IgA anti-CD74 autoantibodies are associated with treatment escalation in peripheral psoriatic arthritis

Psoriatic arthritis (PsA) often requires escalation from conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) to biologic therapy (bDMARDs), yet biomarkers guiding treatment decisions remain limited. Anti-CD74 autoantibodies have shown diagnostic potential in axial spondyloarthritis, but their relevance in PsA is insufficiently characterized.

To evaluate whether IgA anti-CD74 levels are associated with treatment escalation in peripheral PsA (pPsA) and to characterize their relationships with clinical and immunological parameters.

Serum samples from 171 PsA (127 pPsA, 44 axial PsA [axPsA]), 43 non-rheumatic disease controls (NRD), and 43 rheumatoid arthritis (RA) patients were analyzed. IgA anti-CD74 levels were measured by enzyme-linked immunosorbent assay (ELISA). Patients were stratified by disease duration and treatment exposure. Correlation analyses, receiver operating characteristic (ROC) curves, and logistic regression models were performed. A prospective subgroup of 53 early pPsA patients was followed to assess treatment initiation.

IgA anti-CD74 levels were elevated in PsA compared with NRD controls (median 13 vs. 6 U/mL, p < 0.0001), with similar levels in pPsA and axPsA and comparable values in RA, indicating an association with inflammatory disease rather than disease specificity. Anti-CD74 positivity (>15 U/mL) was observed in 31.5% of pPsA and 43.2% of axPsA versus 2.3% of NRD, independent of disease duration. Anti-CD74 levels were associated with treatment escalation in pPsA, with higher levels in bDMARD-treated patients (median 13.0 vs. 11.0 U/mL, p = 0.04). In multivariate analyses, anti-CD74 was independently associated with csDMARD (OR 1.113, p = 0.022) and bDMARD use (OR 1.052, p = 0.02). After false discovery rate (FDR) correction, anti-CD74 remained associated with serum IgA (q = 0.0008) and weakly with IgG (q = 0.0250), but not with C-reactive protein (CRP) or age. Longitudinal associations were not significant after FDR correction (csDMARD initiation: p = 0.047, q = 0.094; bDMARD initiation: p = 0.19, q = 0.1866), indicating these findings are exploratory.

IgA anti-CD74 levels are elevated in PsA and appear to reflect immunological activity not captured by CRP. Their independent association with treatment escalation in pPsA supports further evaluation as a biomarker candidate, although findings remain exploratory and require validation in larger longitudinal cohorts.