Macular thickness is associated with whole brain grey matter volume in a mixed cohort of healthy controls and patients with neurodegenerative disease

Abstract Background Retinal atrophy in CNS neurodegenerative disease (NDD) has been reported by a growing literature base, suggesting that retinal changes may mirror, or even precede, neurodegeneration in the brain. However, understanding of how retinal and brain atrophy are associated in NDD is limited. In a large cohort of healthy aging controls (HC) and patients with confident diagnoses of NDD, we investigated if neuronal tissue loss in the unmyelinated retina is associated with whole brain grey (GM) and white (WM) matter volumes. Method We performed optical coherence tomography (OCT) within six months of structural brain imaging (MRI) in 373 patients with NDD (256 frontotemporal lobar degeneration spectrum (FTLD), 87 Alzheimer’s Disease (AD), 16 Lewy Body Disease (LBD), 12 prion disease, 2 with mixed etiology; age [years, mean±SD]:65.6±8.8, gender [male %]:51%, clinical dementia rating (CDR) [median (IQR)]:1 (0.5)) and 117 HC (age:73.8±9.5, gender:48%). 123 subjects (81 FTLD, 20 AD, 5 prion, 17 HC) had at least one follow up exam with both brain and retinal imaging (follow up time[years, median (IQR)]:1 (0.5)). Linear models correcting for age and total intracranial volume (TIV) were used to investigate the relationship between measurements of retinal thickness (µm), and whole brain GM and WM volumes (mm 3 ). Associations between annualized rate of change were modeled for patients with follow up, and were not corrected for TIV. Result Cross sectionally, macular thickness was associated with GM (0.62±0.20, p = 0.003) and WM (0.30±0.14, p = 0.030). Inner plexiform layer (IPL; 2.00±0.98, p = 0.038) and outer nuclear layer (ONL; 1.10±0.40, p = 0.005) were associated with GM, but not WM. Annualized rate of IPL (4.1±1.7, p = 0.016) and ONL (1.1±0.5, p = 0.034) change were associated with annualized rate of GM change, but not WM. In FTLD patients alone, rates of macular (2.7±0.6, p<0.001) and GCL (6.1±2, p = 0.005) change were also associated with GM change. Conclusion Whole macular thickness and structurally relevant sublayers representing neuron cell bodies (GCL, ONL) and dendritic arbors (IPL, ONL), were associated with GM, but not WM volume changes in a cohort of HC and NDD, with overrepresentation of FTLD. While findings may be limited to FTLD, they support retinal imaging as a biomarker for CNS neurodegeneration.