Nonaxonal content of the circumpapillary retinal nerve fiber layer in experimental glaucoma

This study aimed to determine the region-specific distribution of nonaxonal cells and capillaries within the circumpapillary retinal nerve fiber layer (RNFL) in a nonhuman primate experimental glaucoma model.

Five rhesus macaques (age range: 4.5-9.0 years) with varying degrees of unilateral laser-induced experimental glaucoma were imaged with optical coherence tomography (OCT), euthanized and perfusion fixed for histological analysis. Eyes were enucleated, and six circumpapillary regions (60° apart), approximately 300-1000 µm from the optic disc margin, were dissected and prepared for serial block face scanning electron microscopy. A series of 400 images (0.098 µm/pixel) were obtained at 500 nm intervals to construct a volume of 200 × 200 × 200 µm. Inside the nerve fiber layer, nonaxonal nuclei and capillary lumen were manually outlined. The nearest vessel distance for each nonaxonal cell was determined as the Euclidean distance in 3-dimensional space from each nucleus center to the nearest capillary.

At endpoint, OCT circumpapillary RNFL thickness in the experimental eyes ranged between 49 and 99 µm (control: 103-111 µm). In control eyes, the distribution of nonaxonal cells at the six circumpapillary regions was nonuniform (F_5,24 = 4.05 and p < 0.01). However, the total cell count was similar between the two eyes (control: 858 ± 173, experimental: 960 ± 245, and p = 0.51), and for each region (p > 0.05). Although capillary volume was reduced in experimental eyes (p = 0.02), the capillary volume density was similar between control (0.58 ± 0.31%) and experimental eyes (0.58 ± 0.32% and p = 0.99). Correspondingly, the nonaxonal cells had a shorter vessel distance in experimental compared to control eyes (15.92 ± 4.89 μm, 18.61 ± 5.48 μm, and p = 0.05).

The number of nonaxonal cells and capillary volume density in the circumpapillary RNFL are similar for stable glaucomatous eyes compared to the healthy eyes. Regional differences and changes in nonaxonal contribution to the nerve fiber layer thickness should be considered when assessing clinical RNFL thickness measures.