Evaluating the quantity and spatial density of macrophage-like cells in patients with retinal vascular disease and healthy subjects via non-invasive retinal imaging

Here we investigate the spatial density and distribution of macrophage-like cells (MLCs) in ischemic retinal diseases, including branch retinal vein occlusion (BRVO), central retinal vein occlusion (CRVO), and proliferative diabetic retinopathy (PDR), compared to healthy controls. In this pilot investigation with prospective cross-sectional design, OCT angiography (OCTA) images were obtained from 20 eyes across four groups (BRVO, CRVO, PDR, controls). Using a standardized semi-automated image processing protocol, MLCs were identified and quantified over a 6 × 6 mm area temporal to the macula. Additionally, perfusion densities were measured, and analyses were performed to assess differences in MLC counts and their correlations with perfusion. Patients with BRVO, CRVO, and PDR exhibited significantly higher MLC counts compared to controls (p = 0.002 for all comparisons). Median MLC counts were lowest in controls (132; 95% CI, 63–180), followed by BRVO (382; 95% CI, 290–446), CRVO (688; 95% CI, 507–716), and highest in PDR patients (973; 95% CI, 805–999). A moderate-to-strong negative correlation was found between perfusion density and MLC count, suggesting increased MLC accumulation in areas of reduced perfusion in the superior vascular complex (p = 0.04), deep vascular complex (p < 0.01), as well as combined vascular complexes (p = 0.01). Here, we demonstrate that MLC density is significantly elevated in BRVO, CRVO, and PDR compared to healthy eyes and is inversely correlated with retinal perfusion. By applying consistent imaging and analysis across a wider peripheral retinal field, these findings highlight MLC quantification as a potential biomarker for disease severity and progression in ischemic retinopathies. Future investigations should explore whether modulating MLC responses could offer new therapeutic strategies to improve outcomes in these conditions.