Longitudinal retinal microvascular changes and their association with white matter hyperintensities in neuromyelitis optica spectrum disorder.

BACKGROUND

Neuromyelitis optica spectrum disorder (NMOSD) is an autoimmune astrocytopathy characterized by recurrent inflammatory attacks within the central nervous system. Beyond optic neuritis, diffuse retinal and cerebral alterations may occur. Optical coherence tomography angiography (OCTA) enables the non-invasive assessment of retinal microvasculature and may provide insight into longitudinal retinal changes and their association with cerebral white matter hyperintensities (WMHs).

METHODS

In this longitudinal observational study, 45 aquaporin-4 antibody-positive NMOSD patients and 45 age- and sex-matched healthy controls underwent OCTA imaging. Radial peripapillary capillary (RPC) density was measured in the peripapillary region, while superficial vascular complex (SVC), deep vascular complex (DVC), and foveal avascular zone (FAZ) area were assessed in the macular region. Brain MRI was performed at baseline and at a 1-year follow-up, and WMH burden was rated using the Fazekas scale. Associations between OCTA metrics and WMHs were evaluated using regression models adjusted for age and disease duration, and longitudinal changes were analyzed using mixed-effects models.

RESULTS

At baseline, NMOSD eyes, including those without prior optic neuritis, showed significantly reduced RPC, SVC, and DVC densities and an enlarged FAZ area compared with controls (all p < 0.0001). Lower RPC density was associated with higher periventricular and total WMH burden (standardized β range: -0.45 to -0.52, all p < 0.01). Over 1 year, RPC and SVC densities declined further (both p < 0.0001), whereas DVC density and FAZ area remained stable. RPC density demonstrated the highest discriminatory performance in ROC analyses (AUC = 0.78).

CONCLUSION

OCTA reveals dynamic retinal microvascular alterations in NMOSD and a consistent association between reduced peripapillary capillary density and WMH burden. These findings are observational and do not imply causality, but they support the use of OCTA as a sensitive tool for monitoring retinal involvement in NMOSD.