Abstract
Purpose:
Individual differences in photoreceptor topography include a higher density of cones in the fovea and different rates of decrease of cone density with increasing eccentricity. We investigated the lack of a smooth change in cone density at the border of and extending into the scleral crescent, where there are no apparent retinal pigment epithelial (RPE) cells.
Methods:
To measure cone density at the optic nerve head without the presence of confounding factors such as high myopia or glaucoma, we recruited two subjects < 35 yr with wide scleral crescents despite healthy eyes and low refractive errors. The topography of the scleral crescent was documented with SD-OCT (Spectralis). The hyperreflectivity indicating the lack of melanin-containing RPE cells and the scleral pattern of birefringence were localized with scanning laser polarimetry (GDx). For high resolution imaging at the scleral crescent, a focus series with the Indiana Adaptive Optics Scanning Laser Ophthalmoscope (AOSLO) was used to document the thinned aspect of the scleral crescent and topography change, along with depigmentation. We used the AOSLO to image photoreceptors, the scleral crescent, and the overlying retinal blood vessels. The AOSLO had two channels for imaging photoreceptors via direct backscatter and multiple scattering, using center wavelengths of 785 and 810 nm. We documented cones in the scleral crescent with two operators. Two masked graders counted cones in 6 sampled regions per image, computing cone density with visual angle.
Results:
Cones density changed markedly near the scleral crescent over RPE cells and within the scleral crescent, where there was no evidence of RPE cells. Patches of higher cone density were found near the scleral crescent, but also regions of low density and even cones that appeared sideways in or near the scleral crescent. The ratio of cone density on and near the edge of scleral crescent to farther away and having RPE cells varied from 0.50 to 1.4 within a single image. In contrast, cones sampled within a single image but nearer to the fovea had more uniform coverage, ranging in density by only .83 to .99.
Conclusions:
Cone density, known to decrease systematically with increasing distance from the fovea, increases in some patches near the optic nerve head. Cones that are not over RPE cells have irregular densities and sometimes orientations.