Biomarkers of treatment response in rheumatoid arthritis: from conventional markers to tissue immunophenotype

Rheumatoid arthritis (RA) is a biologically heterogeneous immune-mediated disease characterized by substantial variability in therapeutic response. Despite the availability of multiple conventional synthetic, biologic, and targeted synthetic disease-modifying antirheumatic drugs (DMARDs), many patients fail to achieve adequate disease control or experience secondary loss of efficacy, underscoring the need for predictive biomarkers that can guide treatment selection. This narrative review was based on a structured literature search of PubMed/MEDLINE, Scopus, Web of Science, and Google Scholar, covering publications from January 2000 to June 2026, with earlier landmark studies included when relevant. Literature selection followed PRISMA-informed principles, although the review was not designed as a formal systematic review. Unlike previous reviews that mainly catalogue RA biomarkers by analytical platform, drug class, or clinical use, this review integrates conventional and emerging biomarkers within a tissue-immunophenotype-centered framework. We critically evaluate clinical, serological, pharmacological, molecular, imaging, and tissue-based biomarkers according to biological plausibility, reproducibility, level of validation, clinical actionability, and translational readiness. Established markers such as rheumatoid factor, anti-citrullinated protein antibodies, acute-phase reactants, drug concentrations, and anti-drug antibodies remain clinically useful but provide incomplete insight into mechanism-specific therapeutic response. In contrast, synovial pathotypes, fibroblast and macrophage subsets, B-cell niches, tertiary lymphoid structures, single-cell and spatial omics, and ligand-receptor interaction networks offer a mechanistically richer view of treatment response and resistance. We conclude that precision medicine in RA will require integrated biomarker panels combining clinical, pharmacological, molecular, and synovial tissue data. The key future direction is the development of scalable, externally validated, and clinically interpretable models capable of assigning synovial endotypes and supporting mechanism-based therapeutic selection.