Risk of progression from undifferentiated arthritis to rheumatoid arthritis: the effect of the PTPN22 1858T-allele in anti-citrullinated peptide antibody positive patients

Abstract

Anti-citrullinated peptide antibodies (ACPA) and the C1858T missense single-nucleotide polymorphism (SNP) in the PTPN22 gene are both associated with the development of rheumatoid arthritis (RA). We investigated whether the combination of these two biomarkers yielded better test characteristics to predict progression from undifferentiated arthritis (UA) to RA compared with ACPA alone. A total of 394 individuals with UA from a Dutch population-based inception cohort were included in this study. At baseline, ACPA were measured and the PTPN22 C1858T and HLA-DRB1 genotypes determined. Progression to RA was monitored at 1 yr after entry into the cohort. A priori, UA patients had a 35% (95% CI 30-40%) risk of developing RA, which increased to 66% (95% CI 57-75%) in patients who were ACPA-positive. There was an additional, although non-significant (P = 0.34), increase in RA risk to 76% (95% CI 57-90%) when patients were positive for both ACPA and the PTPN22 1858T-allele. The area under the receiver operator characteristic curve increased from 0.68 for ACPA-status alone to 0.70 for the combination of ACPA-status and the PTPN22 C1858T polymorphism. In logistic regression analysis, ACPA predicted RA-development independent of PTPN22, while the PTPN22 polymorphism had no independent effect. In HLA-DRB1 shared epitope positive, ACPA-positive UA patients, ACPA-levels were significantly increased in PTPN22 1858T allele carriers compared with non-1858T carriers. In this Dutch cohort of UA-patients, the PTPN22 1858T allele does not markedly improve individual decision-making to predict RA-development over ACPA alone, but it is associated with higher ACPA-levels.