMpri : 72.7%, aOR = 1.68, 95% CI = 1.01-2.81, p = 0.048), whereas there was a trend toward less favorable outcomes with primary EVT (EVTpri : 77.4% vs MMpri : 83.3%, aOR = 0.39, 95% CI = 0.12-1.34, p = 0.13) without target mismatch (pinteraction  = 0.06). Similar findings were observed in a propensity score-matched subpopulation. INTERPRETATION: Overall, EVT was not associated with improved clinical outcomes in mild strokes due to LVO, and sICH was increased. However, in patients with target mismatch profile, EVT was associated with increased functional independence. Perfusion imaging may be helpful to select mild stroke patients for EVT. ANN NEUROL 2022;92:364-378.

The bright vessel sign on arterial spin labeling MRI for heralding and localizing large vessel occlusions.

BACKGROUND AND PURPOSE: The significance of a bright vessel sign (BVS) at the site of a large vessel occlusion (LVO) on MR arterial spin labeling (ASL) sequence is not widely reported. We compared the utility of the ASL BVS to the gradient echo (GRE) susceptibility vessel sign (SVS) in heralding and localizing LVOs in a large cohort; most underwent digital subtraction angiography (DSA) and endovascular therapy for acute stroke. METHODS: A total of 171 patients with large hemispheric stroke symptoms had baseline and follow-up MRIs with ASL, GRE, and MR angiogram (MRA). Scans were evaluated for (1) presence versus absence and (2) location of ASL BVS and GRE SVS. For patients who underwent DSA, data comparing presence and location of ASL BVS and GRE SVS to occlusions found on angiography, as well as resolution of the signs after successful recanalization, were also evaluated. RESULTS: Compared to MRA, the sensitivity of the ASL BVS for an LVO was .83, significantly better than .67 for GRE SVS (p = .001). Localization of vessel occlusion was correct 60.4% of the time by ASL compared to 64.4% by GRE (p = .502). For the 107 patients who underwent DSA, the sensitivity of ASL BVS was .80 compared to .64 for GRE SVS (p = .009). Localization of LVO found on DSA was correct 63.5% of the time by ASL BVS compared to 72.9% by GRE SVS (p = .251). CONCLUSION: ASL BVS is significantly more sensitive than GRE SVS for identification of LVO on both MRA and DSA.

Improving patient selection for endovascular treatment of acute cerebral ischemia: a review of the literature and an external validation of the Houston IAT and THRIVE predictive scoring systems.

Outcome after intraarterial therapy (IAT) for acute ischemic stroke remains variable, suggesting that improved patient selection is needed to better identify patients likely to benefit from treatment. The authors evaluate the predictive accuracies of the Houston IAT (HIAT) and the Totaled Health Risks in Vascular Events (THRIVE) scores in an independent cohort and review the existing literature detailing additional predictive factors to be used in patient selection for IAT. They reviewed their center's endovascular records from January 2004 to July 2010 and identified patients who had acute ischemic stroke and underwent IAT. They calculated individual HIAT and THRIVE scores using patient age, admission National Institutes of Health Stroke Scale (NIHSS) score, admission glucose level, and medical history. The scores' predictive accuracies for good outcome (discharge modified Rankin Scale score ≤ 3) were analyzed using receiver operating characteristics analysis. The THRIVE score predicts poor outcome after IAT with reasonable accuracy and may perform better than the HIAT score. Nevertheless, both measures may have significant clinical utility; further validation in larger cohorts that accounts for differences in patient demographic characteristics, variation in time-to-treatment, and center preferences with respect to IAT modalities is needed. Additional patient predictive factors have been reported but not yet incorporated into predictive scales; the authors suggest the need for additional data analysis to determine the independent predictive value of patient admission NIHSS score, age, admission hyperglycemia, patient comorbidities, thrombus burden, collateral flow, time to treatment, and baseline neuroimaging findings.

The role of polyamine metabolism in neuronal injury following cerebral ischemia.

Stroke is a leading cause of morbidity and mortality in the US, with secondary damage following the initial insult contributing significantly to overall poor outcome. Prior investigations have shown that the metabolism of certain polyamines such as spermine, spermidine, and putrescine are elevated in ischemic parenchyma, resulting in an increase in their metabolite concentration. Polyamine metabolites tend to be cytotoxic, leading to neuronal injury in the penumbra following stroke and expansion of the area of infarcted tissue. Although the precise mechanism is unclear, the presence of reactive aldehydes produced through polyamine metabolism, such as 3-aminopropanal and acrolein, have been shown to correlate with the incidence of cerebral vasospasm, disruption of oxidative metabolism and mitochondrial functioning, and disturbance of cellular calcium ion channels. Regulation of the polyamine metabolic pathway, therefore, may have the potential to limit injury following cerebral ischemia. To this end, we review this pathway in detail with an emphasis on clinical applicability.