Patient-Derived Synovium-on-a-Chip Recapitulates Disease and Treatment Response and Permits Study of Resistance
Rheumatoid arthritis (RA) is a chronic inflammatory disease with substantial interpatient heterogeneity in synovial pathology and therapeutic response. Reliable methods to predict drug efficacy for a given patients remain a critical unmet need. We developed a vascularized, immunocompetent synovium-on-a-chip platform that integrates patient-derived fibroblast-like synoviocytes (FLS) with matched peripheral blood leukocytes to provide a physiological human model of RA. The platform reconstructed the lining-sublining synovial architecture, recapitulated key features of RA inflammation, and exhibited transcriptional profiles resembling those of patient synovial tissue. Across patients spanning low, mixed, and high inflammatory endotypes, stromal-vascular organization was associated with distinct immune cell positioning and activation patterns, and on-chip immune phenotypes correlated with histological joint inflammation scores. We then tested infliximab, tocilizumab, and tofacitinib in parallel and integrated cytokine secretion, extracellular matrix remodeling, immune cell survival and immune cell recruitment into a composite treatment response score. On-chip scores for clinically matched treatments correlated with clinical improvement measured by {Delta}DAS28-ESR (Pearson r = 0.844). To investigate the cellular basis of heterogeneous drug responses, single-cell RNA sequencing identified SFRP1 remodeling FLS that persisted under treatment and exhibited high inflammatory programs in poorly responding chips. Cell-cell communication analysis linked reduced FN1-CD44 signaling toward SFRP1 remodeling FLS to higher NF-{kappa}B activity and drug resistance. We tested the functional relevance of this finding in chips derived from an independent TNFi-nonresponsive patient, where fibronectin supplementation reduced IL-6 secretion and restored the composite treatment response. Together, these findings support patient-derived synovium-on-a-chip as a platform for individualized drug evaluation and identify ECM-mediated niche signaling as a potential therapeutic target for overcoming drug resistance in RA.