Disentangling patient, disease, and treatment effects: proteomic and metabolomic differences in early versus established psoriatic arthritis

Abstract

Objectives: Psoriatic arthritis is a heterogeneous inflammatory disease with unclear pathobiology. Biological mechanisms may change over time with disease progression or treatment and may correlate with specific characteristics of PsA. This study aimed to identify differences in proteomic and metabolomic profiles between DMARD-naïve (DN) and DMARD-failing (DF) PsA patients. Methods: In the TOFA-PREDICT trial discovery cohort, baseline data from 40 DN and 40 DF PsA patients with active disease (all fulfilling the Classification Criteria for Psoriatic Arthritis) were analysed. The aim was to identify differences in proteomic and metabolomic profiles using univariate (Welch’s t-test) and multivariate (XGBoost and sPLS-DA) approaches; potential mediation of these differences by patient, disease and treatment characteristics; and biological functions of identified markers through enrichment analysis. Results: A total of 36 proteins and 25 metabolites differentiated DN from DF PsA patients. Mediation analysis indicated that most differences are independent of clinical characteristics or current use of medication. Eleven proteins and two metabolites were (partially) mediated by the Psoriasis Area and Severity Index, CRP level, ESR, tender joint count 68, disease duration or smoking. Enrichment analysis highlighted an overrepresentation of immune and inflammatory proteins, while metabolomic markers were predominantly glycerophospholipids, with no significant pathway enrichment. Conclusion: This study reveals distinct proteomic and metabolomic profiles between DN and DF PsA patients and highlights markers potentially involved in PsA pathogenesis. These findings emphasise the biological heterogeneity of PsA patients and the need to consider the disease phase in proteomic and metabolomic analysis.