Immune–hemostatic crosstalk in diabetic retinopathy: amplifiers, secondary signatures, and evidence boundaries

Diabetic retinopathy (DR) develops within a neurovascular environment in which inflammatory and hemostatic processes intersect, but the abnormalities reported across these systems do not carry equal causal weight. This Review evaluates where immune–hemostatic crosstalk contributes to retinal injury and where it more likely reflects barrier failure, systemic diabetes, or advanced disease. We assess temporal precedence, retinal localization, experimental directionality, interventional support, replication, and agreement with human observations. Among the selected mechanisms reviewed here, endothelial activation and adhesion-dependent leukostasis have the strongest experimental support for selected early retinal vascular endpoints, although their contribution to persistent human retinal nonperfusion remains unresolved. Nonperfusion, hypoxic stress, vascular endothelial growth factor (VEGF) signaling, blood–retinal barrier dysfunction, and neuroglial injury interact rather than follow a fixed sequence. Fibrinogen/fibrin is most convincingly placed downstream of barrier failure, where it can exert secondary inflammatory effects. Neutrophil extracellular traps have context-dependent actions across retinal pathological states, whereas most platelet findings remain secondary/systemic signatures. Human VWF–ADAMTS13 evidence is predominantly associative, whereas ADAMTS13 supplementation and suppression have both shown protective effects on selected retinal endpoints in non-equivalent experimental models. Thrombin–PAR1 signaling is an emerging mechanism with incomplete human validation. This evidence-stratified synthesis therefore identifies experimentally supported amplifiers without recasting DR as a unified immunothrombotic disease.