Neurodegeneration in Glaucoma: Microstructural Magnetic Resonance Imaging Evidence Within and Beyond the Visual Pathway

Purpose: To investigate structural and microstructural brain changes in glaucoma using multimodal magnetic resonance imaging across primary, secondary, and higher-order visual brain regions, and their associations with glaucoma diagnosis, optical coherence tomography-derived retinal nerve fiber layer (RNFL) thickness, ganglion cell layer (GCL) thickness, and/or IOP. Methods: From the UK Biobank, we identified glaucoma cases (n = 1465) and 10-fold age- and sex-matched controls (n = 14,650). Magnetic resonance imaging modalities comprised T1-weighted structural, diffusion tensor imaging, and neurite orientation dispersion and density imaging. Associations with glaucoma status and ophthalmic measures (RNFL, GCL, and IOP) were assessed using regression models adjusted for age, sex, polygenic risk score, and Townsend Deprivation Index, with false discovery rate correction. Results: Glaucoma was associated with reduced gray matter volume in primary visual regions (lateral geniculate nucleus, optic chiasm, intracalcarine cortex, and occipital pole) and diffusion tensor imaging/neurite orientation dispersion and density imaging abnormalities in the posterior thalamic radiation (all P < 0.001). Glaucoma was also associated with secondary regions (lateral occipital cortex, lingual gyrus, and occipital fusiform gyrus) and microstructural changes in the inferior fronto-occipital and inferior longitudinal fasciculus (all P < 0.001). Higher-order and supporting regions were also associated with glaucoma, including the right putamen and paracingulate gyrus (P < 0.05). The RNFL (P < 0.05) and GCL (P < 0.01) correlated linearly with most primary visual regions, whereas the IOP showed no significant associations. Conclusions: Glaucoma is associated with structural and microstructural brain changes beyond the eye to primary, secondary, and higher-order brain regions. These changes correlate with RNFL and GCL thinning but not IOP. Involvement of the occipital pole is consistent with the plausibility of trans-synaptic degeneration in glaucoma.