Journal of Stroke最新文献

Background and purpose: Accurate classification of ischemic stroke subtype is important for effective secondary prevention of stroke. We used diffusion-weighted image (DWI) and atrial fibrillation (AF) data to train a deep learning algorithm to classify stroke subtype.

Methods: Model development was done in 2,988 patients with ischemic stroke from three centers by using U-net for infarct segmentation and EfficientNetV2 for subtype classification. Experienced neurologists (n=5) determined subtypes for external test datasets, while establishing a consensus for clinical trial datasets. Automatically segmented infarcts were fed into the model (DWI-only algorithm). Subsequently, another model was trained, with AF included as a categorical variable (DWI+AF algorithm). These models were tested: (1) internally against the opinion of the labeling experts, (2) against fresh external DWI data, and (3) against clinical trial dataset.

Results: In the training-and-validation datasets, the mean (±standard deviation) age was 68.0±12.5 (61.1% male). In internal testing, compared with the experts, the DWI-only and the DWI+AF algorithms respectively achieved moderate (65.3%) and near-strong (79.1%) agreement. In external testing, both algorithms again showed good agreements (59.3%-60.7% and 73.7%-74.0%, respectively). In the clinical trial dataset, compared with the expert consensus, percentage agreements and Cohen's kappa were respectively 58.1% and 0.34 for the DWI-only vs. 72.9% and 0.57 for the DWI+AF algorithms. The corresponding values between experts were comparable (76.0% and 0.61) to the DWI+AF algorithm.

Conclusion: Our model trained on a large dataset of DWI (both with or without AF information) was able to classify ischemic stroke subtypes comparable to a consensus of stroke experts.

Distal medium vessel occlusions (DMVOs) are thought to cause as many as 25% to 40% of all acute ischemic strokes and may result in substantial disability amongst survivors. Although intravenous thrombolysis (IVT) is more effective for distal than proximal vessel occlusions, the overall efficacy of IVT remains limited in DMVO with less than 50% of patients achieving reperfusion and about 1/3 to 1/4 of the patients failing to achieve functional independence. Data regarding mechanical thrombectomy (MT) among these patients remains limited. The smaller, thinner, and more tortuous vessels involved in DMVO are presumably associated with higher procedural risks whereas a lower benefit might be expected given the smaller amount of tissue territory at risk. Recent advances in technology have shown promising results in endovascular treatment of DMVOs with room for future improvement. In this review, we discuss some of the key technical and clinical considerations in DMVO treatment including the anatomical and clinical terminology, diagnostic modalities, the role of IVT and MT, existing technology, and technical challenges as well as the contemporary evidence and future treatment directions.

Background and purpose: Infarct volume and other imaging markers are increasingly used as surrogate measures for clinical outcome in acute ischemic stroke research, but how improvements in these imaging surrogates translate into better clinical outcomes is currently unclear. We investigated how changes in infarct volume at 24 hours alter the probability of achieving good clinical outcome (modified Rankin Scale [mRS] 0-2).

Methods: Data are from endovascular thrombectomy patients from the randomized controlled ESCAPE-NA1 (Efficacy and Safety of Nerinetide for the Treatment of Acute Ischaemic Stroke) trial. Infarct volume at 24 hours was manually segmented on non-contrast computed tomography or diffusion-weighted magnetic resonance imaging. Probabilities of achieving good outcome based on infarct volume were obtained from a multivariable logistic regression model. The probability of good outcome was plotted against infarct volume using linear spline regression.

Results: A total of 1,099 patients were included in the analysis (median final infarct volume 24.9 mL [interquartile range: 6.6-92.2]). The relationship between total infarct volume and good outcome probability was nearly linear for infarct volumes between 0 mL and 250 mL. In this range, a 10% increase in the probability of achieving mRS 0-2 required a decrease in infarct volume of approximately 34.0 mL (95% confidence interval: -32.5 to -35.6). At infarct volumes above 250 mL, the probability of achieving mRS 0-2 probability was near zero. The relationships of tissue-specific infarct volumes and parenchymal hemorrhage volume generally showed similar patterns, although variability was high.

Conclusion: There seems to be a near-linear association between total infarct volume and probability of achieving good outcome for infarcts up to 250 mL, whereas patients with infarct volumes greater than 250 mL are highly unlikely to have a favorable outcome.

Leukoaraiosis (LA) manifests as cerebral white matter hyperintensities on T2-weighted magnetic resonance imaging scans and corresponds to white matter lesions or abnormalities in brain tissue. Clinically, it is generally detected in the early 40s and is highly prevalent globally in individuals aged >60 years. From the imaging perspective, LA can present as several heterogeneous forms, including punctate and patchy lesions in deep or subcortical white matter; lesions with periventricular caps, a pencil-thin lining, and smooth halo; as well as irregular lesions, which are not always benign. Given its potential of having deleterious effects on normal brain function and the resulting increase in public health burden, considerable effort has been focused on investigating the associations between various risk factors and LA risk, and developing its associated clinical interventions. However, study results have been inconsistent, most likely due to potential differences in study designs, neuroimaging methods, and sample sizes as well as the inherent neuroimaging heterogeneity and multi-factorial nature of LA. In this article, we provided an overview of LA and summarized the current knowledge regarding its epidemiology, neuroimaging classification, pathological characteristics, risk factors, and potential intervention strategies.

Background and purpose: Non-high-density lipoprotein cholesterol (non-HDL-C), which represents the total cholesterol content of all pro-atherogenic lipoproteins, has recently been included as a new target for lipid-lowering therapy in high-risk atherosclerotic patients in multiple guidelines. Herein, we aimed to explore the relationship between non-HDL-C level and the efficacy and safety of ticagrelor-aspirin versus clopidogrel-aspirin in preventing stroke recurrence.

Methods: This study comprised a post hoc analysis of the CHANCE-2 (Ticagrelor or Clopidogrel in High-Risk Patients with Acute Nondisabling Cerebrovascular Events II) trial, from which 5,901 patients with complete data on non-HDL-C were included and categorized by median non-HDL-C levels, using a cutoff of 3.5 mmol/L. The primary efficacy and safety outcomes were recurrent stroke and severe or moderate bleeding within 90 days.

Results: Ticagrelor-aspirin significantly reduced the risk of recurrent stroke in patients with low non-HDL-C (71 [4.8%] vs. 119 [7.7%]; adjusted hazard ratio [HR] 0.54; 95% confidence interval [CI], 0.40-0.74), but not in those with high non-HDL-C (107 [7.3%] vs. 108 [7.6%]; adjusted HR, 0.88; 95% CI, 0.67-1.16), compared with clopidogrel-aspirin (P for interaction=0.010). When analyzed as a continuous variable, the benefit of ticagrelor-aspirin for recurrent stroke decreased as non-HDL-C levels increased. No significant differences in the treatment assignments across the non-HDL-C groups were observed in terms of the rate of severe or moderate bleeding (5 [0.3%] vs. 8 [0.5%] in the low non-HDL-C group; 4 [0.3%] vs. 2 [0.1%] in the high non-HDL-C group; P for interaction=0.425).

Conclusion: CHANCE-2 participants with low non-HDL-C levels received more clinical benefit from ticagrelor-aspirin versus clopidogrel-aspirin compared to those with high non-HDL-C, following minor ischemic stroke or transient ischemic attack.

Background and purpose: The accurate prediction of functional outcomes in patients with acute ischemic stroke (AIS) is crucial for informed clinical decision-making and optimal resource utilization. As such, this study aimed to construct an ensemble deep learning model that integrates multimodal imaging and clinical data to predict the 90-day functional outcomes after AIS.

Methods: We used data from the Korean Stroke Neuroimaging Initiative database, a prospective multicenter stroke registry to construct an ensemble model integrated individual 3D convolutional neural networks for diffusion-weighted imaging and fluid-attenuated inversion recovery (FLAIR), along with a deep neural network for clinical data, to predict 90-day functional independence after AIS using a modified Rankin Scale (mRS) of 3-6. To evaluate the performance of the ensemble model, we compared the area under the curve (AUC) of the proposed method with that of individual models trained on each modality to identify patients with AIS with an mRS score of 3-6.

Results: Of the 2,606 patients with AIS, 993 (38.1%) achieved an mRS score of 3-6 at 90 days post-stroke. Our model achieved AUC values of 0.830 (standard cross-validation [CV]) and 0.779 (time-based CV), which significantly outperformed the other models relying on single modalities: b-value of 1,000 s/mm2 (P<0.001), apparent diffusion coefficient map (P<0.001), FLAIR (P<0.001), and clinical data (P=0.004).

Conclusion: The integration of multimodal imaging and clinical data resulted in superior prediction of the 90-day functional outcomes in AIS patients compared to the use of a single data modality.

Cancer can induce hypercoagulability, which may lead to stroke. This occurs when tumor cells activate platelets as part of their growth and metastasis. Tumor cells activate platelets by generating thrombin and expressing tissue factor, resulting in tumor cell-induced platelet aggregation. Histopathological studies of thrombi obtained during endovascular thrombectomy in patients with acute stroke and active cancer have shown a high proportion of platelets and thrombin. This underscores the crucial roles of platelets and thrombin in cancer-associated thrombosis. Cancer-associated stroke typically occurs in patients with active cancer and is characterized by distinctive features. These features include multiple infarctions across multiple vascular territories, markedly elevated blood D-dimer levels, and metastasis. The presence of cardiac vegetations on echocardiography is a robust indicator of cancer-associated stroke. Suspicion of cancer-associated stroke during endovascular thrombectomy arises when white thrombi are detected, particularly in patients with active cancer. Cancer-associated stroke is almost certain when histopathological examination of thrombi shows a very high platelet and a very low erythrocyte composition. Patients with cancer-associated stroke have high risks of mortality and recurrent stroke. However, limited data are available on the optimal treatment regimen for stroke prevention in these patients. Thrombosis mechanism in cancer is well understood, and distinct therapeutic targets involving thrombin and platelets have been identified. Therefore, direct thrombin inhibitors and/or antiplatelet agents may effectively prevent stroke recurrence. Additionally, this strategy has potential benefits in cancer treatment as accumulating evidence suggests that aspirin use reduces cancer progression, metastasis, and cancer-related mortality. However, clinical trials are necessary to assess the efficacy of this strategy involving the use of direct thrombin inhibitors and/or antiplatelet therapies.