Macrophages in non-infectious uveitis: cell origin, disease stage, evidence provenance, and therapeutic maturity.

Non-infectious uveitis (NIU) comprises a heterogeneous group of immune-mediated disorders in which myeloid cells influence leukocyte entry, inflammatory amplification, tissue injury, resolution, and relapse. The conventional dichotomy of classically activated (M1) and alternatively activated (M2) macrophages does not adequately capture this functional diversity. This narrative review organizes the evidence along three complementary dimensions: cell origin, disease stage, and evidence provenance; therapeutic maturity is evaluated separately. Experimental data support distinct roles for resident retinal microglia, retinal perivascular macrophages, vitreal hyalocytes, uveal macrophages, and recruited monocyte-derived macrophages. Microglial major histocompatibility complex class II signaling appears to amplify locally restimulated T-cell responses, whereas depletion with a broad colony-stimulating factor 1 receptor inhibitor cannot be interpreted as microglia-specific. Perivascular macrophages are positioned to regulate antigen presentation and leukocyte entry at the blood-retinal barrier, but their contribution to human NIU remains poorly defined. Recruited monocyte-derived macrophages contribute to inflammatory amplification in experimental autoimmune uveitis (EAU), while reparative and persistent states remain stage dependent. Human ocular fluid and single-cell studies provide direct observations of inflammatory myeloid states but rarely establish ontogeny or causality. In ankylosing spondylitis-associated acute anterior uveitis, cluster of differentiation 14-positive and intercellular adhesion molecule 1-high (CD14+ICAM1-high) monocytes showed trained-immunity-like transcriptional, functional, epigenetic, and metabolic features, with a computationally inferred transition toward macrophage states in aqueous humor. Clinically, tumor necrosis factor blockade provides the strongest therapeutic evidence. Janus kinase inhibition has randomized phase 2 support, whereas interleukin-6, interleukin-1, P2X purinoceptor 7, chemokine, inflammasome, metabolic, and reparative strategies remain less mature or are not myeloid specific. Progress will require phenotype- and stage-matched studies that combine longitudinal ocular and blood sampling with cell-resolved biomarkers, target engagement measures, and experiments involving delayed treatment in complementary animal models.