Relationship between Psychophysical Measures of Retinal Ganglion Cell Density and In Vivo Measures of Cone Density in Glaucoma

Purpose: Considerable between-individual variation in retinal ganglion cell (RGC) density exists in healthy individuals, making identification of change from normal to glaucoma difficult. In ascertaining local cone-to-RGC density ratios in healthy individuals, we wished to investigate the usefulness of objective cone density estimates as a surrogate of baseline RGC density in glaucoma patients, and thus a more efficient way of identifying early changes. Design: Exploratory cohort study. Participants: Twenty glaucoma patients (60% women) with a median age of 54 years and mean deviation (MD) in the visual field of -5 dB and 20 healthy controls (70% women) with a median age of 57 years and a mean MD of 0 dB were included. Methods: Glaucoma patients and healthy participants underwent in vivo cone imaging at 4 locations of 8.8 degrees eccentricity with a modified Heidelberg Retina Angiograph HRA2 (scan angle, 3 degrees). Cones were counted using an automated program. Retinal ganglion cell density was estimated at the same test locations from peripheral grating resolution acuity thresholds. Main Outcome Measures: Retinal cone density, estimated RGC density, and cone-to-RGC ratios in glaucoma patients and healthy controls. Results: Median cone-to-RGC density was 3.51: 1 (interquartile range [IQR], 2.59: 1-6.81: 1) in glaucoma patients compared with 2.35: 1 (IQR, 1.83: 1-2.82: 1) in healthy participants. Retinal ganglion cell density was 33% lower in glaucoma patients than in healthy participants; however, cone density was very similar in glaucoma patients (7248 cells/mm(2)) and healthy controls (7242 cells/mm(2)). The area under the receiver operator characteristic curve was 0.79 (95% confidence interval [CI], 0.71-0.86) for both RGC density and cone-to-RGC ratio and 0.49 (95% CI, 0.39-0.58) for cone density. Conclusions: Local measurements of cone density do not differ significantly from normal in glaucoma patients despite large differences in RGC density. There was no statistically significant association between RGC density and cone density in the normal participants, and the range of cone-to-RGC density ratios was relatively large in healthy controls. These findings suggest that estimates of baseline RGC density from cone density are unlikely to be precise and offer little advantage over determination of RGC alone in the identification of early glaucomatous change. (C) 2016 by the American Academy of Ophthalmology