Polygenic risk scores of rheumatoid arthritis associated with seropositivity and bone erosions in a Taiwanese population

Polygenic risk scores (PRS) are widely used to estimate disease risks and predict clinical outcomes. However, ethnic differences exist worldwide. We aimed to use PRS for predicting the development of rheumatoid arthritis (RA) in a Taiwanese population and investigate whether PRS of RA may be associated with structural damage. PRS was calculated via GWAS data from 2,042 RA patients and 7,950 controls from the Taiwan Precision Medicine Initiative. LDpred2, PLINK, and PRSice-2 models were used to evaluate RA susceptibility via AUC. Clinical factors, including rheumatoid factor (RF), anti-citrullinated protein antibody (ACPA), medication, and bone erosion identified by X-ray or ultrasound, were compared across PRS quartiles. PRS derived from 97,396 SNPs via LDpred2 had the highest AUC. Participants in the highest quartile of RA-PRS had the highest proportions of rheumatoid factor (RF) and anti-citrullinated protein antibody (ACPA) positivity (74.8% and 65.0%, respectively), bone erosion (86.4%), and use of bDMARDs or tsDMARDs (42.3%) compared to the other quartiles. PRS remained a significant predictor of bone erosion, particularly in patients under 60 years of age, even after adjusting for RF, ACPA, and treatment. PRS is associated with seropositivity, erosive bone disease, and need for advanced therapy among Taiwanese patients with RA.