Delayed Surgical Reversal of Optic Nerve Compression Leads to Exponential Degeneration of Optic Nerve Fibers and Selective Sparing of the Small Fibers

Abstract

Purpose: To evaluate the effectiveness of surgical reversal of experimental optic nerve compression in treating persistent compressive optic neuropathy and to explore the relationship between surgical outcomes and the timing of the procedure.

Methods: Surgical reversal procedures (decompression surgery) were conducted at five time intervals: 1, 3, and 7 days and 2 and 3 weeks following optic nerve compression in a rabbit model. The groups were labeled as DC-1d, DC-3d, DC-7d, DC-2w, and DC-3w, respectively. The study investigated changes in ganglion cell complex (GCC) thickness using spectral-domain optical coherence tomography and the percentage of surviving retinal ganglion cells (RGCs) through immunofluorescence staining and optic nerve axons stained with p-phenylenediamine at 4 weeks after decompression. Additionally, the area distribution of surviving axons was analyzed.

Results: The decline in GCC thickness was halted following decompression. The remaining thickness of the GCC in group DC-1d was found to be statistically significantly higher at 2, 3, and 4 weeks postonset compared to the no-decompression group. Similarly, GCC thickness in group DC-3d was significantly higher at 3 and 4 weeks postonset. The percentage of surviving RGCs and axons at 4 weeks postonset exhibited an exponential correlation with the onset time of decompression, with R² values of 0.72 and 0.78, respectively. The surviving axon area declined following delayed decompression.

Conclusions: Persistent substantial compression on the optic nerve leads to exponential degeneration of the optic nerve, initially affecting larger optic nerve fibers. Early intervention aimed at relieving the compression on the optic nerve may offer potential benefits in mitigating the degenerative effects and conserving visual function.

Compressive optic neuropathy (CON) can arise from diverse etiologies such as inflammation,^1,2 vascular conditions,^3,4 trauma,^5,6 or tumors.^7–11 Approximately four new cases per 100,000 individuals are affected by CON.^8,12 Sustained optic nerve (ON) compression can lead to direct mechanical damage and potential ischemia and demyelination,¹³ culminating in visual function loss. 

In clinical practice, decompression surgery is often considered a primary treatment option for patients with radiologic evidence of compression.¹⁴ However, the decision to undergo immediate surgery may be influenced by economic constraints, limited medical resources, bad general condition, and potential surgical risks.^15,16 As a result, some patients may choose to delay surgery, placing clinicians in the position of providing accurate prognostic recommendations. However, conflicting opinions regarding the impact of decompression surgery and the urgency of the procedure persist. First, the outcomes of decompression surgery are inconsistent; while some patients experience improvement, others deteriorate,^17,18 leading to inconclusive findings regarding the influence of surgery on visual acuity.¹⁹ Second, some researchers have found no significant linear relationship between the duration of preoperative vision symptoms and visual improvement postsurgery,²⁰ while others argue that early decompression surgery results in better visual outcomes.^8,21 The variability in the pathogenesis and severity of different types of compressive optic neuropathy contributes to these disparate viewpoints, underscoring the need for the development of a generalized predictive model for optic nerve decompression outcomes. 

Thus, the primary objective of this study is to establish a sequence of decompression time points to assess the impact of decompression surgery on the optic nerve under persistent compression and to explore the association between surgical outcomes and the timing of the surgical reversal of experimental optic nerve compression. The findings of this research may provide valuable insights and evidence-based clinical recommendations to address the aforementioned inquiries. 

Method

Animals

Belgian gray rabbits were obtained from the Animal Center of the Institute of Surgery, Wenzhou Medical University, Wenzhou, China, with approval from the Animal Ethics Committee of Wenzhou Medical University. The rabbits used in the study were male, were approximately 6 to 8 weeks old, and weighed between 1.8 and 2.5 kg. They were housed in an individually ventilated cages (IVC) with a room temperature of 22°C, alternating light and dark every 12 hours, and provided with ample food and water. All animal experiments adhered strictly to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Surgical Protocol for Optic Nerve Decompression in Experimental Rabbits

The optic nerve compression model (3/4 comp type) was established according to the protocols outlined in our previous publications.²² The protuberance on the inner ring of this type of artificial optic canal implant (AOCI) was 1.5 mm, which was three-fourths of the inner diameter of AOCI (Fig. 1A). This experimental design induces severe compression on the optic nerve. Both the induction of optic nerve compression and its surgical reversal (optic nerve decompression surgery) were performed by the same experienced surgeon. Anesthesia was induced by intramuscular injection of 0.2% sodium pentobarbital (1.0 mL/kg) and 0.025% dimethylphenidate hydrochloride (0.2 mL/kg). The rabbits were positioned supine on the operating table, with their backs and foreheads facing upward. Prior to surgery, the forehead hair was shaved using a razor, and the surgical area was disinfected with povidone iodine. The disinfection procedure involved applying povidone iodine from the upper edge of the nose tip to the forehead, from the back to the front edge of the ears, and from the sides to the upper edge of the eye orbits, while ensuring to avoid contact with the mucous membranes of the mouth, nose, and eyes. Throughout the anesthesia, surgery, and recovery process, the rabbits were covered with sterile wipes to maintain their core body temperature. 

Figure 1.

 

Diagram of optic nerve decompression time interval and changes of retinal ganglion cell complex thickness. (A) Diagram of decompression time interval. Solid dots: indicate the onset of compression on the optic nerve. Dashed line: represents the onset of the decompression procedure. (B) Representative OCT images of the rabbit retina. The OCT scan lines pass through the ON head and are delineated by the green lines. The GCC thickness was measured at locations 4000 µm from lower edge of the ONH, directed downward toward the peripheral retina. GCC refers to the ganglion cell complex, including the inner plexiform layer (IPL), ganglion cell layer (GCL), and RNFL layers, as indicated by double-end red arrows. (C) GCC thickness change curve of different groups. The GCC thickness of the no-decompression group and the groups undergoing decompression at 1 day, 3 days, 7 days, 2 weeks, and 3 weeks after the onset of compression is represented as a percentage of the GCC thickness in the CON eyes, compared to NO DC group, n = 3–8. Data are presented as mean ± SEM. *P < 0.1, **P < 0.05, ***P < 0.01, ****P < 0.001, independent-samples t-test. The orange arrows indicate the decompression time nodes.

Specific Surgical Procedure

• 1. If the original surgical wound from the optic nerve compression in the experimental rabbit had not completely healed, the blood scab was removed and the skin tissue was dissected along the path of the blood scab coverage. If the surgical wound had completely healed, a 3-cm surgical incision was made along the original surgical healing scar using a surgical blade.
• 2. The fibrous hyperplastic tissue overlying the periosteum was bluntly separated, the original periosteal suture was located and removed, and an appropriate length of surgical suture was used to traction the periosteum bilaterally to fully expose the operative field.
• 3. The hyperplastic fibrous connective tissue was cut with microsurgical scissors and bluntly separated with hemostatic forceps to locate the original orbital septal suture. Extra care was taken in this step to avoid bleeding due to the disrupted clear surgical path caused by the hyperplastic fibrous tissue and the large amount of neovascularization.
• 4. After locating and removing the orbital septal suture, a blunt dissection downward was performed, taking care to identify the gray pigmentation at the junction of the optic nerve and the sclera and locate the AOCI. The AOCI was carefully separated from the surrounding fibrous connective tissue and fully exposed.
• 5. The needle holder was used to fragment the panels on both sides of the AOCI; the upper part of the AOCI was removed, releasing the optic nerve compression; and the lower part of the AOCI was retained without pulling or compressing the optic nerve throughout.
• 6. The orbital septum and periosteum were closed with 5-0 sutures and the skin with 3-0 sutures. Tobramycin dexamethasone ophthalmic ointment was applied to the surgical wound postoperatively. The animal was kept warm until it woke up and then returned to the rearing cage with adequate food and water.

The optic nerve decompression (DC) procedures were conducted at five distinct time intervals: 1 day, 3 days, 7 days, 2 weeks, and 3 weeks following the initiation of optic nerve compression. A control group without decompression was also included. Consequently, the study encompassed six distinct groups, denoted as DC-1d, DC-3d, DC-7d, DC-2w, DC-3w, and NO-DC. For the data collection of retinal ganglion cell (RGC) count, ON count, and ON area measurements, the sample size for each group was 3 to 4. Due to the numerous measurement points and the increased mortality rate of rabbits under repeated anesthesia, the sample size for recording ganglion cell complex (GCC) thickness changes ranged between 3 and 8 for each of the six groups. 

Spectral-Domain Optical Coherence Tomography Imaging

To address the limitations associated with measuring the thin retinal nerve fiber layer (RNFL) and to monitor changes in retinal ganglion cells, the thickness of the GCC was assessed using the spectral-domain optical coherence tomography (SD-OCT) technique. The GCC encompasses the inner plexiform layer, ganglion cell layer, and nerve fiber layer within the retinal structure.²³ Although the entire optic nerve of the rabbit is myelinated, the retinal ganglion cells acquire a myelin sheath structure only after traversing the intraretinal environment.²⁴ Consequently, the nerve fibers in the retina do not possess a myelin sheath structure, and therefore the thickness of the GCC is not influenced by myelination. Moreover, in vivo OCT measurements have been extensively used in various rabbit models to monitor changes in retinal thickness.²⁵ SD-OCT recordings were performed at five distinct time points: 3 days prior to surgery and 1, 2, 3, and 4 weeks postsurgery. The experimental rabbits’ pupils were dilated following the previously outlined protocols. SD-OCT images were acquired using a Heidelberg Spectralis OCT system (Heidelberg Engineering, Heidelberg, Germany) with a light source at 870 nm. The axial resolution of the OCT scanner was 12 µm. The GCC thickness was measured vertically at a distance of 4000 µm from the center of the optic nerve papilla (ONH). 

Immunofluorescence Staining of Retinal Ganglion Cells

Subsequent to anesthesia, the experimental rabbits were humanely euthanized via an intravenous injection of potassium chloride (10%, 20 mL). Both eyes were carefully extracted in their entirety, and then the cornea was incised along the corneal edge. The lens was extracted, and the remaining eyes were immersed in 4% paraformaldehyde for 24 hours at 4°C. After fixation, the eyes were uniformly sectioned into a four-leaf clover shape based on the location of the optic papilla, and eight circular retinal slices were obtained from the four lobes utilizing a corneal ring drill blade. Two slices were extracted from each lobe, from the nearest to the farthest point, with the optic papilla serving as the circle's center. Immunofluorescence staining with RBPMS (a specific marker of RGCs) was subsequently conducted, and all stained retinal discs were imaged using a Zeiss LSM710 system (Carl Zeiss Meditec, Sartrouville, Germany) under a 20× objective for confocal images. For the manual counting of RBPMS-positive RGC cells, the ImageJ software (National Institutes of Health, Bethesda, MD, USA) was utilized, and the average of the cell counts from the two images of each disc was recorded as the final reading. 

ON Semi-Thin Sections and Quantification of Surviving Axons

The frontal skull and mesencephalon were meticulously dissected to expose the optic chiasm. Subsequently, the optic nerve was exposed. Using a biting forceps, the posterior orbital wall was bitten off, facilitating the careful and gradual separation of the optic nerve from the orbital contents. The isolated optic nerve was then sectioned into tissue samples approximately 1 mm in length using a razor blade. In the experimental group, samples were obtained at the proximal bulbar end of the decompression, the decompression site, and the distal bulbar end of the decompression. In the control group, samples were obtained 1 mm and 3 mm posterior to the eye. 

Following sampling, the tissue blocks were gently rinsed with PBS to eliminate any residual blood and hair. Subsequently, the cleaned tissue blocks were immersed in 2% electron microscope fixative (Wuhan Servicebio Technology Co., Ltd, Wuhan, China) and promptly prepared for further analysis. The tissue blocks underwent a process of gradient dehydration, resin infiltration, and embedding. Excess resin blocks were then removed, and thin slices with a thickness of approximately 1.5 µm were meticulously cut. These semi-thin sections were stained with 1% p-phenylenediamine (PPD) to enhance visualization.²⁶ When examining intact optic nerves, myelin sheaths appeared as dark rings, but the axoplasm was not stained. However, in previously damaged optic nerves, dark brown discs were observed interspersed with the axons.²⁷ The stained sections were then captured using a DM4B microscope under a 100× oil objective. 

The semi-thin circular sections of the optic nerve were obtained at the center of the circle and at the midpoint of the line from the center to the edge, resulting in a total of nine evenly divided areas. Individual images were captured at a size of 125.22 × 94.26 µm, and the number of surviving axons was manually counted using ImageJ software. The average count from the nine regions was considered the final number of surviving axons. The fabricated optic nerve samples were photographed using laser scanning confocal microscopy (A1-RHD25; Nikon, Tokyo, Japan), with three random positions selected for each sample. Pictures were taken at a magnification of 200×. The area of nonapoptotic optic nerve axons was measured using ImageJ software, and statistical analysis was performed. To ensure objectivity, the researchers strictly adhered to a double-blind principle during the counting process. 

Statistics

All data and graphs were analyzed using GraphPad Prism 7.0 (GraphPad Software, La Jolla, CA, USA). Numerical data are presented as means ± SEM in the figures. For multiple comparisons, one-way ANOVA followed by Dunnett's multiple comparisons test was employed. An independent-samples t-test was used to compare the special group with the control group. Statistical significance was denoted as follows: *P < 0.1, **P < 0.05, ***P < 0.01, ****P < 0.001. 

Results

Halting of GCC Thickness Thinning Following Optic Nerve Decompression

After the onset of optic nerve compression, a gradual decline in GCC thickness was observed over time, as demonstrated in our previous publication.²² At 4 weeks postonset (wpo) without decompression surgery, the GCC thickness had reduced to 73.0% (NO DC group in Fig. 1C). This represents the lowest point of GCC reduction under continuous compression. Our earlier studies have shown that this level remains consistent even after 3 months of compression.²² However, the decline in GCC thickness can be halted following decompression. Specifically, at 4 weeks after decompression, the GCC thickness was approximately 92.0% of the baseline for group DC-1d, 92.0% for group DC-3d, 86.4% for group DC-7d, 81.8% for group DC-2w, and 82.7% for group DC-3w. The thinning of the GCC in group DC-1d was statistically significant at 2, 3, and 4 weeks postonset compared to the no-decompression group. Similarly, the GCC thinning in group DC-3d was statistically significant at 3 and 4 weeks postonset compared to the no-decompression group. These results indicate that decompression on postonset day 1 and day 3 can significantly maintain GCC thickness and prevent further thinning. However, in the DC-7d, 2w, and 3w groups, there was no statistically significant difference in GCC thinning, although the trend suggested that thinning was inhibited. This lack of statistical significance could be due to the later timing of the intervention. 

Significant Reduction in Surviving RGCs Following Delayed Optic Nerve Decompression

The quantification of surviving RGCs in retinal whole mounts of different groups was expressed as a percentage compared with the contralateral eye. At 4 weeks postonset of initial compression, the surviving RGCs were approximately 78.2% ± 7.0% of the contralateral eye for group DC-1d, 60.5% ± 9.6% for group DC-3d, 31.7% ± 8.0% for group DC-7d, 26.2% ± 11.7% for group DC-2w, and 17.4% ± 4.7% for group DC-3w. The percentage of surviving RGCs significantly differed for group DC-7d (P = 0.007), DC-2w (P = 0.01), and DC-3w (P = 0.01) compared to baseline. However, despite the fact that the data for the DC-1d and DC-3d groups also showed a downward trend compared to baseline, with the DC-3d group's percentage being lower than that of the DC-1d group, there was no statistical difference. This suggests a notable difference in the protective effects on RGCs between decompression on postonset day 1 and day 3 and the groups that received decompression after 7 days. Moreover, the percentage of surviving RGCs at 4 weeks postonset exhibited an exponential correlation with the onset time of decompression (Fig. 2B). Specifically, a later decompression time was associated with a reduced number of surviving RGCs. This relationship was represented by the formula Y = 0.74e^(–0.22X) + 0.19, with an R² value of 0.72. 

Figure 2.

 

Analysis of surviving retinal ganglion cells and regression curves after optic nerve decompression. (A) Confocal images depicting surviving RGCs (red) in various retinal areas at 4 weeks after optic nerve decompression (4-wpo) in different decompression groups. Scale bar: 100 µm. (B) Exponential regression curves illustrating the percentage of surviving RGCs (n = 3–8) at 4 weeks after optic nerve decompression (4-wpo) in relation to time. Data are presented as mean ± SEM. *P < 0.1, **P < 0.05, ***P < 0.01, ****P < 0.001, Tukey's multiple comparisons test.

Significant Reduction in Surviving ON Axons Following Delayed Optic Nerve Decompression

The morphology of the ON exhibited a trend of decreasing axon density and increased scarring from group DC-1d, DC-3d, DC-7d, and DC-2w to DC-3w (Fig. 3A). The quantification of surviving axons of the ON in different groups was expressed as a percentage compared with the contralateral eye. At 4 weeks postonset of initial compression, the surviving axons were approximately 65.2% ± 17.7% of the contralateral eye for group DC-1d, 70.7% ± 9.5% for group DC-3d, 31.0% ± 3.3% for group DC-7d, 10.7% ± 1.3% for group DC-2w, and 6.7% ± 1.4% for group DC-3w. The percentage of surviving axons significantly differed for group DC-2w (P = 0.02) and DC-3w (P = 0.03) compared to baseline. Similar to the RGC apoptosis data, the percentages for groups treated before 7 days (DC-1d and DC-3d) did not differ significantly from baseline, suggesting that 7 days is a critical time point. Moreover, the percentage of surviving ONs at 4 weeks postonset exhibited an exponential correlation with the onset time of decompression (Fig. 3B). Specifically, a delayed decompression time was associated with a reduced number of surviving axons. This relationship was represented by the formula Y = 0.89e^(–0.1X) – 0.09, with an R² value of 0.78. 

Figure 3.

 

Analysis of surviving ON axons and regression curves after optic nerve decompression. (A) Light microscope images of transverse section of ON stained with PPD of different groups. Scale bar: 20 µm. (B) Exponential regression curves illustrating the percentage of surviving ON axons (n = 3–4) at 4 weeks after optic nerve decompression (4-wpo) in relation to time. Data are presented as mean ± SEM. *P < 0.1, **P < 0.05, ***P < 0.01, ****P < 0.001, Tukey's multiple comparisons test.

Delayed Optic Nerve Decompression Leads to Reduced Size of Surviving Axons

The surviving axons area demonstrated a reduction in size following delayed optic nerve decompression, as depicted in Figure 4A. The surviving axons area was calculated, with the mean area at 4-wpo measuring about 2.523 µm² for group DC-1d, 1.471 µm² for group DC-3d, 1.026 µm² for group DC-7d, 0.893 µm² for group DC-2w, and 0.7816 µm² for group DC-3w. There was a significant difference between each pair of groups, except for the comparison between group DC-2w versus group DC-3w and group DC-2w versus group DC-7d (Fig. 4B). 

Figure 4.

 

The distribution of area of surviving ON axons. (A) Light microscope images: the upper panel shows enlarged images of the optic nerve axons from different groups. The lower panel outlines the area of nonapoptotic optic nerve axons in yellow. Scale bar: 5 µm. (B) Mean area of nonapoptotic optic nerve axons (n = 3–4) in different groups at 4 weeks after optic nerve decompression (4-wpo). Data are presented as mean ± SEM. *P < 0.1, **P < 0.05, ***P < 0.01, ****P < 0.001, Tukey's multiple comparisons test. (C) The violin plot showing the distribution of the area of nonapoptotic optic nerve axons (n = 3–4) in different groups at 4 weeks after optic nerve decompression (4-wpo). The yellow line represents the median. The upper dashed line represents the third quartile, and the lower dashed line represents the first quartile.

The distribution of area was also depicted in Figure 4C, showing a similar pattern across different groups. The figure clearly illustrates that the proportion of axons with an area greater than 2 µm² decreases as the decompression time is delayed. Specifically, the DC-1d group has the highest proportion of large axons, followed by the DC-3d group, with the DC-3w group having the lowest. Therefore, the area distribution demonstrated that the earlier the decompression time, the larger the area of surviving axons. 

Discussion

The etiology and severity spectrum of compressive optic neuropathy are multifaceted,⁸ leading to its classification into various subgroups: mild or severe, transient or persistent, and insidious or sudden onset, resulting in eight different types of compression. Previous studies have indicated that mild ON compression only partially impairs the optic nerve, even after 3 months of sustained compression.²² As per clinical recommendations, mild compression may not necessitate urgent decompression surgery and can be scheduled at a suitable time or simply observed if no worsening occurs. Therefore, in this study, we focus on severe compression. 

Due to its self-resolving nature, transient compression does not necessitate subsequent decompression surgery and was therefore excluded from our experimental design. Consequently, this study focuses solely on severe and persistent compression. To ensure simplicity and reproducibility, we did not distinguish between insidious or sudden onset of compression. Readers should take this consideration into account when interpreting our results. 

Conversely, severe compression, characterized by rapid degeneration of retinal ganglion cells and optic nerve axons, intuitively warrants prompt surgical intervention. Economic constraints, limited medical resources, and uncertainties regarding the efficacy of surgery may lead some patients to postpone immediate surgical intervention.^15,16 Consequently, a critical question arises: should we endorse the patient's decision to delay decompression surgery for several days or provide more assertive guidance, such as advocating for immediate surgery? Additionally, if a surgical delay is deemed acceptable, what is the maximum permissible duration? To bridge the gap between clinical intuition and empirical evidence, we conducted animal experiments to gain insights. 

The Beneficial Effects of Decompression Surgery on Compressive Optic Neuropathy

In this investigation, we induced compression on the optic nerve in line with a previous report²² and subsequently performed surgical decompression to replicate clinical decompression procedures. Our findings revealed that decompression surgery effectively halted the progressive atrophy of the ganglion cell complex, as depicted in Figure 1, consistent with observations from surgical cases. Yumoto et al.¹⁸ documented eight patients with severe visual defects caused by compression of the optic nerve from ethmoid mucoceles. All of these patients underwent decompression surgeries. Among them, five experienced improvements in visual acuity, while the remaining three showed no change (two cases of no light perception [NLP] and one case of hand motion [HM] vision). Notably, endoscopic tumor removal surgeries targeting the lateral orbital apex resulted in visual improvement in approximately 71% of patients with vision loss of varying degrees.²⁸ Furthermore, the removal of apical orbital tumors through an endoscopic transethmoidal approach led to visual acuity improvement in 11 of 12 patients.²⁹ Dysthyroid optic neuropathy, a prevalent condition responsible for over 90% of related cases involving nerve compression due to enlarged extraocular muscles, often demonstrated favorable outcomes following orbital decompression surgery, even in cases of severe vision loss.³⁰ Similar findings have been observed in other studies, underscoring the importance of decompression surgery for favorable visual outcomes.^8,21 Collectively, these findings underscore the beneficial impact of decompression surgery on the optic nerve, mitigating the ongoing impairment resulting from compression. 

Exponential Degeneration of RGC and Axons Under Compression and the Critical Timing of Decompression

Our study has shown that when decompression surgery is performed 1 week after the onset of compression, only about 30% of retinal ganglion cells (31.7% in group DC-7d) and the optic nerve (31.0% in group DC-7d) survive. The rapid rate of degeneration of RGCs and axons under severe compression was found to be exceptionally swift, with the survival curve indicating a mere 4-day half-life period for surviving RGCs. Notably, the survival of RGCs and ONs was found to be exponentially correlated with the timing of decompression, suggesting that the speed of degeneration is most pronounced in the initial days following compression. This trend is consistent with findings from spinal injury studies, where delayed decompression after 3 days resulted in significant cavitation of the spinal cord.³¹ 

In a systematic review by Carlson et al.,²⁰ dichotomous data revealed a significant difference in the odds of improvement for shorter versus longer compression durations.^32,33 However, linear regression analysis from continuous data did not show a significant association between duration and the probability of vision improvement.^17–19 This discrepancy may be attributed to the prolonged cutoff times used in the dichotomous studies, with most employing 1 year as the cutoff, potentially masking the impact of earlier intervention. Our results indicated that the thickness of GCC stabilizes at a low level after 1 month of compression (Figs. 2, 3), suggesting the potential need for timely decompression surgery, particularly in cases of severe compression. However, given the limitations in extrapolating from animal models, the recommendation for timely surgery should be interpreted with caution by clinicians, who should consider patient- and disease-specific circumstances. 

Differential Axon Degeneration Under CON: Larger Axons Vulnerable, Smaller Axons Resilient

The data presented in Figure 4 illustrate that the DC-1d group exhibits the highest proportion of larger-diameter axons. Furthermore, it is evident that as the onset time of decompression increases, there is a greater loss of larger axons. These findings uncover an intriguing phenomenon wherein larger axons demonstrate earlier degeneration in comparison to smaller axons (see Fig. 4). This suggests that smaller axons display a greater resilience against compression, a trend that has also been observed in glaucoma studies. 

Early investigations into glaucoma, involving both human and nonhuman primate studies, revealed that RGCs with larger cell bodies and axons were more susceptible to early injury.^34,35 For instance, in 1987, researchers conducted a chronic experimental glaucoma study on monkeys, observing that axons larger than the normal mean diameter atrophied more rapidly in glaucomatous eyes. Similarly, in 2000, histologic examinations of human glaucoma subjects indicated a significantly higher presence of smaller axons compared to larger axons in optic nerves.¹⁶ Additionally, further studies by Glovinsky et al.³⁶ on monkeys demonstrated a greater reduction in the size of RGCs in the whole-mount retina, particularly larger-diameter RGCs, which are presumed to give rise to larger-diameter axons. 

It is important to note that the size of the soma or axon does not necessarily indicate a specific type of RGC. While there have been numerous studies on magnocellular versus parvocellular layers^37–39 and Off versus ON RGCs,⁴⁰ the susceptibility of specific types of RGCs to ocular hypertension remains a topic of ongoing debate.⁴¹ 

Potential Contributions of Surgical Trauma, Inflammation, and Ischemia to Observed Outcomes

In our previous studies on the model of optic nerve compression injury,²² we used an artificial optic canal implant without any protrusions as the surgical control group. OCT measurements of the retinal GCC thickness were taken at 1, 2, 3, and 4 weeks postsurgery. A downward trend was observed only at 3 weeks postsurgery, but the decrease was merely 0.05% compared to the preoperative level, with no significant difference statistically. Electrophysiologic flash visual evoked potential (F-VEP) data also showed no significant differences in P2 amplitude at various postoperative time points compared to presurgery. Therefore, the surgical trauma and secondary damage caused by the implantation surgery and implants in this study can be considered negligible in subsequent analyses. 

In the NO-DC group, we observed a chronic and progressive apoptosis of optic nerve axons and RGC bodies. Thus, we hypothesize that apart from the optic nerve compression injury caused by the artificial optic nerve implants, the subsequent local inflammation and ischemic responses around the optic nerve and microenvironment could also contribute to persistent optic nerve damage. Similar secondary inflammatory and ischemic responses have been confirmed in studies of optic nerve damage caused by glaucoma,^42,43 tumors,⁴⁴ and trauma.⁴⁵ Additionally, gene sequencing work related to optic nerve injury published from our laboratory also found that upregulated genes postinjury are primarily enriched in inflammation and ischemia pathways.⁴⁶ It has been demonstrated that neuroinflammatory responses in the visual pathway involve multiple molecules and signaling pathways, including the close interactions via autocrine/paracrine routes among glial cells, immune cells, vascular endothelial cells, and retinal ganglion cells.^42,45 However, the precise timing and pathways through which inflammation and ischemic responses impact optic nerve axon degeneration and long-term RGC apoptosis postinjury require further investigation—one of the future directions for our research group. 

Limitations

In this study, we focused exclusively on severe and persistent optic nerve compression without distinguishing between insidious or sudden onset. Addressing insidious compression requires an intricate design to achieve a gradual increase in compression over several months, which demands extreme precision given the optic nerve's approximate 2-mm diameter. However, since insidious compression is also common in clinical practice, we will endeavor to investigate this type in future studies. 

Conclusions

In conclusion, the persistent substantial compression on the optic nerve can lead to exponential degeneration of the optic nerve, particularly affecting larger optic nerve fibers initially. This underscores the critical importance of timely decompression surgery to preserve RGCs and axons. Early intervention aimed at relieving the compression on the optic nerve may offer potential benefits in mitigating the degenerative effects and conserving visual function. 

Acknowledgments

Supported by National Key Research and Development Program of China (2021YFA1101200). 

Disclosure: X. Jiang, None; B. Xu, None; S. Yao, None; Z. Wang, None; M. Liu, None; Y. Zhang, None; W. Wu, None; E. Wu, None