Polygenic risk scores are associated with radiographic progression in patients with rheumatoid arthritis

Objective

To investigate whether polygenic risk score (PRS) using data from a genome-wide association study (GWAS) for rheumatoid arthritis (RA) susceptibility can be a predictor for radiographic progression.

Methods

We constructed the PRS using GWAS summary data for disease susceptibility to predict Sharp/van der Heijde score (SHS) changes in first five years from the onset (the top quartile of SHS changes was defined as severe progression and the remaining as non-severe progression). We selected the best model in a training set (n = 500) and validated it in a testing set (n = 740). We evaluated the performance of PRS in univariable and multivariable analyses with other factors to predict severe progression.

Results

PRS constructed of 43,784 SNPs significantly differed between severe and non-severe progression in both training (P =?0.0064) and testing sets (P =?0.017). The patients with the top quintile PRS had a higher risk for severe progression compared to those with the bottom quintile (odds ratio (OR) 1.90, P =?0.0022), which was higher when restricted to younger-onset patients (OR 5.06, P =?0.00038). The top quintile PRS and ACPAs positive groups had significantly higher proportion of patients with severe progression compared to the remaining groups (P =?0.00052, and 0.0022, respectively). Multivariable analysis showed that PRS (P =?0.00019) as well as sex (female) (P =?0.0033), ACPAs (P =?0.0023), and BMI (P =?0.031) were independent risk factors.

Conclusion

PRS using GWAS data for RA susceptibility is associated with radiographic progression.