Epigenome-wide DNA Methylation Profiling in Aneurysmal Subarachnoid Hemorrhage and Delayed Ischemic Neurologic Deficit

Overview

The goal of this observational study is to investigate DNA methylation changes in adults with bleeding from brain aneurysm, which is called aneurysmal subarachnoid hemorrhage (aSAH), and their association with delayed ischemic neurologic deficit (DIND). The main questions it aims to answer are: Are there specific DNA methylation changes in peripheral blood that differentiate patients with aSAH from healthy individuals? Can DNA methylation changes in peripheral blood predict the development of DIND following aSAH? Researchers will compare blood DNA methylation profiles of aSAH patients to healthy controls and also do subgroup analysis of patients with DIND versus those without DIND to see if there are distinct methylation patterns associated with aSAH and DIND. Participants with aSAH will: * Have a blood sample collected shortly after admission to hospital. * Undergo epigenome-wide DNA methylation profiling using the Infinium MethylationEPIC v2.0 BeadChip microarray. * Be monitored for the development of DIND, defined by clinical symptoms and radiographic vasospasm. This study aims to identify potential epigenetic biomarkers for aSAH susceptibility and DIND risk, which could improve early diagnosis and risk stratification in affected patients.

This is a prospective observational case-control epigenome-wide association study (EWAS) designed to investigate DNA methylation changes in peripheral blood of patients with aneurysmal subarachnoid hemorrhage (aSAH) and their potential association with delayed ischemic neurologic deficit (DIND). The study aims to identify epigenetic biomarkers that may contribute to the pathophysiology of aSAH and DIND, as well as to assess whether these methylation patterns can serve as predictive markers for disease progression. Study Design and Procedures Population: * Patients diagnosed with aSAH will be prospectively recruited from a tertiary neurosurgical center. * The prospectively recruited patients will be anonymized and classified into DIND and non-DIND groups based on clinical and radiological criteria. * Healthy control subjects will be matched for age, sex, race, and ethnicity using a publicly available dataset (GSE246337). Data Collection and Processing: * Peripheral whole blood samples (10 mL) will be collected within days 1-4 post-hemorrhage before the onset of vasospasm. * DNA will be extracted using the salting-out method, followed by bisulfite conversion for methylation profiling. * Genome-wide DNA methylation analysis will be performed using the Infinium MethylationEPIC v2.0 BeadChip microarray, covering \~850K CpG sites. * Genotyping will be performed using Illumina Global Screening Array (GSA-24) v3.0 consisting of over 654 000 unique loci. PLINK v1.9 software will be used for the variant filtering and statistical analyses. We consider in this analysis only bi-allelic SNPs on autosomes. If this analysis yields significant associations between SNPs and DIND, methylation quantitative trait loci will be analyzed for examination of long-range interactions between relevant SNPs and differentially methylated CpG sites. These SNP-methylation interactions could provide unique and novel data on the vasospasm mechanism. * Data processing will be performed using SeSAME R package with rigorous quality control measures, including: * Removal of probes spanning SNPs with a minor allele frequency \>1%. * Filtering out probes with low bead count or detection p-values. * Exclusion of sex chromosome-associated probes and incorrectly mapped probes. Diagnosis of DIND: * DIND will be clinically defined as a new focal neurological deficit or Glasgow Coma Scale (GCS) decline of ≥2 points, lasting \>1 hour and not attributable to other causes such as hydrocephalus, hyponatremia or rebleeding. * Radiological confirmation of vasospasm will be based on cerebral angiography, CT angiography (CTA), or transcranial Doppler (TCD) findings. Statistical Analysis Plan Differentially Methylated Probes (DMPs): * Identification of DMPs will be conducted using linear regression models, with Benjamini-Hochberg correction for multiple comparisons. * Only CpG sites with an absolute Δβ-value \>5% and an adjusted p-value \<0.05 will be considered significant. * Epigenetic clocks (Horvath clock, Hannum clock, SkinBlood clock, PhenoAge clock, GrimAge clock) will estimate biological age based on DNA methylation patterns at specific CpG sites Machine Learning Analysis: * LASSO regression will be applied for feature selection to identify methylation patterns most predictive of aSAH vs. healthy controls and DIND vs. non-DIND. * Performance of the predictive models will be assessed using the R² metric. Functional Enrichment Analysis: * Gene Set Enrichment Analysis (GSEA) using "GENE2FUNC" function of Functional Mapping and Annotation of Genome-Wide Association Studies platform (FUMA GWAS) and Genomic Regions Enrichment of Annotations Tool (GREAT) will be used to identify biological pathways associated with significant methylation changes. * Pathway analysis will focus on immune response, neuroinflammation, endothelial dysfunction, and phosphorous metabolism, as these processes have been implicated in aSAH pathophysiology. Quality Assurance and Data Validation Internal Data Validation: * Automated data checks will be used to identify inconsistencies in methylation profiles. * Comparative analysis against reference datasets will ensure that methylation patterns align with known biological mechanisms. Source Data Verification: * Methylation profiles will be cross-checked with publicly available databases to ensure consistency with previous epigenome-wide association studies. * Patient demographic and clinical data will be verified against electronic medical records to ensure accuracy. Handling Missing Data: * If methylation data for specific CpG sites is missing or below detection thresholds, these sites will be excluded from analysis. * Missing clinical data will be handled using multiple imputation methods to minimize bias. Sample Size Justification Cohort Composition: * 61 aSAH patients (32 with DIND, 29 without DIND). * 61 matched healthy controls from GSE246337. The relatively small sample size was chosen based on previous EWAS studies in aSAH Limitations and Future Directions Tissue-Specific Limitations: o The study uses peripheral blood DNA, which may not fully reflect methylation changes in the cerebral vasculature. Future studies should validate findings in the arterial wall of the affected vessel using low-mortality rat models or cerebrospinal fluid. DIND Subgroup Limitations: o Due to the relatively small sample size, findings related to DIND will require external validation in larger cohorts. Conclusion This study aims to provide novel insights into epigenetic changes associated with aSAH and DIND. By identifying differentially methylated CpG sites, the study seeks to improve biomarker discovery for early risk stratification and potential therapeutic targets in aSAH patients.