Genome-Wide Study of Early Disability Severity in Multiple Sclerosis

Background and Objectives: Multiple sclerosis (MS) is a chronic neurological disorder characterized by demyelination and neuroaxonal injury. MS genetic susceptibility and severity long after disease onset is well characterized, but the genetic architecture of disability severity near disease onset is unclear. We performed the first genome-wide association study of early MS severity to identify causal loci, prioritize effector genes and associated pathways, and understand the relationships of early MS severity and various traits. Methods: The primary outcome was the first age-related MS severity score (ARMSS), an age-adjusted severity measure, in 7,209 individuals of Swedish discovery cohorts and independent Italian and multi-national European replication cohorts. Replicated loci underwent conditional analysis, fine-mapping, time-to-event MS phenotype association analyses, and multi-omics-based gene prioritization. Using the genome-wide summary statistics, we also conducted heritability analyses, polygenic risk score association with other MS phenotypes, gene-level and pathway analyses, cross-trait analyses with disease-related traits, Mendelian randomization, and drug-repurposing screening. Results: Two suggestive loci were replicated: rs1234317 at chromosome 1 (effect allele T: beta = 0.100, SE = 0.0189, P = 1.08E-07) and rs929769 at chromosome 5 (effect allele T: beta = -0.0956, SE = 0.0184, P = 1.95E-07). Fine-mapping supported rs929769 as a single-variant 95% credible set, and rs1234317 and rs6657136 as a two-variant 95% credible set with two-causal-SNP model having the highest posterior probability. The SNPs did not show a significant change in effect sizes when adjusted for heritable traits and exhibited attenuated associations with long-term MS disability progression. TNFSF4 was the top prioritized effector for rs1234317, TNFSF18 for rs6657136, and ADAMTS2 for rs929769. Early MS severity showed partial genetic overlap with MS susceptibility or late MS severity, but no significant correlation with a range of neurological, psychiatric, autoimmune, cognitive, and aging traits. Pathway analyses identified ion-channel regulation and folic acid metabolism, and Mendelian randomization suggested nominal bidirectional associations with educational attainment. Drug screening nominated several drugs that can be repurposed. Discussion: This study identifies candidate causal variants of early MS severity and their effector genes: TNFSF4, TNFSF18, and ADAMTS2. It provides insights into the genetic architecture and therapeutic measures for the early phase of MS severity.