Improved classification of rheumatoid arthritis with a score including anti-acetylated ornithine antibodies

Abstract

The presence of rheumatoid factor (RF) or anti-cyclic citrullinated peptide (anti-CCP) autoantibodies contributes to the current rheumatoid arthritis (RA) classification criteria. These criteria involve stratification on antibody levels, which limits reproducibility, and underperform in the RA patients without RF and anti-CCP. Here, we have explored if two anti-acetylated peptide antibodies (AAPA), anti-acetylated lysine (AcLys) and anti-acetylated ornithine (AcOrn), could improve the performance of the current criteria. The analysis was done in 1062 prospectively-followed early arthritis (EA) patients. The anti-AcOrn were more informative than the anti-AcLys, the conventional RA antibodies and the anti-carbamylated protein antibodies. The anti-AcOrn produced a classification that did not require antibody levels and showed improved specificity (77.6% vs. 72.6%, p = 0.003) and accuracy (79.0% vs. 75.8%, p = 0.002) over the current criteria. These improvements were obtained with a scoring system that values concordance between anti-AcOrn, RF and anti-CCP. No significant gain was obtained in sensitivity (80.2% vs. 78.8%, p = 0.25) or in improving the classification of the RA patients lacking RF and anti-CCP, although the anti-AcOrn ranked first among the analysed new antibodies. Therefore, the anti-AcOrn antibodies could contribute to the improvement of RA classification criteria by exploiting antibody concordance.