Genetic Risk Stratification in Optic Neuritis: Advancing Early Prediction of Multiple Sclerosis
Optic neuritis (ON) is a common inflammatory disorder of the optic nerve and frequently represents the first clinical manifestation of multiple sclerosis (MS). Despite its importance, accurately distinguishing ON cases that will later evolve into MS from those associated with other immune-mediated conditions remains challenging at initial presentation. This distinction is clinically critical, as MS-associated optic neuritis (MS-ON) and non-MS ON differ substantially in prognosis, therapeutic urgency, and long-term neurological outcomes. The article under discussion addresses this unmet need by evaluating whether a multiple sclerosis genetic risk score (MS-GRS), when combined with basic demographic factors, can improve prediction of future MS diagnosis following a first episode of optic neuritis.
Genetic Architecture of Multiple Sclerosis and Rationale for Risk Scoring
Multiple sclerosis is a highly polygenic autoimmune disease, with approximately 20% of its heritability attributable to common genetic variants identified through large-scale genome-wide association studies (GWAS). These studies have highlighted both strong human leukocyte antigen (HLA) class II associations and numerous non-HLA loci implicated in immune regulation. The aggregation of these variants into a composite genetic risk score provides a quantitative measure of inherited susceptibility. Prior work has shown that MS genetic risk scores can discriminate MS cases from controls, but their utility in predicting disease evolution among patients with undifferentiated optic neuritis had not been previously established.
Study Design and Cohort Integration
The investigators leveraged data from the UK Biobank, analyzing over 480,000 genotyped individuals to identify cases of optic neuritis, multiple sclerosis, and MS-associated optic neuritis. Importantly, the primary analysis focused on individuals presenting with undifferentiated ON and no prior MS diagnosis. The MS-GRS was constructed using 307 non-HLA variants and an interaction-aware HLA model, reflecting contemporary understanding of MS genetic architecture. To ensure robustness, findings were externally validated in two independent cohorts: Geisinger (United States) and FinnGen (Finland), enabling cross-population assessment of model performance.
Discriminatory Power of the MS Genetic Risk Score
Across all datasets, the MS-GRS was significantly higher in individuals with MS compared to healthy controls and those with non-MS ON. In the UK Biobank, the full MS-GRS demonstrated a receiver operating characteristic area under the curve (ROC-AUC) of approximately 0.72 for distinguishing MS cases from controls, improving further when combined with demographic covariates such as age and sex. Notably, the MS-GRS distribution for individuals with non-MS ON fell between that of healthy controls and MS-ON cases, suggesting partial genetic overlap and supporting its relevance for intermediate-risk stratification.
Prediction of Future MS Following Optic Neuritis
The central finding of the study is that MS-GRS independently predicts future MS diagnosis among patients presenting with undifferentiated optic neuritis. In multivariable Cox proportional hazards models, each standard deviation increase in MS-GRS was associated with a 1.29-fold increase in MS risk, even after adjusting for age and sex. When patients were stratified into quartiles of predicted risk, cumulative MS incidence ranged from approximately 4% in the lowest-risk group to over 40% in the highest-risk group during long-term follow-up. These results demonstrate that genetic risk meaningfully refines prognostic assessment beyond conventional clinical variables.
External Validation and Generalizability
Crucially, the predictive model derived from the UK Biobank replicated well in both the Geisinger and FinnGen cohorts, despite differences in MS prevalence, healthcare systems, and follow-up duration. In all cohorts, individuals in the highest predicted risk quartile consistently exhibited substantially greater rates of MS conversion compared to those in the lowest quartile. While performance metrics varied modestly across datasets, the preservation of risk gradients supports the model’s generalizability and suggests potential applicability across diverse clinical settings, particularly in populations of predominantly European ancestry.
Clinical Implications and Future Directions
This study provides compelling evidence that integrating a genetic risk score with simple demographic factors can enhance early MS risk stratification in patients presenting with optic neuritis. Clinically, such an approach could inform time-sensitive decisions regarding corticosteroid therapy, referral urgency, and early initiation of disease-modifying treatments. Nevertheless, important limitations remain, including underrepresentation of non-European ancestries and the absence of contemporaneous imaging or biomarker data. Future prospective studies, incorporating neuroimaging and broader ancestral diversity, will be essential before routine clinical implementation. Overall, the work represents a significant step toward precision medicine in neuro-ophthalmology and multiple sclerosis care.
Disclaimer: This blog post is based on the provided research article and is intended for informational purposes only. It is not intended to provide medical advice. Please consult with a healthcare professional for any health concerns.
References:
Loginovic, P., Wang, F., Li, J. et al. Applying a genetic risk score model to enhance prediction of future multiple sclerosis diagnosis at first presentation with optic neuritis. Nat Commun 15, 1415 (2024). https://doi.org/10.1038/s41467-024-44917-9
