Exploratory Multi-Omics Profiling of Rheumatoid Arthritis Heterogeneity

Rheumatoid arthritis (RA) is a clinically heterogeneous disease with highly variable treatment responses despite widespread use of conventional and biologic disease-modifying therapies. Existing clinical and serological markers only partially explain this variability and are not routinely used to personalise drug choice. We hypothesised that integrated plasma proteomic- metabolomic profiling could capture dimensions of RA molecular heterogeneity that are invisible to standard phenotyping and highlight characteristic patterns of metabolic dysregulation. We profiled 384 proteins and 1,063 metabolites in 80 RA patients and 40 healthy controls. Unsupervised consensus clustering of the integrated omics matrix identified two stable molecular subtypes within the RA cohort. Anti-citrullinated protein antibody (ACPA) status, which in this cohort was evenly split overall (40 ACPA-positive, 40 ACPA-negative), was similarly distributed across the two subtypes, suggesting that the molecular groupings reflect heterogeneity orthogonal to conventional serology. Protein-metabolite association network analysis revealed a substantial reduction in association edges in RA versus controls, with pathway-specific effects: pronounced connectivity loss in amino-acid metabolism but paradoxical gains in inflammatory mediators including glutamate and sphingolipids. Preserved correlation strength among surviving edges is consistent with selective pathway disconnection rather than uniformly weakened coordination. These hypothesis-generating findings, limited by cross-sectional design, medication heterogeneity, and single-centre recruitment, suggest that RA comprises molecularly distinct subtypes not captured by standard serological categories and is associated with extensive metabolic network reorganisation. This multi-omics framework illustrates how network-level signatures could support future precision stratification efforts and motivates further investigation of metabolic network restoration as a potential therapeutic avenue.