Evaluating Polygenic Risk Scores for Multiple Sclerosis: A Comparative Study Across European and African American Populations

Multiple sclerosis (MS) is a chronic autoimmune disorder characterized by inflammation and demyelination in the central nervous system, leading to a range of neurological symptoms and disabilities. The etiology of MS is complex, involving both genetic and environmental factors. Advances in genomic technologies have significantly contributed to our understanding of the genetic architecture of MS, revealing that polygenic inheritance plays a crucial role in the disease's development. To date, 333 single nucleotide polymorphisms (SNPs) have been associated with MS risk, mostly contributing modest increases individually. However, their cumulative genetic burden can be quantified using Polygenic Risk Scores (PRS), which aggregate the effects of numerous genetic variants into a single score, predicting disease risk.

This study aimed to determine whether a PRS for MS, developed primarily from European ancestry data, could effectively stratify individuals of non-European ancestries into distinct MS risk categories. Given the underrepresentation of non-European populations in genetic research, this study provides an essential assessment of PRS utility in diverse groups, thereby contributing to more equitable and accurate genetic risk assessment in MS.

A cross-sectional genetic study was conducted within the All of Us Research Program, which aims to enroll an ethnically diverse group of 1 million Americans. Data from 413,457 participants were analyzed, with 173,153 meeting the criteria for inclusion (having both whole genome sequencing [WGS] and electronic health record [EHR] data). Participants were classified into genetic ancestry groups: European, African, and Latino/admixed American (L/A).

Genomic Data and PRS Calculation
WGS data were centrally quality controlled using standardized pipelines. Ancestry was determined via principal components analysis. The PRS was computed using 282 independent genetic risk variants associated with MS, based on allele counts and their reported effect sizes.

MS diagnosis was ascertained using ICD-10 and SNOMED codes. Participants were divided into quintiles based on their PRS. Multivariable logistic regression models, adjusted for age, sex, and genetic principal components, were used to evaluate the relationship between PRS and MS risk.

Cohort Characteristics
Out of the 173,153 participants, the distribution of genetic ancestry was as follows: 55.4% European, 21.8% African, and 18.7% L/A. To ensure equal population sizes for comparison, 32,428 participants were randomly sampled from each group.

Unadjusted Analysis
The prevalence of MS was 1.0% in the European group, 0.56% in the African group, and 0.46% in the L/A group. In Europeans, the proportion of MS cases increased significantly across PRS quintiles. However, in Africans, the increase was not statistically significant.

Adjusted Analysis
In Europeans, the highest PRS quintile was associated with a 141% increase in MS risk compared to the lowest quintile (Odds Ratio [OR]: 2.41). In the L/A population, the highest PRS quintile corresponded to a 156% increase in MS risk (OR: 2.56). Conversely, in the African population, the PRS did not significantly stratify MS risk.

This study highlights significant disparities in PRS utility across different genetic ancestries. The PRS effectively stratified MS risk in European and L/A populations but not in African populations. This discrepancy may stem from the limited representation of African populations in existing genome-wide association studies (GWAS), leading to PRS that are less applicable to this group. Additionally, differences in genetic architecture and gene-environment interactions may contribute to these disparities.

The findings underscore the necessity for tailored genetic risk prediction tools and inclusive genetic research that considers diverse populations. Large, diverse datasets like the All of Us Research Program are invaluable for developing inclusive genetic risk prediction models and advancing personalized medicine for all populations.

Conclusion

While the PRS for MS is effective in European and L/A populations, it is less effective for African ancestry individuals. This emphasizes the urgent need for inclusive genetic research and the development of ancestry-specific PRS to ensure equitable benefits from genetic risk prediction and personalized healthcare. Future studies should focus on expanding genetic diversity in research cohorts, improving the representation of underrepresented populations in GWAS, and refining PRS methodologies to enhance their applicability across all populations.

References:
Rivier, C. A., Payabvash, S., Zhao, H. Y., Hafler, D. A., Falcone, G. J., & Longbrake, E. E. (2024). Differential Results of Polygenic Risk Scoring for Multiple Sclerosis in European and African American Populations. medRxiv, 2024-06.