Rosand, Jonathan M.

Background: Identification of clinical and imaging characteristics of ischemic stroke that can predict the underlying etiology is critical for accurate stroke subtyping. We sought to investigate the relationship between stroke characteristics and stroke etiology in 7,591 patients in the NINDS Stroke Genetics Network Study. Methods: We determined the probability of having a major etiology in the presence of certain stroke characteristics that were previously linked to specific stroke mechanisms. We excluded patients with multiple competing major etiologies. All patients had brain imaging, intra- and extra-cranial vascular imaging, and complete cardiac assessment. We defined major etiologies according to the Causative Classification of Stroke System Criteria without reference to the stroke characteristics under study. Results: History of 1 ipsilateral ischemic events within the last month, the presence of internal watershed infarcts or multiple, ipsilateral, unilateral, and temporally separate infarcts were associated with large artery atherosclerosis (Table). Multiple acute infarcts in different cerebral circulations were more common in patients with major cardiac source of embolism as well as in those with uncommon/cryptogenic causes. Stereotypic lacunar TIAs within the last week and lacunar syndrome were associated with lacunar infarction Discussion: Clinical and imaging stroke characteristics appear to predict stroke etiologies. This information may find utility in developing algorithms for a guided stroke evaluation as well as help identify the most likely causative etiology in the presence of multiple competing etiologies.

Background: Fundamental advances in stroke care will require pooling imaging phenotype data from multiple centers, to complement the current aggregation of genomic, environmental, and clinical information. Sharing clinically acquired MRI data from multiple hospitals is challenging due to inherent heterogeneity of clinical data, where the same MRI series may be labeled differently depending on vendor and hospital. Furthermore, the de-identification process may remove data describing the MRI series, requiring human review. However, manually annotating the MRI series is not only laborious and slow but prone to human error. In this work, we present a recurrent convolutional neural network (RCNN) for automated classification of the MRI series. Methods: We randomly selected 1000 subjects from the MRI-GENetics Interface Exploration study and partitioned them into 800 training, 100 validation and 100 testing subjects. We categorized the MRI series into 24 groups (see Table). The RCNN used a modified AlexNet to extract features from 2D slices. AlexNet was pretrained on ImageNet photographs. Since clinical MRI are 3D and 4D, a gated recurrent unit neural network was used to aggregate information from multiple 2D slices to make the final prediction. Results: We achieved a classification accuracy (correct/total cases) of 99.8%, 98.5% and 97.5% on the training, validation and testing set, respectively. The averaged F1-score (percent overlap between predicted cases and actual cases) over all categories were 99.8% 98.2% and 94.4% on the training, validation and testing set. Conclusion: We showed that automated annotation of MRI series by repurposing deep-learning techniques used for photographic image recognition tasks is feasible. Such methods can be used to facilitate high throughput curation of MRI data acquired across multiple centers and enable scientifically productive collaboration by researchers and, ultimately enhancing big data stroke research.

Introduction: Genetic studies of early-onset disease have been an effective strategy to identify novel pathways and drug targets relevant to later-onset disease. Few studies have investigated the role of common genetic variation in the etiology of early-onset ischemic stroke (IS). Methods: We performed a GWAS meta-analysis of 38 studies from 10 countries, comprised of 5,847 IS cases of European ancestry under age 60 and 32,533 controls. Results: We identified two genome-wide significant (p< 5 x 10 -8 ) loci (see Figure). The ABO locus has previously been associated with venous thrombosis and ischemic stroke in predominantly older adults, but the effect size of our top SNP (OR 1.18; p = 9.1 x 10 -12 ) is larger than the effect size for this same SNP in MEGASTROKE (OR: 1.05; p = 6.5 x 10 -5 ). The lead SNP at the BCL7A/MLXIP locus is a novel GWAS finding for stroke (OR 1.14, 95% CI 1.08-1.19; p = 1.7 x 10 -8 ) and is noteworthy because of prior reports linking SNPs in these genes to BMI and blood pressure. Conclusions: We identified a novel locus that is near variants associated with BMI and blood pressure. Further studies are needed to confirm this locus, examine subtype specificity, and determine its function. The larger effect size observed at the ABO in this early-onset IS sample compared to older-onset IS samples is consistent with a larger role for prothrombotic mechanisms in early-onset IS.

Introduction: White matter hyperintensity volume (WMHv) is an important and highly heritable cerebrovascular phenotype; however, manual or semi-automated approaches to clinically acquired MRI analysis hinder large-scale studies in acute ischemic stroke (AIS). In this work, we develop a high-throughput, fully automated WMHv analysis pipeline for clinical fluid-attenuated inversion recovery (FLAIR) images to facilitate rapid genetic discovery in AIS. Methods: Automated WMHv extraction from multiple subjects relies on significant pre-processing of medical scans, including co-registration of the images. To reduce the effects of anisotropic voxel sizes, each FLAIR image is upsampled using bi-cubic interpolation. Brain extraction is performed using RObust Brain EXtraction (ROBEX). Images are then registered to an in-house FLAIR template using Advanced Normalization Tools (ANTs). The spatial covariation of WMH is learned through principal component analysis (PCA) of manual outlines from 100 subjects. Areas of leukoaraiosis are identified and separated from other lesions using the PCA modes. Volumes are then computed using non-interpolated slices for each subject. Standard deviation (SD) in WMHv (9 subjects; 6 raters each) is calculated as a measure of variability. Good agreement between automated and manual outlines is assessed in 358 subjects (automated WMHv within 3SD of manual WMHv). Results: As part of the MRI - Gen etics I nterface E xploration (MRI-GENIE) study, WMHv were calculated on a set of 2703 FLAIR images of patients from 12 independent AIS cohorts (sites). Results are shown in Figure 1. Comparing manual and automated WMHv shows that 88% of the automated WMHv fall within 3 SD from the manual WMHv, suggesting good agreement. Conclusion: WMHv segmentation using a fully-automated pipeline for analysis of clinical MRIs is both feasible and accurate. Ongoing analysis of the extracted WMHv is expected to advance current knowledge of risks and outcomes in AIS.

Introduction: Even though neuroinflammation is increasingly recognized as an essential contributor to ischemic brain injury, the exact underlying mechanisms remain unclear. Higher expression of the vascular cell adhesion molecule 1 (VCAM-1) increases leukocyte-brain endothelium interaction and has been associated with larger infarct size following acute ischemic stroke (AIS). The activation of the TGF-β signaling pathways can down-regulate VCAM-1 expression and ameliorate deleterious tissue outcome during neuroinflammation. Hypothesis: We sought to investigate whether genetic variation in the TGF-beta pathway and adhesion molecule genes is associated with acute stroke lesion size. Methods: We completed genome-wide association (GWA) testing and diffusion-weighted imaging lesion volume (DWIv) analysis in a discovery cohort of 532 AIS patients of European ancestry enrolled within 48 hours of symptom onset. An independent European ancestry cohort of 724 AIS patients with GWA data and automated DWIv served as replication cohort. GWA testing per SNP was performed using linear regression modeling of natural log-transformed DWIv adjusted for age, sex and relevant principal components. We selected 42 inflammatory genes in the TGF-beta pathway for a gene-based analysis using VEGAS (Versatile Gene-based -Association Study) software. A pre-specified discovery phase Bonferroni-corrected threshold was set at p<0.001. In the replication phase, 14 SNPs overlapping were tested at the Bonferroni-corrected p-value threshold of p<0.004. Results: Of all genes in the TGF-beta pathway, VCAM1 (p=0.0006) was significantly associated with DWIv in the discovery AIS cohort (age: 66 ± 14.9 years, sex: 63.4% male, DWIv: 2.2cm 3 (IQR: 0.6-11.7cm 3 )). A single SNP within the VCAM1 gene boundaries (rs3176876 (BETA=0.2341, p=0.003)) was significantly associated with DWIv in the replication AIS cohort (age: 65 ± 14.1 years, sex: 68.6% male, DWIv: 4cm 3 (IQR: 1.3-17.2cm 3 )). Conclusion: The genetic variant rs3176876 in the VCAM1 gene is associated with larger infarct lesion size measured on brain MRI of AIS patients. These findings suggest genetic contribution to the pro-inflammatory mechanisms in acute cerebral ischemia and warrant further investigation.

Introduction: White matter hyperintensity (WMH) is a highly heritable trait and a significant contributor to stroke risk and severity. Vascular risk factors contribute to WMH severity; however, knowledge of the determinants of WMH in acute ischemic stroke (AIS) is still limited. Hypothesis: WMH volume (WMHv) varies across AIS subtypes and is modified by vascular risk factors. Methods: We extracted WMHv from the clinical MRI scans of 2683 AIS subjects from the MRI-Genetics Interface Exploration (MRI-GENIE) study using a novel fully-automated, volumetric analysis pipeline. Demographic data, stroke risk factors and stroke subtyping for the Causative Classification of Stroke (CCS) were performed at each of the 12 international study sites. WMHv was natural log-transformed for linear regression analyses. Results: Median WMHv was 5.7cm 3 (interquartile range (IQR): 2.2-12.8cm 3 ). In univariable analysis, age (63.1 ± 14.7 years, β=0.04, SE=0.002), prior stroke (10.2%, β=0.66, SE=0.08), hypertension (65.4%, β=0.75, SE=0.05), diabetes mellitus (23.1%, β=0.35, SE=0.06), coronary artery disease (17.6%, β=0.04, SE=0.002), and atrial fibrillation (14.6%, β=0.48, SE=0.07) were significant predictors of WMHv (all p<0.0001), as well as smoking status (52.2%, β=0.15, SE=0.05, p=0.005), race (16.5% Non-Caucasian, β=0.25, SE=0.07) and ethnicity (8.2% Hispanic, β=0.30, SE=0.11) (all p<0.01). In multivariable analysis, age (β=0.04, SE=0.002), prior stroke (β=0.56, SE=0.08), hypertension (β=0.33, SE=0.05), smoking status (β=0.16, SE=0.05), race (β=0.42, SE=0.06), and ethnicity (β=0.34, SE=0.09) were independent predictors of WMHv (all p<0.0001), as well as diabetes mellitus (β=0.13, SE=0.06, p=0.02). WMHv differed significantly (p<0.0001, unadjusted) across CCS stroke subtypes: cardioembolic stroke (8.0cm 3 , IQR: 4.2-15.4cm 3 ), large-artery stroke (6.9cm 3 , IQR: 3.1-14.7cm 3 ), small-vessel stroke (5.8cm 3 , IQR: 2.5-13.5cm 3 ), stroke of undetermined (4.7cm 3 , IQR: 1.6-11.0cm 3 ) or other (2.55cm 3 , IQR: 0.9-8.8cm 3 ) causes. Conclusion: In this largest-to-date, multicenter hospital-based cohort of AIS patients with automated WMHv analysis, common vascular risk factors contribute significantly to WMH burden and WMHv varies by CCS subtype.

Introduction: In observational studies, higher plasma high-density lipoprotein cholesterol (HDL-C) has been associated with increased risk of spontaneous intracerebral hemorrhage (ICH). Common DNA sequence variants within the cholesteryl ester transfer protein ( CETP ) gene decrease CETP protein activity and increase plasma HDL-C; as such, medicines that inhibit CETP and raise HDL-C are in clinical development to combat coronary artery disease. Hypothesis: Common CETP DNA sequence variants associated with higher HDL-C also increase risk for ICH. Methods: We performed a two-stage case-control genetic association study in Caucasians. The discovery phase utilized data on 12 independent loci within CETP (+/- 50 kilobases) from 3 genome-wide association studies of ICH. Replication involved direct genotyping in 5 additional studies. We also constructed a genetic risk score with 7 independent CETP variants and tested it for association with HDL-C and ICH risk. We used principal component analysis to account for population structure and a Bonferroni-adjusted p<0.004 (12 tests) to declare statistical significance. Results: The discovery phase included 1149 ICH cases (43% lobar hemorrhages) and 1238 controls. Twelve variants were nominally associated (p<0.05) with ICH, with the strongest association at the rs173539 locus (Figure 1: OR 1.25, 95%CI 1.11-1.41; p=6.0x10 -4 ) and no heterogeneity across studies (I 2 =0%). This association was replicated in 1625 cases (43% lobar hemorrhages) and 1845 controls (OR 1.12, 95%CI 1.02-1.24; p=0.03). A genetic score of independent CETP variants known to increase HDL-C by ~2.85 mg/dL was strongly associated with ICH risk (OR 1.86, 95%CI 1.44-2.40; p=1.4x10 -6 ). Conclusion: Genetic variants in CETP associated with increased HDL-C raise the risk of ICH. Given ongoing therapeutic development in CETP inhibition and other HDL-raising strategies, further exploration of potential adverse cerebrovascular outcomes is warranted.

Introduction: The MRI-Genetics Interface Exploration (MRI-GENIE) study is the first international collaboration that aims to facilitate genetic discoveries in clinical cohorts of patients with acute ischemic stroke (AIS). We have amassed the largest-to-date collection of AIS cases with brain MRI scans and genome-wide genotyping to test the role of genetic susceptibility in MRI-based cerebrovascular traits. Objective/Hypothesis: To elucidate the genetic architecture of white matter hyperintensity (WMH) burden in AIS patients. Methods: Using a novel automated algorithm, we extracted WMH volume (WMHv) from clinical MRI scans of 2704 AIS patients (age 63.1 ± 14.7 years, 60.6% male) of European ancestry. Quality control (QC) measures were undertaken per subject and per SNP, excluding subjects with non-European ancestry and poor genotyping, as well as SNPs deviating from Hardy-Weinberg equilibrium and high levels of missingness. Imputation to the Haplotype Reference Consortium (HRC version r1.1) was conducted for 1712 remaining subjects with 2.8 million SNPs on the Michigan Imputation Server. After exclusion of poorly imputed SNPs (R 2 <0.5) and SNPs with minor allele frequency < 1%, 7.7 million SNPs remained for further analysis. Genome-wide association testing of natural log-transformed WMHv on the allelic dosage per SNP was adjusted for age, sex and principal components 1-10. Results: Genome-wide association testing has identified a novel locus on chromosome 2 (T allele at rs72856504) near the LDL Receptor related Protein 1B gene (LRP1B) that was significantly associated with WMHv burden in AIS (β=0.54, SE=0.098, p=3.65*10 -8 ). Conclusion: We have identified a novel locus (T allele rs72856504) on chromosome 2 near the LRP1B gene, which is specific for WMH in AIS and has not been previosuly described in stroke-free WMH cohorts. A replication effort involving additional independent cohorts of AIS patients with brain MRI and genome-wide genotyping is ongoing.

Glial fibrillary acidic protein (GFAP) is cleared by the Food and Drug Administration (FDA) to determine need for head computed tomography (CT) within 12 h after mild traumatic brain injury (TBI) (Glasgow Coma Score [GCS] 13–15); S100 calcium-binding protein B (S100B) serves this function in Europe. This phase 1 biomarker cohort analysis of the multi-center, observational Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACK-TBI) study compares GFAP's diagnostic performance, measured on a rapid point-of-care platform, against protein S100B to predict intracranial abnormalities on CT within 24 h post-injury across the spectrum of TBI (GCS 3–15). Head CT scan performed in TBI subjects and blood was collected for all consenting subjects presenting to 18 United States level 1 trauma centers. Plasma was analyzed on a point-of-care device prototype assay for GFAP and serum was analyzed for S100B. In 1359 patients with TBI (GCS 3–15), mean (standard deviation [SD]) age = 40.1 (17.0) years; 68% were male. Plasma GFAP levels were significantly higher in CT+ TBI subjects (median = 1358 pg/mL, interquartile range [IQR]: 472–3803) than in CT- TBI subjects (median = 116 pg/mL, IQR: 26–397) or orthopedic trauma controls ( n = 122; median = 13 pg/mL, IQR: 7–20), p < 0.001. Serum S100B levels were likewise higher in CT+ TBI subjects (median = 0.17 μg/L, IQR: 0.09–0.38) than in CT- TBI subjects (median = 0.10 μg/L, IQR: 0.06–0.18), p < 0.001. Receiver operating characteristic curves were generated for prediction of intracranial injury on admission CT scan; area under the curve (AUC) for GFAP was significantly higher than for S100B in the same cohort (GFAP AUC – 0.85, 95% confidence interval [CI] 0.83–0.87; S100B AUC – 0.67, 95% CI 0.64-0.70; p < 0.001). GFAP, measured on a point-of-care platform prototype assay, has high discriminative ability to predict intracranial abnormalities on CT scan in patients with TBI across the full injury spectrum of GCS 3–15 through 24 h post-injury. GFAP substantially outperforms S100B.

ObjectiveObservational studies point to an inverse correlation between low‐density lipoprotein (LDL) cholesterol levels and risk of intracerebral hemorrhage (ICH), but it remains unclear whether this association is causal. We tested the hypothesis that genetically elevated LDL is associated with reduced risk of ICH.MethodsWe constructed one polygenic risk score (PRS) per lipid trait (total cholesterol, LDL, high‐density lipoprotein [HDL], and triglycerides) using independent genomewide significant single nucleotide polymorphisms (SNPs) for each trait. We used data from 316,428 individuals enrolled in the UK Biobank to estimate the effect of each PRS on its corresponding trait, and data from 1,286 ICH cases and 1,261 matched controls to estimate the effect of each PRS on ICH risk. We used these estimates to conduct Mendelian Randomization (MR) analyses.ResultsWe identified 410, 339, 393, and 317 lipid‐related SNPs for total cholesterol, LDL, HDL, and triglycerides, respectively. All four PRSs were strongly associated with their corresponding trait (all p < 1.00 × 10‐100). While one SD increase in the PRSs for total cholesterol (odds ratio [OR] = 0.92; 95% confidence interval [CI] = 0.85–0.99; p = 0.03) and LDL cholesterol (OR = 0.88; 95% CI = 0.81–0.95; p = 0.002) were inversely associated with ICH risk, no significant associations were found for HDL and triglycerides (both p > 0.05). MR analyses indicated that 1mmol/L (38.67mg/dL) increase of genetically instrumented total and LDL cholesterol were associated with 23% (OR = 0.77; 95% CI = 0.65–0.98; p = 0.03) and 41% lower risks of ICH (OR = 0.59; 95% CI = 0.42–0.82; p = 0.002), respectively.InterpretationGenetically elevated LDL levels were associated with lower risk of ICH, providing support for a potential causal role of LDL cholesterol in ICH. ANN NEUROL 2020 ANN NEUROL 2020;88:56–66