In primates, the retinal neurons and ocular morphology continue to mature early in life.
45–49 During this period of development and emmetropization, there is scleral shell expansion, with associated changes in ONH and retinal layer structure.
37,50,51 Myopia is a refractive condition that occurs when the scleral shell expands beyond the refractive power of the anterior segment and is linked to many ocular pathologies, including glaucoma.
20–27 However, the pathophysiology of this association remains elusive. Animal models may aid in determining features of the ONH and macula inner retinal thickness, which places some eyes at a higher risk of glaucoma. In this experiment, using year old animals with anisometropic axial elongation, we show that longer eyes have a larger BMO with greater scleral crescent areas, and a more anteriorly positioned lamina surface. Although measures of the peripapillary nerve fiber layer and MRW were not related to axial length, longer eyes had thinner macula inner retinal thickness.
OCT morphometrics are typically analyzed for one eye, as measures from the genetically identical fellow eye are highly correlated.
52 Several population studies in children showing inter-ocular differences in OCT metrics, are often not adequately corrected for ocular magnification.
53,54 Hence, to evaluate axial length-specific associations, in addition to single eye analysis, we analyzed the interocular difference for significant relationships using ocular magnification corrected measures. Such interocular differences are commonly analyzed in refractive error studies to better identify the effects of monocular treatments, with the assumption that eye developments in monocularly treated animals are independent of each other. When applied to morphological data, the interocular difference metric reduces the influence of individual variability in structural dimensions and biological development. Hence, the interocular difference agreeing with the single, right eye, analysis further supports the idea that scleral expansion alters posterior structure as measured by OCT.
Human, in vivo OCT and cadaveric studies show that during infancy, along with globe expansion, there is an increase in the size of the ONH.
16,51,55 The increase in ONH size is linear in eyes with low to moderate myopia, but becomes nonlinear in individuals with high myopia.
15,56 The monkeys in the current study did not have signs of pathological myopia, and, hence, only the linear component was realized for the relationship between BMO area and axial length (see
Figs. 3A,
3B). Although the size of the BMO reflects on the underlying scleral opening,
57 due to variability in visualization of the scleral canal, we were unable to accurately quantify the scleral opening diameter for the current study. Addition of this metric would have enabled quantification of the relative positions of the ONH openings, whose offset increases with axial length.
58–60 A morphological feature related to opening-offset development is scleral crescent,
61 for which the distance from the temporal edge of the crescent to the fovea is not related to axial length, but the distance from the edge of the nerve to the fovea is.
19 In the primate model, based on manual delineation on an OCT enface slab, the scleral crescent was linearly related to axial length (see
Figs. 3E,
3F). However, the development of a scleral crescent in young rhesus monkeys is heterogenous, with some animals not developing this feature, but having large interocular differences in axial length (see
Supplementary Data).
Although the size of the ONH has been hypothesized to be a risk factor for glaucoma development, this has not been substantiated, even in anisometropic eyes.
62–65 However, in addition to ONH enlargement, axial elongation is associated with thinning of the posterior pole sclera, peripapillary scleral flange, and lamina cribrosa.
66–68 As the scleral canal enlarges with axial elongation, the lamina is thought to become taught, and is anteriorly displaced. This is supported by human studies showing that hyperopic eyes have a more posterior lamina than myopia.
18 Similarly, the right eye and interocular difference data show that the lamina of longer eyes have a more anterior ALCS position when referenced to the BMO. However, the position of the BMO referenced to the BM plane was not correlated with axial length. This suggests that the anterior displacement of the lamina is likely a reflection of radial rather than lateral strain. In fact, the relationship between BMO area and ALCS position had a negative slope and was statistically significant (
R2 = 0.22,
P = 0.01, data not shown).
Although several features of the ONH were related to axial length, the MRW and peripapillary RNFL thickness were not, which is comparable to previous work in this animal model.
43 Specifically, whereas the RNFL thickness from standard circular 12 degree diameter scans is shown to be negatively correlated with axial length, this was not the case when lateral scaling was included, and scan paths extrapolated from a fixed distance from the BMO.
43 These findings are similar to human studies, which show that the relationship between circumpapillary RNFL thickness scans from 12 degree diameter circular scans and axial length can be explained by ocular magnification.
69–71 As circumpapillary RNFL thickness and MRW correlate with total retinal ganglion cell axon count in the optic nerve,
72 the data suggest that axial elongation in these subjects did not result in retinal ganglion cell loss. Hence, the results of the current study suggest that stretching of axons from mild to moderate axial elongation alone has minimal effect on retinal ganglion cell content.
The relationship between choroid thickness and axial length/refractive error has been extensively studied, showing that longer or more myopic eyes have a thinner choroid.
73–76 In these previous studies, choroid thickness is most commonly measured in the macula region, where the changes are influenced by axial length and other factors, including, but not limited to defocus, age, medication use, and time of day.
77–83 The relationship between axial length and choroid thickness has also been reported for the peripapillary region,
84 and, as with previous studies, the current work suggests that longer eyes have a thinner peripapillary choroid, although this relationship was not appreciated when interocular comparisons were made (see
Figs. 3G,
3F). The latter likely reflects the larger variability of this measure in comparison to the effect size for rhesus monkeys and our relatively small sample size, as the relationship between axial length and choroid thickness was also significant when only left eye data were analyzed (
R2 = 0.50,
P < 0.01, plot not shown, see
Supplementary Data File). Although several studies show that choroid thickness is not decreased in patients with open-angle glaucoma,
84–86 it is thinner in patients with peripapillary atrophy,
87 and could be associated with vascular compromise to the ONH.
In addition to the ONH and peripapillary region, OCT assessment of the macula region, including the inner retinal thickness, provides valuable information on retinal ganglion cell content in healthy and diseased eyes.
88,89 Although the association has a lot of variability, with more significant axial elongation, human studies suggest greater thinning of the total retinal thickness in the periphery.
3,90,91 For the macula region, whereas some studies show no relationship with axial length, others suggest a negative correlation with the nuclear layers, including the GCIPL, inner nuclear layer, and outer nuclear layer.
3–8,92–94 In the present study, both total retinal and GCIPL thickness were negatively correlated with axial length. The slope of the function was steeper for total retinal thickness than GCIPL thickness, suggesting greater outer than inner retinal thinning with axial elongation. Together with the ONH data, these findings suggest that although axial elongation does not result in retinal ganglion cell loss, the cell density is reduced in the macula region.
The study has several additional limitations to those already mentioned. The current data are cross-sectional in nature, and longitudinal changes, as demonstrated in other species, cannot be ascertained.
95 Whereas both the ONH and macula regions were imaged, widefield imaging could provide added information on the extent of retinal thinning with eccentricity with axial elongation. Due to the limited number of subjects, sector-based analysis was not performed for statistical reasons. Our laboratory has reported on the repeatability of most OCT metrics used for this study. However, in this cross-sectional study, we could not assess repeatability of new metrics, including the crescent area. Although assumptions can be made on cellular content, the study did not quantify retrobulbar axon counts or retinal nuclear layer cell density. In addition, the study did not induce a pathological state to determine if morphological differences in longer versus shorter eyes place them at higher risk of disease. To address these limitations, our future studies will include longitudinal data collection, with glaucoma risk assessment when animals reach sexual maturity.