Early Peripapillary Choroidal Thinning during Myopia Development: A Potential Biomarker for Progressive Myopia Identified in Tree Shrews

Abstract

Purpose

The purpose of this study was to investigate longitudinal thickness changes of the peripapillary tissues in tree shrew eyes with induced high myopia from juvenile age to early adulthood.

Methods

Juvenile tree shrews were randomly assigned to either a control group (16 eyes, 8 animals) or a group with binocular −10 diopter (D) lens wear (18 eyes, 9 animals). Refraction, biometry, and optical coherence tomography scans centered on the optic nerve head (ONH) were obtained weekly for 19 weeks. Data were split into three temporal phases representing myopia onset (phase I, week 1), juvenile high myopia development (phase II, weeks 2–5), and sustained high myopia into early adulthood (phase III, weeks 6–19).

Results

Most eyes in the myopia group developed and maintained a spherical equivalent (SE) closely matching the lens power (−10.4 ± 2.03 D). Unexpectedly, two animals (n = 4 eyes) developed progressive myopia (−18.2 ± 2.9 D). Notably, eyes with progressive myopia showed rapid choroidal thinning compared to non-progressive eyes (−3.03 ± 0.6%/day vs. −1.64 ± 0.9%/day, P < 0.001) despite comparable SE and axial length changes during phase I. During phase III, these eyes exhibited accelerated thinning of sclera (−0.13 ± 0.01%/day vs. −0.02 ± 0.04%/day, P < 0.001) and choroid (−0.12 ± 0.03%/day vs. −0.02 ± 0.04%/day, P < 0.001). The observed chorioscleral thinning was more pronounced in sectors that are closer to the posterior pole.

Conclusions

Sustained negative lens wear can induce progressive myopia in tree shrews. Profound choroidal thinning during early myopia development is a potential biomarker for future chorioscleral thinning, axial elongation, and myopia progression. The asymmetric chorioscleral thinning may contribute to pathologic ONH remodeling and increased glaucoma risk later in life.

Myopia is a growing global health concern due to its rising prevalence, with projections indicating that by 2050, approximately 5 billion people will be affected, including 1 billion with high myopia.¹ Glaucoma,^2,3 maculopathy,^4,5 and choroidal neovascularization⁶ are sight-threatening conditions associated with myopia. Despite its growing burden, the precise mechanisms driving its onset, progression, and comorbidities remain poorly understood. This knowledge gap poses significant challenges for developing effective therapies and predicting complications.

The optic nerve head (ONH) region, encompassing the lamina cribrosa (LC) and peripapillary tissues, is crucial for investigating the interaction between myopia and glaucoma. Both conditions involve biomechanical^7–10 and vascular^11–14 alterations within the ONH. In our previous work,¹⁵ experimental high myopia was shown to induce profound thinning of the peripapillary sclera and choroid at a juvenile age. From a biomechanical standpoint, thinning of the peripapillary sclera can induce ONH and LC deformations,¹⁶ which may increase the risk of biomechanical insult to retinal ganglion cell axons even at normal intraocular pressure (IOP) levels.^17,18 Additionally, thinning of the peripapillary choroid can compromise blood flow, resulting in hypoxia, which may induce pathological connective tissue remodeling¹⁹ and subsequently increase the risk of axonal insult and development of glaucoma.^20,21 Because myopia typically develops during childhood, whereas glaucoma develops later in life, aging is key when investigating the interaction between myopia and glaucoma.¹⁶ However, age-related ONH remodeling following juvenile high myopia remains unclear. To the best of our knowledge, no study has longitudinally investigated the remodeling of peripapillary tissues from the onset of juvenile myopia through adulthood.

Although clinical studies are ideal for this purpose, they are often impractical due to challenges in recruiting and retaining participants for long-term studies. Translational studies utilizing animal models present a viable alternative, providing valuable insights into the remodeling mechanisms of peripapillary tissues that may contribute to the increased glaucoma risk in myopic eyes. The tree shrew serves as a reliable model for myopia due to its well-developed visual system, evolutionary proximity to primates, and rapid maturation.^22–25 Moreover, the primate-like collagenous LC of tree shrews provides an opportunity to explore ONH remodeling during myopia development, aging, and glaucoma.²⁶ In this study, we aimed to longitudinally investigate the impact of sustained experimental high myopia on the thickness of the peripapillary tissues from a juvenile age to early adulthood in tree shrews.

Methods

Experimental Groups

Seventeen northern tree shrews (Tupaia belangeri) were the subjects of this study. Eight animals were subjected to normal visual conditions (n = 16 eyes, normal control group), whereas experimental high myopia was induced in nine animals (n = 18 eyes, myopia group) using binocular −10 diopter (D) lenses starting at the age of 24 days of visual experience (DVE) and sustained until 157 DVE (about 6 months of age). Baseline characteristics of study subjects are provided in Supplementary Table S1. It is worth noting that by the age of 6 months, tree shrews are considered young adults as they become sexually mature at approximately 3 to 5 months of age,^27,28 with their eyes reaching adult size by 20 weeks of age.²⁵ The detailed protocols and procedures to mount the goggles with lenses onto a surgically installed pedestal have been previously described.^23,29,30 All animals were housed in individual cages under a 10/14-hour dark/light cycle with continuous access to food and water.

Refractive and Biometric Measurements

Refractive and biometric measurements were performed weekly on conscious animals using the Nidek ARK-700A infrared autorefractor (Marco Ophthalmic, Jacksonville, FL, USA) and the Lenstar biometer (LS-900; Haag-Streit, Mason, OH, USA). Biometric data were analyzed using tree shrew-specific refractive indices.³¹

Optical Coherence Tomography Imaging

A total of 48 radial B-scans centered on the ONH (Fig. 1) were acquired in both eyes of each animal using the spectral-domain optical coherence tomography (OCT) system Spectralis OCT2 (Heidelberg Engineering, Heidelberg, Germany) with enhanced-depth imaging mode. Baseline OCT B-scans were acquired at 24 DVE, and subsequent B-scans were acquired on a weekly basis using the follow-up mode until 157 DVE (6 months of age). The “follow-up” mode ensured that B-scans were acquired at the same location as used during the baseline scan throughout the longitudinal study.

Figure 1.

Thickness calculation of peripapillary tissues. (A) En face confocal scanning laser ophthalmoscopy image showing the 48 radial B-scans centered on the optic nerve head and circumferential band used for thickness measurements. (B) A representative distortion-corrected B-scan showing the segmented interfaces of the peripapillary tissues and the 50-µm band (marked as a thick turquoise circle in A), where the thickness was quantified. ASCO, anterior scleral canal opening; BMO, Bruch's membrane opening; Ch-RPE, choroid-retinal pigment epithelium complex; RNFL, retinal nerve fiber layer; RRL, remaining retinal layers.

Prior to OCT imaging, animals were anesthetized using an intramuscular injection of 0.02 mL of xylazine (7.5 mg/kg) and continuous low flow of isoflurane (0.5–1% in 100% oxygen at 1 L/min). Animals were placed on a custom-built imaging platform equipped with a heating pad to maintain body temperature. To dilate the pupil and anesthetize the cornea, topical drops of 2.5% phenylephrine hydrochloride, 1% tropicamide, and 0.5% proparacaine were administered. A custom-made rigid contact lens (plano sphere, base curve 4.0 mm, diameter 6.3 mm; Conforma Laboratories Inc., Norfolk, VA, USA) was then placed on the cornea to prevent corneal dehydration and to create a smooth imaging surface. Finally, the imaging platform was aligned with the OCT system to capture images of the ONH including the peripapillary region.

The acquired OCT B-scans (see Fig. 1) were then automatically segmented using an artificial intelligence (AI)-based software (Reflectivity; ABYSS Processing, Singapore) to extract the interfaces of the peripapillary sclera, choroid-retinal pigment epithelium (Ch-RPE) complex, retinal nerve fiber layer (RNFL), and the remaining retinal layers (RRL). The extracted interfaces were anatomically corrected using our previously reported and validated nonlinear distortion correction approach.³² Three-dimensional models were reconstructed from the anatomically corrected tissue interfaces and used to calculate thickness values within a 50-µm circumferential band centered on the anterior scleral canal opening centroid and located 1000 µm away from it (see Fig. 1). The band location was carefully selected to avoid the steep curvature and variable contours of ocular tissues near the neural canal opening, which can compromise measurement accuracy. Additionally, the nonlinear distortion correction approach³² was empirically derived from a tree shrew population with a wide range of axial lengths (7.37 to 8.07 mm), ensuring that the OCT images can be accurately corrected and the 50-µm band quantified across animals and across time as myopia is induced and the eye elongates. Furthermore, the 50-µm band width provides sufficient spatial resolution to detect myopic thickness changes while reducing any remaining distortion effects caused by axial elongation enabling a rigor quantification over time. Global thickness values were calculated and averaged across the entire band, whereas sectoral thickness values were calculated and averaged within the band at the following eight distinct sectors: temporal (T), superotemporal (ST), superior (S), superonasal (SN), nasal (N), inferonasal (IN), inferior (I), and inferotemporal (IT). All procedures adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were also approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee.

Data and Statistical Analysis

In a previous work,¹⁵ thickness changes of peripapillary tissues occurred mostly in the first week of myopia development (24 to 31 DVE) followed by a relatively stable phase until high myopia was fully developed at 59 DVE. Based on this trajectory, we defined three phases for analyzing our data, representing the onset of myopia (phase I = 24 to 31 DVE), the development of juvenile high myopia (phase II = 31 to 59 DVE), and sustained high myopia into early adulthood (phase III = 59 to 157 DVE).

The longitudinal development of the spherical equivalent (SE) and axial length (AL) was examined for each eye to evaluate the individual patterns of myopia development and progression. Longitudinal thickness results were presented as group-level global means. Thickness data were also quantified in eight sectors and the mean sectoral thickness changes at the end of each phase were presented as polar plots to evaluate the symmetry of thickness changes around the ONH. Group differences in longitudinal outcome variables were evaluated using a linear mixed-effects model that incorporated interaction effects to draw statistical inferences where each pair of eyes was nested within its corresponding animal to account for inter-eye correlation. In addition, each eye was nested within time points to account for repeated measurements. Additionally, rates of change in the outcome variables were computed from the coefficient estimates of the fitted statistical model for each group at each phase and further analyzed using a random intercept model to evaluate group differences. Statistical analyses were performed in RStudio (R software version 4.0.2)³³ using the lme4³⁴ and lmerTest³⁵ packages, with the significance level set to 0.05.

Results

All eyes in the myopia group exhibited mild myopia by the end of phase I (SE = −1.03 ± 0.89 D), which progressed further, reaching an SE of −8.9 ± 1.58 D by the end of phase II and −12.3 ± 3.8 D by the end of phase III. In contrast, control eyes underwent emmetropization, with their SE values being significantly different from lens-treated eyes at all phases (P < 0.001). The progression of SE in lens-treated eyes was accompanied by significantly greater axial elongation relative to the control eyes at all phases (P < 0.001).

Whereas most eyes in the myopia group developed and maintained SE levels (−10.4 ± 2.03 D) closely matching the power of the negative lenses (−10 D) used to induce myopia, a group of four eyes (2 animals) unexpectedly showed continued myopia progression, reaching significantly higher SE levels than the rest of eyes in the myopia group (−18.1 ± 2.1 D, P < 0.001) by the end of phase III. This progression in SE was also reflected in AL changes from baseline (Fig. 2). Due to this divergent outcome, subsequent analyses will classify these four eyes with significantly higher SE levels as the “progressive myopia” group, distinct from the remaining eyes in the myopia group categorized as the “non-progressive myopia” group. The Table summarizes the rates of change in AL and SE for phases I to III.

Figure 2.

Individual progression patterns of spherical equivalent (A) and axial length changes from baseline (B) in control, non-progressive, and progressive myopic eyes. Vertical lines and colored backgrounds delineate the phases of myopia onset (phase I = 24–31 DVE), juvenile high myopia (phase II = 31–59 DVE), and sustained high myopia into adulthood (phase III = 59–157 DVE). Asterisks denote significant differences in terms of time-group interaction with Roman numerals indicating the phase(s) where significance was observed.

Table.

Summary of Rates of Change in Axial Length (AL) and Spherical Equivalent (SE) for the Control (C), non-Progressive Myopia (nPM), and Progressive Myopia (PM) Groups During Myopia Onset (phase I = 24–31 DVE), Juvenile High Myopia (phase II = 31–59 DVE), and Sustained High Myopia Into Early Adulthood (phase III = 59–158 DVE)

		Rate of Change			P Value
	Phase	C	nPM	PM	N vs. nPM	N vs. PM	nPM vs. PM
AL (µm/day)	I	6.7 ± 1.6	13.5 ± 2.80	15.4 ± 0.8	0.0003*	0.002*	0.6
	II	7.2 ± 0.5	12.9 ± 0.60	14.7 ± 0.5	0.0001*	0.0001*	0.2
	III	2.2 ± 0.2	2.7 ± 0.30	3.9 ± 0.5	0.04	0.0001*	0.002*
SE (D/day)	I	−0.02 ± 0.01	−0.36 ± 0.01	−0.4 ± 0.01	0.0001*	0.0001*	0.9
	II	−0.02 ± 0.01	−0.25 ± 0.02	−0.34 ± 0.03	0.0001*	0.0001*	0.06
	III	−0.001 ± 0.004	−0.02 ± 0.01	−0.07 ± 0.02	0.07	0.0001*	0.0001*

^*Asterisks denote significant differences in pairwise group comparisons.

Figure 3 illustrates the longitudinal changes in global mean thickness along with the rates of thickness changes for each group, tissue, and phase. Specifically, no significant differences (P > 0.05) were observed in RNFL thickness changes during phase I and phase II (see Figs. 3A, 3B). However, during phase III, a slight but significant thickening was noted in the control (0.04 ± 0.01%/day) and non-progressive myopia groups (0.04 ± 0.01%/day) compared to the progressive myopia group (−0.01 ± 0.01% /day). In contrast, RRL thickness changes differed significantly among the groups (see Figs. 3C, 3D). Both myopic groups showed significantly faster RRL thinning during phase I compared to the control group (P < 0.001), with a significantly greater thinning rate in the progressive myopia group compared to the non-progressive group (P < 0.001). During phase II, RRL of the non-progressive myopia group continued to thin significantly faster compared to the control eyes (P = 0.003), but, beyond this point, RRL thickness remained stable across all groups. The Ch-RPE complex and sclera exhibited similar thinning trends (see Figs. 3E–H), with one notable distinction. During phase I, both the progressive and non-progressive myopia groups, exhibited significantly higher rates of Ch-RPE and scleral thinning compared to the control group (P < 0.001). Notably, the Ch-RPE complex thinned about twice as fast in the progressive myopia group compared to the non-progressive group (P < 0.001; see Figs. 3E, 3F) despite comparable SE and AL in both groups during phase I (see Fig. 2). After a relatively stable phase II, the progressive myopia group demonstrated significantly faster and progressive thinning of the Ch-RPE and sclera compared to the non-progressive myopia (P = 0.03 and 0.01, respectively) and the control (P = 0.009 and 0.008, respectively) groups during phase III (see Figs. 3E–H). Across all tissues, significant group differences in rate changes (see Figs. 3A, 3C, 3E, 3G) mirrored the significant time-group interactions (see Figs. 3B, 3D, 3F, 3H).

Figure 3.

Global group-level time course of thickness changes and corresponding rates for (A, B) retinal nerve fiber layer (RNFL), (C, D) remaining retinal layers (RRL), (E, F) the choroid-RPE complex, and (G, H) the sclera. Error bars represent standard error of mean. Vertical lines and colored backgrounds delineate the phases of myopia onset (phase I = 24–31 DVE), juvenile high myopia (phase II = 31–59 DVE), and sustained high myopia into adulthood (phase III = 59–157 DVE). In A, C, E, and F, the asterisks denote statistically significant time-group interactions with Roman numerals indicating the phase(s) where significance was observed. In B, D, F, and H, the asterisks denote significant differences in pairwise group comparisons.

To examine the spatial symmetry in the thinning patterns around the ONH, thickness changes were compared in eight distinct sectors. Figure 4 illustrates polar plots for sectoral group-level thickness changes at the end of each phase for all tissues. A schematic representation of the eight sectors is placed next to each polar plot to indicate significant comparisons. It is crucial to consider that the posterior pole in tree shrews is located nasally and slightly inferior relative to the ONH,^36,37 while it is located temporally relative to the ONH in human eyes. Therefore, the nasal sector points toward the posterior pole in our polar plots and asymmetric changes toward the nasal direction would be expected to occur toward the temporal direction in human eyes.

Figure 4.

Group-level sectoral thickness changes at the end of each phase: myopia onset (phase I = 24–31 DVE), juvenile high myopia (phase II = 31–59 DVE), and sustained high myopia into adulthood (phase III = 59–157 DVE). Asterisks on the schematic polar representation next to each polar plot denote significant differences in group comparisons based on the group-time interaction effects for each phase at each sector. RNFL, retinal nerve fiber layer; RRL, remaining retinal layers.

Consistent with our global findings shown in Figure 3, RNFL exhibited significant sectoral thickness changes among the groups only in phase III. Specifically, the progressive myopia group exhibited significant thinning compared to the non-progressive myopia group (S, SN, N, and IN) and the control group (ST, S, SN, I, and IT). In contrast, sectoral RRL thickness changes showed significant differences only in phases I and II. Namely, the progressive and non-progressive myopia groups exhibited significantly greater thinning than the control group in all sectors during phase I, with further thinning in all sectors in the non-progressive group during phase II compared to the control group. A significantly greater thinning was observed in the progressive myopia group compared to the non-progressive myopia group during phase I in the S and N sectors (P = 0.01). Overall, both RNFL and RRL underwent modest thickness changes with no distinct asymmetry observed in their thinning patterns. For the Ch-RPE and sclera, both myopic groups exhibited significant thinning in all sectors compared to the control group in phase I. Although scleral thinning was comparable between the two myopic groups (P > 0.05), the progressive myopia group displayed significantly greater Ch-RPE thinning in all sectors compared to the non-progressive group during phase I (P < 0.01). Phase II was relatively stable, with no significant differences observed among the groups. However, phase III revealed a distinct asymmetric thinning pattern in the progressive myopia group, with sectors closer to the posterior pole (SN, N, and IN) experiencing more pronounced thinning of both the Ch-RPE and the sclera.

Discussion

We have conducted a comprehensive investigation into the longitudinal thickness changes of peripapillary tissues from the onset of myopia at a juvenile age to early adulthood. To the best of our knowledge, this is the first study to leverage OCT imaging in the tree shrew model of myopia for this purpose, providing novel insights into the pathophysiological mechanisms underlying myopia development and progression. Our results showed that most eyes with sustained high myopia developed and maintained an SE closely matching the power of the negative lenses used for myopia induction. However, a subset of eyes exhibited continuous myopia progression, ultimately attaining an SE of −18.2 ± 2.9 D. This subset of eyes with progressive myopia exhibited profound choroidal thinning after the onset of induced myopia during phase I, nearly doubling that observed in the non-progressive group. Remarkably, this disparity in choroidal thinning between the progressive and non-progressive groups occurred despite comparable levels of refraction and AL changes by the end of phase I. After that, the peripapillary tissues thinned at comparable rates in progressive and non-progressive groups during phase II as their SE reached the target level of high myopia set by the −10 D lens. During sustained negative lens wear until early adulthood in phase III, a clear divergence re-emerged with the progressive myopia group exhibiting progressive and asymmetric thinning of the Ch-RPE and sclera, with sectors closer to the posterior pole being the most affected.

The current findings align with our previous work,¹⁵ where thinning of peripapillary tissues was heterogeneous (more profound in the choroid and sclera relative to the RNFL and RRL) and occurred predominantly after the onset of myopia during phase I. We previously suggested that the tri-lamellar collagen-elastin structure of Bruch's membrane³⁸ and its potential growth ability^39,40 may act as a shearing band to potentially shield the retina against the shearing forces generated in the sclera. During phase II, eyes with progressive and non-progressive myopia showed relatively stable thickness profiles despite ongoing axial elongation. We previously suggested two possible explanations for this discrepancy: (i) either thinning ceases only in the peripapillary region but continues in the periphery or (ii) an intrinsic mechanism counteracts the passive stretching effects induced by axial elongation.¹⁵ However, continuous axial elongation in the progressive myopia group throughout phase III was accompanied by significant thinning of Ch-RPE, sclera, and RNFL, with a similar trend observed in the RRL, albeit not reaching statistical significance. Considering these findings, we propose that intrinsic homeostatic mechanism(s) may actively intervene to inhibit or mitigate the effects of passive stretching forces on the retina during myopia development. However, once a critical threshold is exceeded during myopia progression, such protective mechanism(s) may fail.

Another key finding of the current study was the temporal correlation between the profound early choroidal thinning after the onset of myopia (phase I) and the development of progressive myopia during early adulthood (phase III). Lee et al.⁴¹ observed similar choroidal changes in a cohort of myopic children, in which a subset of eyes progressed to high myopia. (−4.03 ± 2.44 at baseline vs. −5.68 ± 2.52 after 4 years). Interestingly, significant choroidal thinning was observed in the subset of eyes with myopia progression, whereas it remained relatively stable in the non-progressive group.⁴¹ Previous studies^42–46 have shown that choroidal thickness changes precede, and consequently can predict, the direction of eye growth. For example, several experimental studies have shown that the choroid thins^15,47,48 during myopia development and thickens^48,49 during recovery from myopia. Similarly, choroidal thickening was shown to be associated with decelerated eye growth attained with atropine treatment,⁵⁰ orthokeratology,^51,52 multifocal contact lens wear,⁴⁴ and red-light therapy.^53,54 Our findings suggest that the extent of choroidal thinning during early myopia development may be predictive of future scleral thinning and axial elongation, resulting in the development of progressive myopia. A key question is what factors dictate the extent of initial choroidal thinning in response to a myopiagenic stimulus. Given that our experimental subjects were housed under identical conditions and exposed to the same myopiagenic stimulus (−10 D lens), it is plausible to consider the involvement of genetic determinants. This aligns with emerging evidence that although environmental factors may initiate myopia, certain genetic loci may modulate its progression and final severity.^55–58 Decoding these genetic factors could provide valuable insights into the underlying mechanisms and aid in early identification of individuals with an elevated risk of developing progressive myopia.

Pertinent to its predictive potential, the role of the choroid is increasingly being investigated in the regulation of normal eye growth, the development of myopia, and myopia progression.^42,59 Due to its intermediate location between ocular tissues, the choroid can indirectly influence axial elongation by acting as a conduit or barrier for retinal/RPE signals to the sclera. Prior research has suggested that choroidal accommodation can shift the retinal position to match the focal plane during induction or recovery from myopia.^43,59,60 However, the instant nature of choroidal accommodation cannot explain the temporal gap between early choroidal changes and later scleral thinning and axial elongation observed in the current study. Additionally, the observed asymmetric pattern of scleral thinning closely aligns with that of the choroid, suggesting a mutual interaction between scleral and choroidal responses during myopia development and progression. The sclera, the outermost load-bearing layer of the eye, primarily defines the shape of the eye.¹⁶ However, several studies revealed a more complex choroidal structure than previously believed, with growing evidence suggesting that it can actively influence and modulate scleral growth and remodeling.^59,60

One possible mechanism that may underlie the long-term effect of choroidal thinning on myopia progression is that the reduction in choroidal blood flow may lead to scleral hypoxia, initiating a cascade of events leading to scleral biomechanical weakening and hence excessive axial elongation.⁶¹ Although we did not assess the changes in choroidal blood flow, clinical and experimental studies reported a strong correlation between choroidal thickness and choroidal blood flow.^62–64 Because the hypoxic response is a localized phenomenon, it is reasonable to think that areas with significant choroidal thinning, and consequently reduced blood flow, would be most susceptible to scleral hypoxia, which aligns with the co-localized asymmetric thinning patterns observed here in the choroid and sclera. Zhou et al.⁶⁴ observed myopia development with underlying scleral hypoxia in guinea pigs when choroidal blood flow was intentionally reduced, either surgically or pharmacologically. This mechanism is further supported by the finding of upregulated hypoxia-inducible transcription factor (HIF-1α),^65–67 a primary mediator of cellular response to hypoxic stress in most mammals. Moreover, pharmaceutical up- or downregulation of peroxisome proliferator-activated receptor γ (PPARγ) was shown to modulate collagen expression levels and induced myopia through an inverse relationship between changes in PPARγ and HIF-1α changes, further supporting the role of scleral hypoxia in myopia.⁶⁸ Another potential mechanism may reside in the secretory function of the choroid. Apart from providing the blood supply, several growth factors are produced within the choroid, including the fibroblast growth factor⁶⁹ and transforming growth factor-β,⁷⁰ in addition to various enzymes, including matrix metalloproteinase⁷¹ and tissue plasminogen activator.⁷² The formerly mentioned growth factors and enzymes are known to modulate the scleral glycosaminoglycan (GAG) synthesis and extracellular matrix remodeling.^73,74 Supporting this speculated mechanism, scleral GAG synthesis has been shown to increase and decrease when co-cultured with choroid from fast- and slow-growing eyes, respectively.⁷⁵ Although the observed temporal association between choroidal thinning and myopia progression does not confirm or exclude causation, it suggests that early choroidal thinning may serve as a biomarker for eyes at risk of progressive myopia. Further research is necessary to elucidate the proposed choroidal mechanisms and their potential role in myopia development and progression.

One crucial aspect of our findings is the observed asymmetric thinning pattern in sectors adjacent to the posterior pole in both the choroid and sclera during phase III. This asymmetric pattern was also present in RRL to a lesser extent, although it did not reach statistical significance. These findings support the notion that myopia induces asymmetric remodeling of the ONH.^16,76,77 Our previous study identified asymmetric remodeling patterns in the ONH region in juvenile high myopia (24 to 59 DVEs), where changes in border tissue configuration from internally to externally oblique were more pronounced toward the posterior pole.⁷⁸ However, no clear asymmetric thinning patterns of peripapillary tissues were observed during this timeframe of myopia development in the present study (phase II) and our previously published study.¹⁵ Because the asymmetric thinning pattern has been observed only during phase III, it is possible that aging plays a key role in determining not only the thinning extent but also the deviation from a symmetric thinning pattern. Additionally, these asymmetric thinning patterns were only observed in eyes with progressive myopia, where the risk of developing associated complications is highest. If this asymmetric thinning pattern persists and continues to progress into late adulthood, it may focally disrupt retinal homeostasis and increase the risk of developing glaucoma even at normal IOP. Supporting this finding, a recent 3-year longitudinal study in young adults revealed an accelerated RNFL thinning in highly myopic glaucomatous eyes compared to those with glaucoma alone.⁷⁹ The observed thinning in this human cohort occurred predominantly in the temporal sector,⁷⁹ which matches the asymmetric thinning pattern seen here in sectors that are closest to the posterior pole (nasal in tree shrews).

Whereas this is the first study of its kind to investigate peripapillary thickness changes in myopic tree shrew eyes from juvenile age to early adulthood, limitations exist. First, experimentally induced myopia may not perfectly mimic human myopia regarding some of its critical aspects, such as the development rate and the magnitude of axial elongation. Additionally, although monocular myopia models are often used in animal studies, we used a binocular myopia model which mimics human myopia more closely. It remains unclear whether sustained monocular lens defocus would produce similar patterns of tissue remodeling as observed here. Second, our findings are limited to the peripapillary region due to the restricted field of view of the used OCT imaging system. Acknowledging regional variations of myopic changes, it remains unknown if our findings extend to regions beyond the peripapillary region. Other areas of interest for myopic remodeling are the posterior pole, visual streak, and the area centralis. However, in the tree shrew, reliable imaging of these regions is challenged by the lack of a fovea or other distinct tissue-level landmarks, as well as by the positional constraints imposed by the animal's long nose and the anesthesia cone used during OCT imaging. A wide-angle OCT imaging system may provide further insights into thinning patterns across the posterior segment, including the posterior pole, visual streak, and the area centralis regions, while still leveraging the ONH anatomic landmark. Additionally, histological studies hold immense potential to characterize key features of progressive myopia, including staphyloma formation and microstructural changes at the cellular level. Third, our sample size is relatively small, particularly in the progressive myopia group, warranting cautious interpretation. Last, whereas we propose a potential correlation between structural changes (for example, choroidal thinning) and functional alterations (scleral and retinal hypoxia), further research is needed to elucidate structure-function correlations.

In summary, this study showed that sustained experimental high myopia can lead to progressive myopia in tree shrews. Eyes with progressive myopia exhibited profound choroidal thinning after the onset of myopia followed by progressive chorioscleral thinning and axial elongation during aging from juvenile age to young adulthood. Additionally, an asymmetric pattern of chorioscleral thinning was evident in the progressive myopia group. Collectively, a combination of vascular insufficiency (reduced choroidal blood flow) and biomechanical weakening (asymmetric thinning) may supersede the ability of the intrinsic homeostatic mechanism(s) to counteract these effects, potentially contributing to the increased risk of developing pathologic myopia and glaucoma later in life.