Background The B-cell chemoattractant CXCL13 is a key factor in the pathology of synovitis in rheumatoid arthritis (RA). Local expression of the chemokine is indeed involved in the formation and maintenance of synovial ectopic lymphoid structures . Recently, elevated circulating CXCL13 has emerged as a candidate marker of worst radiographic and ultrasonographic outcomes [2,3]. However, whether variability in expression levels of CXCL13 in the target tissue, i.e. the synovial membrane, is associated with different biomolecular and clinical patterns of the disease is unknown. Furthermore, to which extent circulating CXCL13 reflects synovial synthesis and specific markers of synovitis remains to be clarified.
Objectives To analyse the biomolecular and clinical features of RA in relationship with synovial and serum expression of CXCL13.
Methods Synovial samples from 30 RA patients were evaluated by qPCR and IHC. CXCL13 mRNA was analysed in relationship with cellular and molecular markers of inflammation, lymphocyte infiltration and activation, bone remodeling and clinical parametres of disease activity and severity. Circulating CXCL13 protein levels were assessed in paired serum samples by colorimetric ELISA.
Results Synovial CXCL13 mRNA was mildly correlated with histologic markers of inflammation and pro-inflammatory cytokine expression, with significant differences being detected for sublining macrophages (rho 0.51, p=0.03) and IL-1β (rho 0.55, p<0.01) only. Accordingly, low correlation was found with the ESR but not with other parametres of systemic disease activity. In contrast, the association with the degree of B cell, T cell and plasma cell infiltration as well as with markers of lymphocyte activation (IFN-γ, IL-2 and AID, the enzyme required for immunoglobulin affinity maturation and class switching) was highly significant (rho>0.8, p<0.001 for all). Patients with the highest levels of synovial CXCL13 mRNA were more frequently ACPA-positive (70.5% vs 30.8%, p=0.04). Furthermore, the frequency of radiographic erosions was increased in the high CXCL13 expression group (86.7% vs 46.7%, p=0.02), irrespective of disease duration and serum autoantibodies. Accordingly, synovial CXCL13 and the local RANKL/OPG ratio significantly co-varied (rho 0.52, p<0.01). In paired serum samples, the levels of CXCL13 protein positively correlated with synovial CXCL13 mRNA (rho 0.55, p=0.02). Circulating CXCL13 was associated with markers of local and systemic inflammation. However, differently from acute phase reactants, high serum CXCL13 (categorised on the basis of median values) was further associated with significant elevations in synovial B cell (p=0.02) and plasma cell (p=0.03) infiltration, expression of AID (p=0.04) and the RANKL/OPG ratio (p=0.01), higher ACPA titres (p=0.04) and increased frequency of erosive disease (p=0.02).
Conclusions Elevated synovial and serum levels of CXCL13 in RA are related to each other and associate with an immunologically-active pattern of synovial inflammation, not strictly related to increased disease activity, but coupled with clinical and molecular signs of disease severity.
Corsiero E, et al. Immunol Lett 2012;145:62-7.
Meeuwisse CM, et al. Arthritis Rheum 2011;63:1265-73.
Bugatti S, et al. Arthritis Res Ther 2012;14:R34.
Disclosure of Interest None Declared