Kittner, Steven J.

Abstract This study aimed to investigate the influence of stroke lesions in predefined highly interconnected (rich‐club) brain regions on functional outcome post‐stroke, determine their spatial specificity and explore the effects of biological sex on their relevance. We analyzed MRI data recorded at index stroke and ~3‐months modified Rankin Scale (mRS) data from patients with acute ischemic stroke enrolled in the multisite MRI‐GENIE study. Spatially normalized structural stroke lesions were parcellated into 108 atlas‐defined bilateral (sub)cortical brain regions. Unfavorable outcome (mRS > 2) was modeled in a Bayesian logistic regression framework. Effects of individual brain regions were captured as two compound effects for (i) six bilateral rich club and (ii) all further non‐rich club regions. In spatial specificity analyses, we randomized the split into “rich club” and “non‐rich club” regions and compared the effect of the actual rich club regions to the distribution of effects from 1000 combinations of six random regions. In sex‐specific analyses, we introduced an additional hierarchical level in our model structure to compare male and female‐specific rich club effects. A total of 822 patients (age: 64.7[15.0], 39% women) were analyzed. Rich club regions had substantial relevance in explaining unfavorable functional outcome (mean of posterior distribution: 0.08, area under the curve: 0.8). In particular, the rich club‐combination had a higher relevance than 98.4% of random constellations. Rich club regions were substantially more important in explaining long‐term outcome in women than in men. All in all, lesions in rich club regions were associated with increased odds of unfavorable outcome. These effects were spatially specific and more pronounced in women.

Background: The heterogeneity of ischemic stroke has relevance for determination of risk for recurrence, prediction of outcomes, and identification of genetic risk factors. Among the many systems created for classifying patients into subtypes of stroke, the Trial of Org10172 Acute Stroke Trial (TOAST) and Causative Classification of Stroke (CCS) systems are among the most commonly used. Both systems generate similar subtype diagnoses although the two systems use different criteria for subtype assignment. The purpose of this study was to assess the level of agreement between subtype classification made using these two systems. Methods: Study subjects included 14,106 adult men and women accrued in the NINDS Stroke Genetics Network (SiGN) across 21 centers in the US and Europe classified independently using both with TOAST and CCS. Kappa statistics were calculated for the five major ischemic stroke subtypes common to both systems. Results: The overall agreement between TOAST and CCS was moderate (agreement rate, 70%; kappa = 0.60, 95% confidence interval [CI], 0.59 to 0.61). Agreement varied widely across study sites, ranging from 28% to 90%. Agreement on specific subtypes was highest for large artery atherosclerosis (kappa, 0.71; 95% CI, 0.70 to 0.73) and lowest for small artery occlusion (kappa = 0.56, 95%CI, 0.54 to 0.58). Conclusion: Agreement between TOAST and CCS diagnoses was only moderate. Caution is warranted when comparing or combining results based upon the two systems. Re-phenotyping using one system should be encouraged when possible to ensure valid results. Table: Cross-classification of stroke subtypes by TOAST and CCS

Objectives: Genetic and environmental influences contribute to differences in ischemic stroke phenotype across race-ethnic groups. The Causative Classification System (CSS) is an ischemic stroke phenotype characterization tool developed for etiological and causative classification. The CCS phenotypes across ethnic groups were analyzed and compared. Methods: The NINDS S troke G enetics N etwork (SiGN) is a large consortium aimed at identifying subtype-specific ischemic stroke susceptibility genes. Over 12,000 subjects at 22 sites across US and Europe were enrolled, with CSS locally adjudicated. Central CCS quality control was performed on all participants. Analysis was performed on 7,674 patients with complete workups. Findings: Significant ethnic differences were found in stroke etiology across ethnic groups (p <0.0001) with large artery atherosclerosis present most commonly in Asians, cardioembolic sources among Caucasian Europeans, and lacunar strokes among African Americans. African Americans were the youngest ethnic cohort analyzed (see Table 1). Conclusion: Significant differences in stroke subtypes are present across ethnic groups. Future genetic analyses may contribute to our understanding of these differences.

BACKGROUND AND PURPOSE: Ischemic stroke is a heterogeneous complex disease that results from the interaction between numerous genetic and environmental risk factors. Stroke heritability is approximately 40% in the overall population, similar to that of cancer, although there may be important differences in the etiologies of young (<49 years old) versus older onset ischemic stroke. The goal of this study was to conduct a mechanistic genome-wide association (GWA) analysis contrasting the gene pathways involved in young- and old-onset stroke. METHODS: Our analysis included 946 young subjects (448 ischemic stroke subjects and 498 age and sex matched controls, mean age of stroke onset 41 years) and 2558 older subjects (1070 ischemic stroke subjects and 1488 controls, mean age of stroke onset 66.5 years). All subjects were of European ancestry. The young onset cases and controls were genotyped by Illumina Omni1-Quad array with imputation up to 1.8 M SNPs and the old onset cases and controls by a combination of 550 v1 and v3 as well as 610 Illumina arrays with imputation up to 5.0 M SNPs. Single SNP association testing was performed within each contributing study by logistic regression adjusting for age, sex and population structure. Numbers of genes assigned with significant SNPs (P<=0.0001) were 18 and 25 with young- and old-onset groups, respectively, which were used as an input to pathway analysis by the Pathway Studio (Ariadne Genomics, Co.) RESULTS: SNP-level analyses did not reveal any significant associations with each group that met Bonferroni adjusted thresholds for statistical significance. In analyzing the ‘suggestively associated SNPs’, protein-protein interaction pathway analysis in the young-onset group revealed a strong segregation of significant hub proteins (CAT, RUNX1, and EIF4E) with more than 100 connections, while in the old-onset group the interaction networks were dispersed without much segregation. We observed that protein-protein interaction networks have a power-law connectivity distribution in both young- and old-onset groups. We identified 3 significant and focused pathways, ROS metabolism, melanogenesis, and IGF1R->CdEBPA/FOXO1A signaling, enriched for young-onset stroke (P<0.04). In the old-onset group, 10 pathways were identified involving broader signaling pathways. CONCLUSIONS: Young-onset stroke may involve genetic predisposition characterized by a few distinct and finite pathways. Our data suggests that young onset stroke may involve a few pathways that relate to a small number of key hub proteins, including ROS metabolism, melanogenesis, and IGF1R->CEBPA/FOXO1A signaling. Our analysis is consistent with the hypothesis that different pathways may be involved in young versus old onset stroke.

Introduction: Estrogens have been associated with risk of stroke in both men and women. Estradiol and estrone are synthesized from testosterone via the aromatase enzyme in both men and postmenopausal women. Genetic variation affecting function of the aromatase gene (CYP19A1) has been linked to elevated estradiol levels. We tested the hypothesis that these variants, as well as a modifying gene, are associated with ischemic stroke in three cohorts: the Vitamin Intervention for Stroke Prevention trial (VISP), Women’s Health Initiative (WHI), and Genetic Susceptibility to Early-Onset Stroke study (GEOS). Methods: Genomic data for individuals of European descent with ischemic stroke was compared against an equal number of controls: VISP (1723 cases - 65% men), WHI (436 cases - all postmenopausal women), GEOS (557 cases - all men age <55). Genotyping was funded through the Genomics and Randomized Trials Network (GARNET) and Gene Environment Association Studies (GENEVA) consortia. All cohorts used the Illumina Omni1-Quad chip for genotyping with imputation as needed. Eight candidate single nucleotide polymorphisms (SNPs) in CYP19A1 on chromosome 15 and an additional regulatory SNP in a modifier locus on chromosome 8 were chosen for analysis due to coding status or previous association with estradiol levels. Analyses were adjusted for age, hormone therapy use (women), and the top three principal components. Association results were determined as odds ratios for individual SNPs (effect allele = minor), with significance level of p<0.05. Results: GWAS results are tabulated. There were no significant differences in allele frequency for four SNPs in CYP19A1 among stroke patients and controls. The additional four SNPs were monomorphic in all three populations. The modifier gene SNP (rs1864729) was less common among young men with prior stroke in GEOS, but not men in VISP or in women. Conclusions: Aromatase genetic variation was not associated with risk of stroke in these three cohorts. One SNP that modifies aromatase function was less common among young, but not older men with stroke. Replication in other studies is necessary.

Background: Causative Classification of Stroke (CCS) system is a novel, web-based, and fully-computerized algorithm that integrates clinical, diagnostic, and etiologic stroke characteristics in an evidence-based manner (available at http://ccs.mgh.harvard.edu ) and provides in a given patient both phenotypic subtypes where abnormal evaluation findings are simply organized in major etiologic categories and causative subtypes where abnormal evaluation findings are filtered through a decision making process to identify the most likely cause of stroke. In this study, we sought to identify the concordance between phenotypic and causative CCS subtypes in the NINDS Stroke Genetics Network (SiGN). Methods: A total of 24 adjudicators from 15 US and European sites certified in CCS retrospectively determined CCS subtypes through chart review in 7134 patients. The CCS software provided causative subtypes in 3 confidence levels as “evident”, “probable” or “possible” based on the weight of causal evidence. We estimated how often CCS classified a given major abnormal evaluation finding as the causative stroke mechanism. Results: CCS assigned 55% of patients with the phenotype of moderate-to-severe atherosclerotic stenosis into the causative category of “evident large artery atherosclerosis”. Likewise, 50% of patients with a major cardiac source (such as atrial fibrillation) were classified into “evident cardio-aortic embolism”, and 74% of those with imaging evidence of a typical lacunar infarct into “evident small artery occlusion”. In 20% to 30% of patients, phenotypic and causative categories were the same but the causative subtype was assigned with a lower level of confidence. Six to 20% of patients with a given phenotypic feature were classified into a different causative category. Failure to investigate for other subtypes in the presence of a positive test finding and the presence of multiple competing etiologies were key contributors to the observed disconcordance between phenotypic and causative subtypes. Conclusions: Our findings show that the presence of a major abnormality in stroke evaluation does not necessarily mean that it is always the cause. The ability of CCS system to discriminate between phenotypic and causative characteristics of stroke etiology provides the opportunity to refine stroke phenotypes for use in research studies.