Failed resolution as a mechanistic basis of difficult-to-treat rheumatoid arthritis: stromal-myeloid crosstalk in the synovium and emerging resolution-restoring therapies.

Modern disease-modifying therapy has transformed rheumatoid arthritis. Even so, many patients cycle through biological and targeted synthetic drugs of distinct mechanisms without achieving sustained remission. This is the difficult-to-treat phenotype defined by the European Alliance of Associations for Rheumatology. We argue that in a large fraction, refractory disease is not simply inflammation that is too strong or too drug-resistant. It is better understood as an active failure of the program meant to end inflammation. Single-cell and spatial dissection of the synovium lets us recast the rheumatoid joint as a niche of failed resolution. In this niche, imprinted pathological fibroblasts and a macrophage compartment stripped of its pro-resolving, efferocytic subsets both work against resolution. Their reciprocal, largely feed-forward signaling, we propose, assembles a self-sustaining stromal-myeloid circuit: a "failed-resolution switch". Lipid-mediator, metabolic, and epigenetic determinants reinforce it, and synovial microarchitecture stabilizes it. A switch built from redundant, mutually reinforcing arms would explain the therapeutic ceiling in refractory disease. Blocking any single mediator can lower inflammatory output while leaving resolution unrestored. Because the framework describes failed resolution rather than every route to refractoriness, it applies most directly to the persistent-inflammatory subset of difficult-to-treat disease. For that subset, it defines a measurable mechanistic state that can be stratified and which is open to intervention. It also motivates resolution-restoring strategies: pro-resolving mediators, macrophage reprogramming, and fibroblast- and crosstalk-directed interventions. Early clinical evidence, from the instructive failure of the granulocyte-macrophage colony-stimulating factor (GM-CSF) blockade to a proof-of-concept signal for programmed cell death protein 1 (PD-1) agonism, points toward reversing the disease rather than only suppressing it.