Genetics Takes the Lead: How Polygenic Risk Scores Are Transforming Our Understanding of Multiple Sclerosis
Multiple sclerosis (MS) — a complex, immune-mediated disease of the central nervous system — affects nearly three million people worldwide. Despite decades of research, predicting who develops MS and how the disease will progress has remained elusive. But a new study published in Brain (2023) by Hengameh Shams and colleagues at the University of California, San Francisco, and collaborators across Europe and the U.S. offers a major leap forward.
Using cutting-edge genomics and statistical tools, the team developed powerful polygenic risk scores (PRS) that quantify a person’s inherited susceptibility to MS — and even link this genetic risk to subtle brain changes detectable years after diagnosis.
What Is a Polygenic Risk Score (PRS)?
A PRS is a single number that sums up the combined effects of millions of small DNA variations associated with a disease. Unlike single-gene disorders, complex diseases such as MS arise from the interaction of hundreds of genes and environmental factors. A PRS provides a snapshot of one’s genetic liability — not a destiny, but a probability.
Shams et al. leveraged the largest multiple sclerosis genome-wide association study (GWAS) to date, encompassing over 47,000 MS cases and 68,000 controls, to build the most accurate PRS model yet.
The Study: Building and Testing the MS-PRS
The researchers constructed their PRS using 8.6 million genetic variants from the International Multiple Sclerosis Genetics Consortium dataset. Advanced machine learning methods (notably LDPred2) were used to fine-tune the model and account for complex genetic correlations across the genome.
They tested the model in three independent European-ancestry datasets:
UK Biobank (UKBB): ~253,000 individuals
Kaiser Permanente Northern California (KPNC): A well-characterized case-control cohort
UCSF EPIC: A longitudinal imaging cohort with 10 years of follow-up
The results were striking.
Key Findings: A Powerful Genetic Signal
Strong predictive performance:
UK Biobank: AUC = 0.73 (95% CI: 0.72–0.74)
Kaiser Permanente: AUC = 0.80 (95% CI: 0.76–0.82)
(An AUC of 0.8 means the model can correctly distinguish MS cases from controls 80% of the time — impressive for a complex disease.)
Top 10% genetic risk:
UK Biobank: 5× higher odds of MS
KPNC: 15× higher odds of MS
compared to the middle of the risk distribution.
Earlier onset and stronger disease activity:
People with higher PRS developed MS earlier and experienced more frequent relapses over time.
Beyond Prediction: Linking Genes to Brain Changes
One of the most intriguing aspects of the study was connecting genetic risk to brain imaging phenotypes.
In the UCSF-EPIC cohort, higher polygenic scores were associated with accelerated thalamic atrophy — a hallmark of neurodegeneration in MS — over a 10-year period.
While causation remains uncertain (Mendelian randomization analyses suggested shared genetic influences rather than direct causality), the overlap between genetic susceptibility and brain tissue loss underscores how inherited risk can manifest as measurable structural damage.
The Immune System Connection: Pathways That Drive Risk
When the researchers dissected PRS into pathway-specific scores, they found enrichment in immune-related signaling cascades — notably:
T cell receptor and MHC class II antigen presentation
Interleukin-12 and interferon gamma signaling
Complement activation
Cell adhesion and extracellular matrix remodeling
Interestingly, pathways related to viral infection, particularly those involving Epstein–Barr virus (EBV) response, also showed significant association. This aligns with growing evidence that EBV plays a causal role in triggering MS in genetically predisposed individuals.
Family Patterns: Inherited Risk in Action
By studying 35 multicase families, the team demonstrated that PRS could capture heritable risk patterns.
Affected siblings consistently had higher PRS than their unaffected siblings, and families where one parent had a high PRS showed increased transmission of risk to offspring.
These familial findings reinforce the polygenic nature of MS and suggest that PRS may one day help identify at-risk relatives for early monitoring or preventive interventions.
Clinical Implications: From Research to Real-World Use
Including PRS in models alongside traditional risk factors — such as Epstein–Barr virus exposure, smoking, and obesity — improved risk prediction by up to 26%.
This integrated approach could help clinicians stratify individuals by risk, prioritize high-risk patients for closer follow-up, and tailor preventive strategies, especially in those with family history or early symptoms.
The authors note that PRS could become a valuable screening tool in adolescence or early adulthood — the critical window before typical MS onset (ages 20–40). However, they also emphasize the importance of extending studies to non-European populations, where current PRS models perform poorly due to limited representation in genetic databases.
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:
Shams, H., Shao, X., Santaniello, A., Kirkish, G., Harroud, A., Ma, Q., ... & Oksenberg, J. R. (2023). Polygenic risk score association with multiple sclerosis susceptibility and phenotype in Europeans. Brain, 146(2), 645-656.
