Potential novel biomarkers of disease activity in rheumatoid arthritis patients: CXCL13, CCL23, transforming growth factor α, tumor necrosis factor receptor superfamily member 9, and macrophage colony-stimulating factor

Objective To determine whether the plasma levels of a range of inflammatory proteins have utility as biomarkers of disease activity in rheumatoid arthritis (RA) patients. Methods Plasma proteins (n = 163) were profiled in 44 patients with RA diagnosed according to the American College of Rheumatology 1987 criteria (22 with active and 22 with quiescent disease) and in 16 age- and sex-matched healthy controls. The utility of a subset of differentially expressed proteins as predictors of RA disease activity was investigated using partial least-squares discriminant analysis, and their response to therapeutic intervention was evaluated in plasma from an additional cohort of 16 patients with active RA treated with anti–tumor necrosis factor α (anti-TNFα). Results The protein profiling study identified 25 proteins that were differentially expressed in plasma samples from patients with active RA (P for the false discovery rate ≤ 0.01) compared with those with quiescent RA, including the previously described interleukin-6 (IL-6), oncostatin M, and IL-2, and the 5 less-established markers macrophage colony-stimulating factor (M-CSF), tumor necrosis factor receptor superfamily member 9, CCL23, transforming growth factor α, and CXCL13. Systemic levels of these 5 markers correlated with the C-reactive protein level, erythrocyte sedimentation rate, rheumatoid factor level, tender joint count in 68 joints, and Disease Activity Score in 28 joints (DAS28), and their combined plasma levels were shown to be good predictors of disease activity (κ = 0.64). In anti-TNFα–treated RA patients, plasma levels of CXCL13 were reduced after 1 and 7 days of therapy, and levels of CCL23, M-CSF, and CXCL13 showed a statistically significant positive correlation with the DAS28 score. Conclusion This exploratory study for biomarker discovery led to the identification of several proteins predictive of RA disease activity that may be useful in the definition of disease subphenotypes and in the measurement of response to therapy in clinical studies.