Association of Polygenic Risk Scores With Radiographic Progression in Patients With Rheumatoid Arthritis.

ABSTRACT

OBJECTIVE: To investigate whether polygenic risk scores obtained using data from a genome-wide association study (GWAS) of rheumatoid arthritis (RA) susceptibility can be predictors of radiographic progression. METHODS: We constructed polygenic risk scores using GWAS summary data on associations of single-nucleotide polymorphisms with RA susceptibility. The polygenic risk scores were stratified into quintiles based on levels of significance (ranging from top quintile of polygenic risk scores to bottom quintile). In addition, change in the Sharp/van der Heijde score (SHS) of radiographic progression over the first 5 years after onset of RA was assessed. The change in SHS over 5 years was stratified according to quartiles, with the top quartile of change in SHS defined as severe radiographic progression (score change of >35 points) and the remaining 3 quartiles defined as nonsevere radiographic progression. Polygenic risk scores were assessed for their ability to predict the SHS status over 5 years in a training set (n = 500 RA patients) for selection of the best model, and in a testing set (n = 740 RA patients) for validation of the data. We evaluated the performance of the polygenic risk score as a predictor of severe radiographic progression in univariable and multivariable analyses with inclusion of other factors. RESULTS: Polygenic risk scores constructed from 43,784 single-nucleotide polymorphisms significantly differed between patients who experienced severe radiographic progression and those with nonsevere radiographic progression in both the training set (P = 0.0064) and the testing set (P = 0.017). Patients with polygenic risk scores in the top quintile had a higher risk of severe progression compared to those with polygenic risk scores in the bottom quintile (odds ratio [OR] 1.90, P = 0.0022), and the risk of severe radiographic progression was even higher when restricted to patients who were younger at disease onset (OR 5.06, P = 0.00038). The group with polygenic risk scores in the top quintile and the anti-citrullinated protein antibody (ACPA)-positive group had significantly higher proportions of patients with severe radiographic progression (P = 0.00052 and P = 0.0022, respectively) compared to the remaining groups. Multivariable analysis showed that polygenic risk score (P = 0.00019) as well as female sex (P = 0.0033), ACPA positivity (P = 0.0023), and body mass index (P = 0.024) were independent risk factors for severe radiographic progression. CONCLUSION: A polygenic risk score that is derived from GWAS data on RA susceptibility is associated with the level of severity of radiographic progression in patients with RA.