Glaucoma is a leading cause of blindness globally [1], with neovascular glaucoma (NVG) being one of the most complicated forms of this illness. The primary pathological manifestations of patients with NVG include hyperplasia of intraocular fibrovascular tissue, obstruction of the trabecular meshwork, and peripheral anterior iris adhesion, which could lead to the obstruction of aqueous humor outflow and an increase in intraocular pressure [2] (Fig. 1). NVG can cause deficits in visual function as well as severe pain, both of which will affect quality of life [3]. Clinically, NVG is divided into three stages: Rubeosis, invasion of anterior chamber angle without closure and invasion of anterior chamber angle with closure. NVG is often secondary to diabetic retinopathy (DR), central retinal vein occlusion (CRVO), intravascular inflammation, and tumors [4, 2]. The production of vascular endothelial growth factor (VEGF) secondary to retinal ischemia is the most common causes of NVG [5, 6]. VEGF [7, 8], Ephrin type-A receptor 2 (EphA2) [9], platelet derived growth factor (PDGF) [10], hypoxia-inducible factor-$\alpha$ (HIF-$\alpha$) [11], and fibroblast growth factors (FGFs) [12, 13] are all considered to be related to the onset of NVG.

Fig. 1.

Examples FC and FFA of NVG in a patient with ischemic central retinal vein occlusion. FC, fundus camera; FFA, fluorescence fundus angiography.

Due to the complex etiology and pathogenesis of NVG, there are no precise indicators or guidelines which apply to the development and outcomes of this disease. Therefore, clinical treatment is complicated, and etiological therapies mainly included the elimination of retinal ischemia by photocoagular, cryotherapy and anti-VEGF therapy. Principles of clinical treatment include treatment of the primary diseases and timely elimination of retinal ischemia to control clinical symptoms. The core strategy of therapy involves reduction of intraocular pressure and retinal ischemic maculopathy, and eliminating neovascularization.

Magnetic resonance imaging (MRI) is well established in clinical work as a highly sensitive, non-invasive method of imaging soft tissues in the human body and observing their morphology and pathological changes. MRI offers advantages over X-ray or other imaging methods [14], and functional MRI (fMRI) allows voxel-based morphometry (VBM) for the analysis of brain morphology in fine spatial detail. VBM provides measurement of any changes in local gray matter volume (GMV), white matter volume (WMV) and the volume of the whole brain. Altered morphology in brain tissue may reflect the mechanism, development trend and clinical manifestations of disease, providing important information for practical clinical work. At present, in addition to cognitive diseases, VBM has been applied in ophthalmic diseases, including optic neuromyelitis [15], advanced monocular blindness [16], amblyopia [17], comitant strabismus [18], glaucoma [19], and acute eye pain [20] (Table 1, Ref. [15, 16, 18, 20, 21, 22, 23]).

Early studies have shown extensive and complex changes to cortical structure in glaucoma patients, and that these were significantly correlated with progression of the disease [24]. However, there is no grading system or other method for assessing changes in brain morphology in patients with NVG. VBM analysis was applied in the present research to explore whether there is a significant discrepancy of cortical volume between NVG patients and matched healthy controls (HCs), and to look for associations between those changes and the clinical manifestations and pathogenesis of NVG.

Nineteen patients (9 males, 10 females) with bilateral NVG (secondary to ischemic central retinal vein occlusion) who had been diagnosed by the Ophthalmology Department of the First Affiliated Hospital of Nanchang University (Nanchang, China) were recruited to this research. Inclusion criteria for patients with NVG included: (1) slit lamp microscopy or gonioscopy examination shows neovascularization in the surface of iris and (or) anterior chamber angle; (2) intraocular pressure (IOP) $\ge$21 mmHg (mean of at least three non-contact tonometry measurements); (3) retinal ischemia indicated by fundus fluorescein angiography.

The exclusion criteria for this group included: (1) the presence of acute keratitis, acute conjunctivitis or any other ocular inflammation (2) retinal ischemia caused by additional angiopathy (such as carotid stenosis); (3) a history of drugs addiction or alcoholism; (4) any condition contraindicating an MRI scan; (5) mental disorder, including but not limited to depression and bipolar disorder.

Nineteen individuals matched for age, gender and educational level met the above exclusion criteria and had no abnormal brain structure, function or metabolism evident on MRI examination.

The Medical Ethics Committee of the First Affiliated Hospital of Nanchang University (Nanchang, China) gave their approval to the study. The study design was conducted in accordance with the Declaration of Helsinki. Prior to participant signed an informed consent declaration.

This is the first study of using VBM to assess gray matter volume abnormalities in NVG patients. VBM has, however, been widely used in research on both neurological and ophthalmological diseases, and has shown clear advantages in the exploration of the pathological mechanism of these diseases.

Studies have established that glaucoma patients generally show atrophy of specific brain areas or changes in functional connectivity between brain areas. Aditya et al. [19] showed varying degrees of volume reduction along the visual pathway in glaucoma. Laura et al. [26] showed reduced cortical thickness and volume in both primary visual cortex and association visual areas in glaucoma patients. Fukuda et al. [27] found decreased gray matter density (GMD) in primary open-angle glaucoma (POAG) patients, particularly in the visual cortical areas (Brodmann areas 17, 18, 19). In patients with POAG, Jiang et al. [28] found altered volume in brain regions which are highly correlated with the retinal nerve fiber layer. Similarly, Ziko et al. [29] found that VBM showed significantly decreased volume in the visual pathway and other brain areas which are related with the visual system. Li et al. [21] suggested that POAG patients with advanced disease show degenerative changes in visual-related cerebral cortex, dorsal and ventral visual pathways.

In this study, the NVG group experienced statistically significant GMV reductions when compared to the HCs in left cingulum anterior/medial frontal gyrus; left middle frontal gyrus, orbital part; and the right temporal: superior temporal gyrus/right frontal inferior orbital part.

The cingulate gyrus occupies a crucial position in the limbic system [30] and exerts an important role in emotional control and pain management. According to current studies, the anterior cingulate gyrus is involved in not only the regulation of heart rate or blood pressure, and other autonomic nerve activities [31]. But it’s also implicated in a variety of sophisticated body and visceral motor processes, as well as pain responses, attention maintenance [32], decision-making processes [33], and pain management [34]. Previous studies have demonstrated that the anterior cingulate gyrus cortex of the right dorsal horn may be related to visual cognition and emotion [35]. Results of the present study indicate decreased gray matter volume of anterior cingulate gyrus in NVG patients. This result suggesting that NVG patients may have dysfunction in visual cognitive function [22].

The frontal gyrus is an important component of the dorsal visual pathway [36]. It is concerned with recognition and judgement based on spatial information about objects and simultaneously plays an important role in optic nerve positioning, multi-retina correspondence, and the formation of scotoma of the fixation point [37]. The frontal gyrus includes superior, medial and inferior frontal gyrus, and abnormalities of its structure or function have been investigated in many ophthalmic diseases, including but not limited to diabetic retinopathy [38], strabismus with amblyopia [39], corneal ulcer [40] and ophthalmectomy [14]. In this study, we found that VBM results indicated significant GMV reduction in all of these regions (medial, superior, and inferior frontal gyrus), which are the same as Wang et al. [41] and it may be related to increased intraocular pressure (IOP).

The medial frontal gyrus is the medial side of the superior frontal gyrus, located anterior to the parasitical central lobule. Research suggests that the medial frontal gyrus is linked to the decision-making process and the subsequent execution process [42]. In addition, the frontal eye field (FEF) area in the center of the medial frontal gyrus regulates spontaneous eye movements and is also linked to emotional transfer [43]. The cortex of the medial frontal gyrus represents one of the components of the default model network which includes the inferior parietal cortex, medial and superior frontal gyrus and precuneus [44, 45]. Furthermore, it plays a fundamental role in specific emotion such as depression and anxiety [46]. The GMV of the medial frontal gyrus was diminished in this study, illustrating that patients with NVG may have visual pathways that are compromised. Jiang et al. [28] found that the frontal gyrus area of patients with primary open-angle glaucoma (POAG) shows increased volume and blood oxygenation level dependent (BOLD ) signals. They concluded that this may be due to edema of local cerebral tissue and microsomal activation, that increased blood flow may be a pathological compensatory response after structural and functional impairment, and that these results may indicate nerve damage in the frontal gyrus. This hypothesis may explain why Williams et al. [47] found increased medial frontal gyrus volume in glaucoma patients in early stage, which is contrary to this study.

The decrease in GMV volume in the superior and inferior frontal gyrus in the present study agrees with the results of other existing studies [21, 48], supporting the finding of damaged cortical visual pathways in NVG patients. These results collectively suggest that NVG patients may have abnormalities in eye movement, cognitive function and visual processing ability.

The superior temporal gyrus is involved in visual signal processing, auditory processing, auditory memory and other physiological processes. The superior and middle temporal gyri are thought to specialize in processing dorsal signals along the visual pathway [49]. Previous studies have observed abnormalities in the superior temporal gyrus in specific ophthalmopathy, including but not limited to optic neuritis [50], diabetic retinopathy [51], comitant strabismus [18], and retinal detachment [23]. In this study, the decrease in GMV in the superior temporal gyrus may indicate atrophy of temporal retinal fibers in NVG patients [29] (Table 4).

Notably, some of the studies also mentioned that beta-amyloid accumulation may be associated with altered GMV in NVG patients [52], which happens to be a typical pathological change in Alzheimer’s patients. There is still a plenty of room to be investigated.

The present study has the following limitations: (1) The participants included in this study may be insufficient. However, selection followed precise inclusion and exclusion criteria, and the differences between potentially confounding factors were not statistically significant; (2) The study samples are all derived from hospital outpatient data, which risks selection bias; (3) Due to equipment and technical limitations, the spatial resolution of imaging may not be optimal.

In summary, NVG patients show decreased GMV values in certain brain areas, the dysfunction of those brain areas may be related to the appearance of corresponding symptom in NVG patients such as the depression and anxiety.

Conceptualization, methodology, writing original draft preparation, writing reviewing and editing—HZ, YDS, RBL. Data curation, formal analysis—QMG, YCP, LJZ, HYS, QYL. Resources, methodology, supervision, projection administration, funding acquisition—YS.

All research methods were approved by the committee of the medical ethics of the First Affiliated Hospital of Nanchang University (2015016) and were in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All subjects were explained the purpose, method, potential risks and signed an informed consent form.

Not applicable.

This research was funded by National Natural Science Foundation of China, grant number 81660158, 81460092, and 81400372; Natural Science Key Project of Jiangxi Province, grant number 20161ACB21017; Health Development Planning Commission Science Foundation of Jiangxi Province, grant number 20175116.

The authors declare no conflict of interest.