Changes in retinal layer thickness with maturation in the dog: an in vivo spectral domain - optical coherence tomography imaging study

Abstract

Background: Retinal diseases are common in dogs. Some hereditary retinal dystrophies in dogs are important not only because they lead to vision loss but also because they show strong similarities to the orthologous human conditions. Advances in in vivo non-invasive retinal imaging allow the capture of retinal cross-section images that parallel low power microscopic examination of histological sections. Spectral domain - optical coherence tomography (SD-OCT) allows the measurement of retinal layer thicknesses and gives the opportunity for repeat examination to investigate changes in thicknesses in health (such as changes with maturation and age) and disease (following the course of retinal degenerative conditions). The purpose of this study was to use SD-OCT to measure retinal layer thicknesses in the dog during retinal maturation and over the first year of life. SD-OCT was performed on normal beagle cross dogs from 4 weeks of age to 52 weeks of age. To assess changes in layer thickness with age, measurements were taken from fixed regions in each of the 4 quadrants and the area centralis (the region important for most detailed vision). Additionally, changes in retinal layer thickness along vertical and horizontal planes passing through the optic nerve head were assessed.

Results: In the four quadrants an initial thinning of retinal layers occurred over the first 12 to 15 weeks of life after which there was little change in thickness. However, in the area centralis there was a thickening of the photoreceptor layer over this time period which was mostly due to a lengthening of the photoreceptor inner/outer segment layer. The retina thinned with greater distances from the optic nerve head in both vertical and horizontal planes with the dorsal retina being thicker than the ventral retina. Most of the change in thickness with distance from the optic nerve head was due to difference in thickness of the inner retinal layers. The outer retinal layers remained more constant in thickness, particularly in the horizontal plane and dorsal to the optic nerve head.

Conclusions: These measurements will provide normative data for future studies.

Keywords: Area centralis; Dog; Maturation; Retina; SD-OCT.

Fig. 1

Changes in retinal layer thicknesses with age in the 4 regional quadrants and area centralis.a. Confocal scanning laser ophthalmoscopy (cSLO) image of a left eye fundus indicating the sites at which retinal layer thicknesses were measured. b. Representative SD-OCT high resolution cross-section image of the retina in the dorsal quadrant in a 12-week-old dog with adjacent bars showing the different layers measured: total retina (TR), Receptor+ (REC+), inner retina (IR), and outer nuclear layer (ONL). c-f. The mean (+/− SD) thicknesses of the measured retinal layers in the dorsal (c), ventral (d), nasal (e) and temporal (f) quadrants with age. Each dataset is fitted with a linear regression model. All layers in each of the regions thinned between 4 and 12 weeks of age. A repeated measures correlation showed there was a significant linear correlation between decrease in thickness of each of the four layers in each of the regions and age. See Additional file 1 - Tables S2A and S2B for r and p-values, respectively, as well as Additional file 1 - Tables S1A and S1B for raw values and percentage changes with age respectively. g. Representative SD-OCT high resolution cross-section images of the retina in the area centralis in a 12-week-old dog with adjacent bars showing the different layers measured: total retina (TR), Receptor+ (REC+), inner retina (IR), and outer nuclear layer (ONL). Note the thicker IR layer in the area centralis region compared to the IR thickness in the dorsal region (b). h. The mean (+/− SD) thicknesses of the measured retinal layers in the area centralis with age. Each dataset is fitted with a linear regression model. IR thinned to about 8 to 10 weeks of age while the REC+ thickened over this period and the ONL and TR showed little change in thickness with age. See Additional file 1 - Tables S1A and S1B for raw values and percentage changes with age respectively, as well as Additional file 1 - Tables S2A and S2B for r and p-values, respectively

Fig. 2

Mean layer thicknesses with age of REC+, ONL, IS/OS and heat map in the AC.a-c. Mean (+/− SD) layer thicknesses with age of a. Receptor+ (REC+), b. Outer nuclear layer (ONL), c. Photoreceptor inner segment/outer segment (IS/OS) for the 5 regions. Note that in the dorsal, ventral, nasal and temporal regions the REC+ thinned with age. In the area centralis it was thinner than in the other regions and trended towards thickening with age (A). The ONL in the area centralis changed little with age (whereas in the other areas in thinned). The change in REC+ thickness with age in the area centralis seemed to be accounted for primarily by thickening of the IS/OS. d, e. Heat maps showing layer thickness of ONL and IS/OS in the region of and surrounding the area centralis in a 12-week-old dog. This shows the regionality of the difference in thickness of these 2 layers

Fig. 3

Retinal layer thicknesses changes with age and eccentricity in the dorso-ventral and naso-temporal planes. a cSLO fundus image showing planes along which retinal layer thickness were measured. The temporal plane is shown in more detail with the corresponding SD-OCT image showing positions where measurements were made. b and c Spider graphs showing the mean (+/− SD) total retina (TR), inner retina (IR), Receptor+ (REC+) and outer nuclear layer (ONL) thicknesses in the ventro-dorsal (b) and naso-temporal (c) axes at 4, 6, 12, 26 and 52 weeks of age. Note that at each age the total retina (TR – shown in black in the figures) thins with distance from the optic rim. This is mostly accounted for by thinning of the inner retina (IR, red). The outer retina (REC+ and ONL) shows little change with distance from the optic rim (also see Fig. 4). The dorsal retina is thicker than the ventral retina at all ages and the nasal retina slightly thicker than the temporal retina at early ages (4 and 6 weeks of age). A linear mixed effects model was calculated to analyze layer thickness changes with increasing distance from the optic rim. Correlation of layer thickness with distance from optic rim and significance values are shown in Additional file 1 - Tables S4A and 4B, respectively. Additional file 1 - Tables S3A and S3B show raw values and percentage changes with eccentricity, respectively. d-e Representative SD-OCT high resolution cross-section images of the retina at 12 weeks of age of the dorsal (d oriented vertically) and nasal (e oriented horizontally) quadrants centered at 1, 3 and 6 mm from the edge of the optic rim. The adjacent bars show the measurement of the different retinal layers. Note the progressive thinning of the TR and IR from 1 to 6 mm from the optic rim while the REC+ and ONL thicknesses show little change

Fig. 4

Spider graphs showing the mean of each retinal layer thicknesses with eccentricity and age. Mean (+/− SD) of retinal layer thicknesses in the ventro-dorsal (a, c, e and g) and naso-temporal (b, d, f and h) axes from 1 to 7 mm from the optic rim at 4, 6, 12, 26 and 52 weeks of age. Total retina (TR; a and b), inner retina (IR; c and d), Receptor+ (REC+; e and f), outer nuclear layer (ONL; g and h) and are shown. The total retina thins with distance from optic nerve rim in all directions (a and b) mostly accounted for by thinning of the inner retina (c and d). The outer retina (ONL and REC+) thins with distance from the optic rim ventrally (e and g) while dorsally (e and g) and in the naso-temporal plane (f and h) there is little thinning with distance from the optic rim