Nonarteritic anterior ischemic optic neuropathy (NAION) is the most common cause of sudden optic nerve-related vision loss and typically occurs in individuals over 55 years of age (1). As reported, the prevalence of NAION in the general population older than 40 was 102.87 per 100 000 people (2). The actual etiology of NAION is controversial. A crowded disk is an acknowledged risk factor (3). Additionally, there are many other systemic risk factors associated with NAION, such as hypertension, hyperlipidemia, atherosclerosis, diabetes mellitus, episodic hypotension, and obstructive sleep apnea (4–10). It is generally believed that NAION is due to acute ischemia of the optic nerve head (ONH) supplied from the posterior ciliary artery circulation (11). Repeated studies demonstrated the loss of peripapillary capillary (RPC) and retinal nerve fiber layers (RNFL) predominantly in the superior sectors from the acute to the chronic phase (12, 13). Changes in vascular and neural structures seemed to correlate with the severity of visual field (VF) defects (12–15).

Although NAION is different from inflammatory optic nerve diseases such as optic neuritis, some evidence has shown that the inflammatory response is involved in the pathological process of NAION. Studies have shown the accumulation of inflammatory cells—both intrinsic, activated microglia and extrinsic macrophages—in rat models of NAION (9). The same findings were observed in a human histological specimen (16) and primate models (17). Histopathologic studies of human specimens found that ischemic optic nerve lesions were initially acellular and later showed macrophage infiltration (15). Other histopathologic findings also suggested the destruction of the blood retinal barrier. Some studies have shown that systemic macrophages, rather than intrinsic microglia, are involved in eliminating degenerating myelin from fragmenting axons (18). This could be a potential therapeutic target to protect the optic nerve and reduce the loss of retinal ganglion cells.

Recently, some groups have reported that macrophage-like cells (MLCs) located at the inner limiting membrane (ILM) were successfully observed in live human retina using adaptive optics optical coherence tomography (AO-OCT) and clinically used optical coherence tomography (OCT) (19–22). MLCs increased in many retinal diseases, including diabetic retinopathy (19), retinal vein occlusion (23, 24), uveitis (25) and central retinal artery occlusion (26). A study based on animal models suggested that MLCs are a potential biomarker for inflammation during retinal vascular disease (27).

Although the observation of inflammatory cells in optic nerve axons is difficult with a clinical OCT device, investigation of peripapillary macrophages located above the vitreous retinal interface can provide new insights into the role of inflammation in NAION. To the best of our knowledge, this study is the first to investigate MLCs in the ONH region in NAION eyes with en face OCT. We aimed to illuminate the number and distribution of macrophage-like cells near the optic disc and the relationship between the cells and the changes in other optical coherence tomography angiography (OCTA) parameters. This study will provide clinical evidence for further understanding the pathological process of NAION.

A total of 25 patients were included in the current study. Twenty-five acute NAION eyes and 25 normal fellow eyes were evaluated. The time interval from NIAON onset to en face OCT acquisition was 9.56 ± 4.30 days (mean± standard deviation). The demographic characteristics, best corrected visual acuity (BCVA), the results of automated perimetry and risk factors for NAION are summarized in Table 1 .

MLCs distributed unevenly around the ONH in both NIAON eyes and normal eyes. In two-ways repeated measures ANOVA, c a higher MLC density in the nasal distribution was found in both NAION eyes (9.62 ± 5.74 vs. 14.35 ± 8.13, p=0.026) and normal eyes (6.74 ± 4.99 vs. 11.83 ± 8.32, p=0.026) when comparing to the temporal. Additionally, more MLCs in the control group were found in the inferior than in the superior region (9.95 ± 6.50 vs. 8.69 ± 5.92, p=0.021). In the affected group, the MLC density was slightly higher in the superior region, and no difference was found between the inferior and superior regions (11.01 ± 6.27 vs. 12.91 ± 7.97, p=0.086) ( Figure 3 ). The MLC count and density showed large variation between individuals, but those values for NAION eyes compared to fellow eyes showed a positive correlation (r=0.612, p=0.001). In addition, the same results were found in the nasal (r=0.582, p=0.002), temporal (r=0.517, p=0.008), superior (r=0.495, p=0.012) and inferior (r=0.619, p=0.001) subregions ( Figure 4 ).

Despite the similar distribution characteristic, MLC density increased significantly in the acute NAION eyes (NAION vs. normal, 11.97 ± 6.66 vs. 9.31 ± 6.10 cell/mm², p=0.028) (examples shown in Figure 5 ). Results of paired sample T test between two groups were shown in Table 2 . In two-ways repeated measures ANOVA, we found that the increase of MLC density was significant in superior, nasal and temporal subregions (NAION vs. normal, p=0.008, p=0.001, p=0.001, respectively). No statistical difference was found between the two groups in inferior region (NAION vs. normal, 11.01 ± 6.27 vs. 9.95 ± 6.50, p=0.315). In NAION group, the increase in the MLC density was uneven among the subregions. The amount of change in the superior subregions was greater than that in the inferior sides (4.13 ± 7.49 vs. 0.94 ± 5.21 cell/mm², p=0.001) ( Figure 6 ). The same result was not found in the comparison between the temporal sector and the nasal sector (3.23 ± 5.69 vs. 1.68 ± 7.87 cell/mm², p=0.301). In addition, the VD of RPC decreased significantly in the affected eyes (NAION vs. normal, 37.54 ± 5.25 vs. 40.56 ± 4.25, p=0.016). In the sectorial analysis, we found that the VD of RPC showed a reduction predominantly in the superior sectors and the temporal sectors when compared to the inferior sectors and the nasal sectors, respectively (superior vs. inferior, -3.54 ± 6.71 vs. -0.37 ± 8.07, p=0.004; temporal vs. nasal, -2.69 ± 8.72 vs. -1.22 ± 6.06, p=0.005) ( Figure 5 , 6 ). The loss of RPC seems to correspond to the increase in the MLC density. Compared with the fellow eye, RNFL thickness increased significantly in four sectors ( Table 2 ). There were no significant differences in the increase in RNFL thickness among the four sectors (p=0.414).

We further investigated the MLC density, VD of RPC and RNFL thickness and their difference between bilateral eyes in subregions according to the VF defect. Most patients (n=21) had inferior VF defect with superior ONH affected and others had superior VF defect with inferior ONH affected. Compared with unaffected regions, significantly increased MLC density was found in affected regions. There was a statistical difference in ΔMLC density (MLC density in NAION eye minus MLC density in fellow normal eye) and ΔVD of RPC (VD of RPC in NAION eye minus VD of RPC in fellow normal eye) between the affected and unaffected subregions. The results are shown in Table 3 .

There was no correlation among the decrease in VD of RPC, the increase in RNFL thickness and the increase in MLC density after controlling the age, sex and disease duration

MLCs around the ONH characterized by en face OCT were detected in the current study. To the best of our knowledge, this is the first study to investigate macrophage-like cells in vivo with acute NAION. We explored the cell distribution characteristics around the ONH, the uneven increase in MLCs and the correlation among the increase in cell density, the decrease in VD of RPC and the VF defect.

In our study, we found that whether in normal eyes or in NAION eyes, the distribution of cells around the optic disc followed a certain pattern. Significantly higher cell density was found in the nasal sectors than in the temporal sectors. In the normal eyes, more MLCs located in the inferior sectors than in the superior sectors. Due to the greater increase in the superior sectors, the same pattern was not found in the affected group. Repeated studies have demonstrated that MLCs located unevenly on the vitreoretinal interface (VRI). A previous study found that the cell density was higher at the macular temporal than at the ONH (20). In our observation of RVO patients, the en face OCT images centered on the macular illustrated that most MLCs in the normal eyes were found in the perifovea region, and only a few cells were found within 3 mm of the fovea. A recent study in a mouse model that combined OCT and confocal immunofluorescence found that the cells located above the membrane are heterogeneous (27). Cells with different identities are not evenly distributed on the VRI. Wild field en face OCT is needed to explore the distribution characteristics of human retina MLCs. Another interesting finding is that although individual differences in cell numbers were evident, there was a positive correlation between the number of cells in both eyes. This result was true in all subregions around the ONH, which was consistent with our previous findings (24). It would be better to design a paired experiment in the MLC investigation to eliminate differences between individuals.

We found that MLC density in the ONH region increased significantly in the NAION group, especially in the affected subregions corresponding to the VF defect. According to previous studies, inflammation plays a role in the pathological process of NAION. Researchers found that the neutrophil-to-lymphocyte ratio and some cytokines, chemokines, and growth factors in the peripheral blood of NAION patients were significantly higher than those of the controls (28) (29). The inflammatory status of patients with NAION suggests that the onset and progression of the disease may be related to inflammation. The local inflammatory response was further studied in animal models. Similar to other central nervous system infarct and spinal cord injury models, NAION and sudden (optic nerve) ON ischemia result in early cytokine-mediated changes followed by sequential inflammatory cellular activation and infiltration. After the induction of ON ischemia, the breakdown of the blood-optic nerve barrier occurs within hours (30–32), followed by the recruitment of extrinsic macrophages and the activation of microglia in the ischemic region (16, 33). A study based on the primate model showed that the activation of the inflammatory response may be related to axon loss (16). Histopathologic studies on human NAION specimens showed that the macrophages infiltrated the affected optic nerve (15). However, the function of inflammation is still controversial. Macrophage activity can be either neurodegenerative and/or neuroprotective in ischemic stroke (34). Early methylprednisolone treatment and granulocyte colony-stimulating factor can be effective in reducing macrophage recruitment and ON vascular permeability and resulting in more survival retinal ganglion cells (30, 35). In the current study, we found an increase in MLCs in living NAION patients using a clinical OCT device, which was consistent with previous studies. MLC on the VRI was observed in several retinal diseases, including diabetic retinopathy (19), retinal vein occlusion (23), uveitis (25), retinal artery occlusion (24, 26) and posterior vitreous detachment (36). Some researchers considered the MLCs above the VRI as potential biomarkers of inflammation during retinal vascular disease (27). Our findings provide clinical insights into the involvement of inflammation in the postischemic pathological process in NAION. However, we did not determine the definite identity and function of these cells in the cross-sessional clinical study. Further studies on animal models combining clinical OCT and confocal immunofluorescence are needed to investigate the exact mechanism.

Acute ischemia of the optic nerve is always followed by severe swelling of the peripapillary RNFL, loss of retinal ganglion cells and loss of the peripapillary RPC, which has been demonstrated by clinical studies based on OCT and OCTA and experimental studies based on animal models (12, 13, 37–44). A few studies found that the change in the RNFL and RPC was predominantly in superior sectors, corresponding to visual field defects (14). In sectorial analysis, we also found that compared with the fellow eye, the decrease in VD of RPC was more significant in the superior sector and the temporal sector. Analysis based on VF affected hemisphere and unaffected hemisphere, the loss in VD of RPC was more in the affected subregions, relating to the VF defect. We speculated that the recruitment of macrophages and the inflammatory response are more severe in the injured area. There was no statistical difference in RNFL thickness between the affected hemisphere and unaffected hemisphere. The mean disease duration was 9.56 days in our study. Previous study demonstrated that in acute NAION, initially the involved part of the disc has edema, followed by the entire disc showing generalized edema in several days (9). Because few patients in very early acute stage were included, the difference were not found in sector analysis of RNFL thickness.

The study is insufficient to draw the exact relationship among the decrease in VD, the loss of RNFL and the aggregation of MLCs. And no correlation was found even after adjusting for patients’ age and disease duration. In NAION eyes, the edema of ONH occurred in the acute stage and the atrophy developed after two months from onset (9, 12). RPC was also gradually lost during this process. Although the NAION eyes showed decreased VD of RPC in the early stage, some study found greater loss of RPC in the chronic stage (42). Considering that during disease duration the RNFL thickness and VD of RPC were changing, it was explainable that no correlation conclusion was drawn in the current cross-sectional study. And the sample size also needs to be further expanded. In addition, follow-up of NAION patients is required to explore the changes in MLCs with the prolongation of the disease course and find out the relationship between the aggregation of MLCs and the loss of RNFL and RPC.

There are some limitations in our study. First, because of the imaging technology, it was difficult to use blinding in processing the MLC images, and it may have some effect on a few images needing manual correction. Second, to cover a larger area, the HD 6 × 6 mm image on the ONH was used in the study. Because the VD of RPC cannot be calculated in the built-in software in the OCT device, the VD in our study was calculated by ImageJ with thresholding strategy used in a previous study. Different methods may influence the values of VD measurement. However, the characteristics of VD in the study are similar to those in other NAION studies. Last but not least, the study did not demonstrate the origin and function of the MLCs on VRI. We only included acute and treatment-naïve subjects. Further follow-up studies may focus on the response to steroids, the behavior of MLCs in the atrophic stage and the impact on visual outcomes.

In general, a significantly increasing density of MLCs on the ILM was found in acute NAION eyes. We also found a potential correlation between the increasing number of MLCs and the loss of RPC. The investigation of MLCs provided insights from clinical studies that inflammation takes part in disease progression and may be helpful for understanding the pathophysiology and for finding therapeutic targets for neuroprotection in NAION. Last but not least, the observation of MLCs is a simple and noninvasive method to evaluate the severity of the inflammatory response in daily clinical practice. As an imaging biomarker, MLCs might contribute to estimating the prognosis of NAION and guiding treatment in the clinic.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by The Institutional Review Board of the Renmin Hospital of Wuhan University (WDRY2021-k162). The patients/participants provided their written informed consent to participate in this study.

CC provided guidance and overall direction throughout the study design, paper writing and revision. WW handled the literature review, data collection, data analysis and interpretation, and manuscript writing. ZY provided expertise on data collection and review the manuscript. XW collected data and reviewed the manuscript. HL collected data and provided guidance on writing. All authors contributed to the article and approved the submitted version.