Axon Content of the Circumpapillary Nerve Fiber Layer in Experimental Glaucoma

Purpose: To determine axonal content from optical coherence tomography (OCT)-derived circumpapillary retinal nerve fiber layer (RNFL) measures using histologically informed models in a non-human primate experimental glaucoma model. Methods: Unilateral experimental glaucoma was induced in 22 animals (mean age, 6.1 ± 1.9 years; 10 females) who were monitored with OCT. At varying stages of neuropathy, animals were euthanized and perfusion fixed. For RNFL histology, circumpapillary tissue (∼0.75 × 0.75 mm) from six equally spaced regions was obtained and processed for serial block face scanning electron microscopy in five animals. A single high-resolution image of the block face was used to quantify axon size (cross-sectional area) and non-axonal tissue area within the RNFL. Using the 95% limits of the histological data as a starting point, models for estimating axon count from OCT RNFL area were optimized and compared to total retrobulbar optic nerve axon counts in 22 animals. Results: The mean circumpapillary RNFL thickness of control eyes was 112 ± 6 µm, and it was 70 ± 24 µm in eyes with experimental glaucoma. Axon size and non-axonal area varied across the circumpapillary regions (F5, 31 = 9.06 and F5, 36 = 3.94, respectively; both P < 0.01), but did not differ between the two eyes (F1, 22 = 0.59, P = 0.45 and F1, 15 = 2.10, P = 0.17). The optimized model for estimating axon content from the circumpapillary RNFL profile had good agreement with the retrobulbar axon count (mean difference = 25; 95% limits of agreement, -334,239 to 334,289). Conclusions: The axon size distribution and non-axonal area in the circumpapillary RNFL are similar between healthy and stable glaucomatous eyes. Building on this, a location-specific model is proposed that estimates axon content from the circumpapillary RNFL profile after accounting for regional variability in axonal and non-axonal components.