Decreased Retinal Ganglion Cell Density in Multiple Sclerosis Revealed by Adaptive Optics Optical Coherence Tomography
- Authors:
- Center:
-
Contributing OfficeCenter for Devices and Radiological Health
Abstract
Multiple Sclerosis (MS) is a debilitating autoimmune disease characterized by lesions found in different regions of the central nervous system (the brain and spinal cord) due to inflammation caused by overactive immune cells. MS also manifests in the retina, another CNS tissue, in which optic nerve pathology (such as optic neuritis) and neurodegenerative processes can affect retinal cell layers. The impact of these different pathologies has been clinically evaluated using optical coherence tomography (OCT), which provides noninvasive cross-sectional imaging of retinal layers with high axial resolution. Adaptive optics (AO) is a technology that can enhance OCT to increase the lateral image resolution. AO dynamically measures optical distortions caused by the eye and corrects these distortions with a deformable mirror. We used AO-OCT to investigate the differences between the retinas of participants with MS compared with healthy volunteers. Specifically, this study targets retinal ganglion cells (RGCs) which are found in the inner retina and are the cells that relay information to the brain for further processing. RGCs can experience similar effects from MS as does the grey matter in the brain. We hypothesized that RGC count will decrease with progression of MS as well as with other retinal pathologies such as optic neuritis. One eye from eight participants, including seven MS people and one healthy volunteer, were scanned using the FDA AO-OCT system. Each participant was scanned in seven different areas of the retina, three temporal and three nasal to the fovea at 4o superior, 0o, and 4o interior as well as the fovea. The locations were designed to capture RGC regional differences in MS patients, which in previous reports is distributed in a horseshoe pattern that preferentially affects the nasal macula. For each location on the retina, 30 AO-OCT videos were taken, each video containing 10 volumes and each volume containing 300x300x1024 pixels. Acquired data were first reconstructed from spectrum data to OCT images. The volumes were registered in three-dimensions, flattened to the inner limiting membrane, and averaged for retinal ganglion cell layer (GCL) and retinal nerve fiber layer (RNFL) quantification. A machine learning algorithm was then implemented for automated GCL soma counting followed by manual review and correction to be compared among the participants with MS and healthy volunteer. Further analysis includes nerve fiber bundle and macrophage counting and comparison between groups. Our results showed among people with MS, those with history of retinal disease have lower counts of RGCs. Furthermore, RGC density correlated with disease severity. For example, in two participants with MS, one participant who previously suffered severe optic neuritis was found to have a significantly lower RGC density (413 cells/mm2 vs 16,040 cells/mm2). Further quantitative comparison of nerve fiber bundle counting revealed that the subject with lower RGC density had loss of RNFL. Full quantification from the eight participants will be presented further in the poster. GCL and RNFL longitudinal quantification powered by AO-OCT may eventually help accelerate the development of therapy and treatment for patients with MS.