Wen-Jia Dong, Ting Su, Chu-Qi Li, Yong-Qiang Shu, Wen-Qing Shi, You-Lan Min,Qing Yuan, Pei-Wen Zhu, Kang-Cheng Liu, Jing-Lin Yi, Yi Shao

1Affiliated Eye Hospital of Nanchang University, Nanchang 330006, Jiangxi Province, China

2Eye Institute of Xiamen University, Fujian Provincial Key Laboratory of Ophthalmology and Visual Science, Xiamen 361102, Fujian Province, China

3Department of Ophthalmology, the First Affiliated Hospital of Nanchang University, Jiangxi Province Clinical Ophthalmology Institute, Nanchang 330006, Jiangxi Province,China

4Department of Radiology, the First Affiliated Hospital of Nanchang University, Nanchang 330006, Jiangxi Province, China

Abstract

INTRODUCTION

Retinal vein occlusion (RVO), sometimes referred to as “eye stroke”, is a thrombosis obstruction in the retinal venous system, which might associate with the central or branch retinal vein[1], and characterized by intraretinal haemorrhages, various degrees of retinal vessel blockage and tortuosity, and cystoid macular edema[2]. The prevalence of epidemiological investigations varies from 0.1% to 0.5%among the middle-aged group and elderly group[3-5]. RVO becomes the 2ndcommon cause leading to vision deterioration among retinal vessels disorders following diabetic retinopathy.Regular ophthalmic examinations for RVO contain fundus fluorescein angiography and optical coherence tomography(OCT)[6], but there’s few tests focus on the neuroimaging.The exploration of RVO-related cerebral progress using neuroimaging is a brand-new prospect in visual neuroscience which could promote to unveil the underlying mechanisms.

Resting-state functional magnetic resonance imaging (rs-fMRI)is a broadly used noninvasive neuroimaging technology which is relied on cerebrum blood flow and metabolize analysis,and it has been improved progressively and enabled scientists to explore functional changes of specific cerebral regions at resting-state. The rs-fMRI has been applied to inspect the intrinsic alterations in several visual-related diseases on account of the blood-oxygen-level-dependent (BOLD) signal with several methods such as amplitude of low-frequency fluctuation (ALFF)[7], diffusion tensor imaging (DTI)[8]and regional homogeneity (ReHo)[9]. However, they cannot provide the data of functional connectivity in the whole-brain net system, and the underlying pathophysiological basis of the entire brain information processing remained unclear.

Voxel-wise degree centrality (DC) is an rs-fMRI analysis method that has been exploited to investigate the cerebrum functional connectome alterations[10]. The DC method detects the functional interrelationships between one cerebral area and the rest within the integral connectome at the voxel level, and a high DC value denotes a node with various direct connections to other nodes[11]. Thusly, DC is a superior network measurement than others for the reason that it calculates the sum of immediate connections in a meshwork for a given voxel and reveals the functional connectivity in cerebral network without a prior election. It has been effectively applied for exploring the concealed pathophysiological mechanisms of some other ocular disorders included glaucoma[12]and strabismus[13]. Hence, the DC is a reliable rs-fMRI technique that has not been utilized in RVO. This research designed to utilize the DC technology to further study the intrinsic cerebral activity of RVO comparing with healthy controls (HCs).

SUBJECTS AND METHODS

Ethical ApprovalThis study was approved by the Medical Ethics Committee of the First Affiliated Hospital of Nanchang University and complied with the Declaration of Helsinki. All participators were apprised the objective, content, potential risks, and signed informed consents.

ParticipantsA total of 21 RVO patients (11 males, 10 females) were enrolled in the First Affiliated Hospital of Nanchang University with the following inclusion criteria: 1)ophthalmoscopy showed ischemic RVO (Figure 1); 2) OCT showed mixed type of macular oedema within subretinal fluid; 3) fundus fluorescein angiography showed occlusion of retinal vein. Ischemic central retinal vein occlusion (CRVO)≥10-disc areas of retinal capillary non-perfusion. Ischemic branch retinal vein occlusion (BRVO) ≥5-disc areas of retinal capillary non-perfusion. The exclusion criteria of RVO were:1) any preceding ocular surgery history; 2) any proof of other ocular diseases; 3) any other systematic diseases.

Twenty-one HCs of comparable age and sex matched up with RVOs were recruited. Inclusion criteria were: 1) no history of ophthalmic diseases; 2) no psychiatric diseases, cerebral infarction diseases and other system diseases; 3) competent of magnetic resonance imaging (MRI) examination.

Magnetic Resonance Imaging Parameters and Data ProcessingMRI examination was scanned with a 3-Tesla MR scanner (Trio, Siemens, Munich, Germany). The whole-brain T1-weights were acquired with spoiled gradient-recalled echo sequence by utilizing these parameters in Table 1.

Figure 1 Fundus photography (A) and fundus angiography (B) in RVO patients.

Table 1 MRI scan parameters

Pre-filter all the data with MRIcro (www.MRIcro.com) and preprocess them with SPM8, DPARSFA and the Resting-state Data Analysis Toolkit. Gathering the residual 230 volumes after deleting the beginning 10 time points. Volumes with the x, y, or z directions ＞2 were eliminated.

Degree CentralityDC based on the individual voxel-wise functional network, was computed by calculating the number of significant threshold correlations between the subjects. The conversion of DC map of each subject has been described in our previous study[13].

Brain-behavior Correlation AnalysisDC values in diverse cerebrum areas between RVOs and HCs were categorized as region of interest (ROI). The correlationship analysis was implemented in RVO patients to inspect the interrelation between the DC signal of respective ROI and clinical features.P＜0.05 was assumed to be statistically significant.

Statistical AnalysisThe clinical features between RVOs and HCs were analyzed utilizing SPSS20.0 software (SPSS,Chicago, IL, USA) with independent samplettest,P＜0.05 was considered to be statistically significant.

To categorize the mean DC values in diverse cerebral regions of RVOs from HCs, the receiver operating characteristic(ROC) curve method was utilized. The correlationship between the DC signal of diverse cerebral regions and the clinical variables in RVOs were investigated with the Pearson’s correlation analysis.

RESULTS

DemographicsThere were 12 CRVO and 9 BRVO patients in RVO group, and 11 left eyes and 10 right eyes. The average duration was 115.24±45.65d.

There were no differences in age (P=0.481) and handiness(P＞0.99) between the RVO and HC group. The hospital anxiety and depression scale (HADS) score were significantly increased in RVOs (13.76±3.67) when compare with HCs(5.52±1.14,P＜0.011).

Figure 2 Comparison of DC values in the RVO and HC group A, B: There were significant differences of DC between two groups. The red parts show increased DC values, the blue regions show decreased DC values; C: The mean DC values in specific areas between RVOs and HCs.

Table 2 Brain areas showed statistical differences in DC between RVO patients and HCs

Degree Centrality DifferencesIn RVO group, the DC signals were significantly elevated in the brain areas including the right superior parietal lobule (RSPL), left precuneus (LP)and right middle frontal gyrus (RMFG), and declined in the left middle temporal gyrus (LMTG) and bilateral anterior cingulated (BAC; Figure 2, Table 2). However, there is no significant differences between CRVO and BRVO (Table 3).

Correlation AnalysisIn RVO patients, the average DC value in BAC were negatively correlated with the HADS score (r=-0.858,P＜0.001; Figure 3).

Receiver Operating Characteristic CurveWe contemplated that the variations of DC signal might be potential valuable diagnostic markers for clarifying the RVOs from HCs.The ROC curve analysis was applied to substantiate this hypothesis, the average DC values of specific cerebral regions were calculated. The respective area under the curve (AUC)of DC signal in different brain regions were as follow: RSPL(0.896,P＜0.001), LP (0.900,P＜0.001), RMFG (0.931,P＜0.001; Figure 4A, RVOs＞HCs); LMTG (0.962,P＜0.001),BAC (0.855,P＜0.001; Figure 4B, RVOs＜HCs).

Figure 3 Correlationship between DC values in specific areas and the HADS score in RVO group The mean DC value of BAC was negatively correlated with the HADS score.

Table 3 DC differences of specific brain areas between CRVOs and BRVOs

DISCUSSION

Rs-fMRI can exhibit spatial patterns of temporal interrelationship efficiently beyond the extent of the data’s point spread function. DC is an innovative and dependable rs-fMRI technique to explore the changes of cerebral functional connectivity, detecting and quantitating sites of activation in the brain. The DC technique has been profitably utilized in a few ophthalmological diseases[12-15](Table 4). To the best of our knowledge, this investigation is the first study to discover the cerebrum functional connectivity referred to RVO patients.In this current research, we proved that the intrinsic cerebral activity patterns of various areas in RVOs were changed when compared with HCs. The RVO group revealed significant incremental DC signal values in RSPL, LP and RMFG, but decreased in the with LMTG and BAC with impaired visual function (Figure 5).

Analysis of Higher Degree Centrality Signal in the Retinal Vein Occlusion PatientsThe parietal lobe is situated behind the frontal lobe and above the temporal lobe, where superior parietal lobule lies on top, and the precuneus is a segment of the superior parietal lobule on the medial surface of hemicerebrum. A pile of evidence about functional neuroimaging researches have consistently disclosed that the superior parietal lobule is associated with many sensory and cognitive process, including visuospatial attention[16-17],visuomotor integration[18], spatial perception[19]and memory[20].Previous study has detected the activity in the superior parietal lobule and precuneus during visual stimulation[21]. Khanet al[22]reported that RSPL damaged in left optic ataxia patient,indicating the relationship with visual-motor transformation in this region. Chenet al[23]has demonstrated that the gray matter volume exhibited a prominent escalation in superior parietal lobe and precuneus of patients with primary open-angle glaucoma. In addition, research of the patients with stroke, a kind of vascular occlusive disorder as well, displayed damage in the superior parietal lobule while visual extinction[24]. In agreement with those previous reports, the elevated DC values in right parietal lobule and LP found in our investigation reflected a significant visual-related activation in this region,which may have been affected by the impaired visual function.The middle frontal gyrus is an expansive gyrus that locates between the inferior and superior frontal sulci, and the frontal eye field (FEF) lies on the back of middle frontal gyrus including a sizable oculomotor area[25], which is competent of initiate eye movement as well as influence the precision or latency[26]. Several researches illuminated that the middle frontal gyrus took a vital part in saccade associated with movement generation[27-28]. Daiet al[29]found that the functional connectivity of glaucoma was increased specifically in the primary visual cortex and middle frontal gyrus. And many retina-involved disorders were detected higher activation in FEF, including progressive retinitis pigmentosa[30], macular hole[31]and age-related macular degeneration[32]. In supporting of the precedent findings, we also discovered that individuals with RVO displayed conspicuous increasing DC values in the RMFG, suggesting activate of the visual processing. This kind of alterations might indicate the cerebral plasticity which recompense for RVO causing visual input deficiency.

Analysis of Lower Degree Centrality Signal in the Retinal Vein Occlusion PatientsThe middle temporal gyrus lies on the lateral exterior of the temporal lobe, abdominal to the superior temporal gyrus, connected with the multi-sensory integration[33]and semantic comprehension processing[34-35].Previous investigation has displayed that the brain activity was obviously higher in the middle temporal gyrus during visual stimulation[36]. And similar activation of the middle temporal gyrus was also observed when fusion stimulus[37-38]. Studies on gaze perception have demonstrated the middle temporal gyrus showed sensitive to the gaze shift[39-40]. Moreover, Yuet al[41]and Liet al[42]detected that the volume of gray matter in middle temporal gyrus was found to be decreased in primary openangle glaucoma. In the present investigation, the DC reduction in the LMTG of RVO group suggested functional damaged in this specific area. Hence, we ulteriorly assumed that the decline may be due to the visual deterioration in RVO patients.

Figure 4 ROC curve analysis in specific brain areas A: The AUC of RSPL were 0.896 (95%CI: 0.793-0.999), LP 0.900 (95%CI: 0.786-1.000),RMFG 0.931 (95%CI: 0.844-1.000) (RVOs＞HCs); B: The AUC of LMTG were 0.962 (95%CI: 0.906-1.000), BAC 0.855 (95%CI: 0. 725-0.984) (RVOs＜HCs).

Figure 5 The altered brain regions in RVO group The DC values in RVOs of these areas were increased: 1-LP (t=4.2945), 2-RSPL(t=4.4701) and 3-RMFG (t=6.1133), and these areas decreased: 4-BAC(t=-4.4668) and 5-LMTG (t=-4.8749).

Figure 6 Relation between rs-fMRI imaging and clinical features and symptoms in RVO Retinal vein occlusion in RVO contributes to vision loss and anxiety, further leads to changes in anterior cingulate activity.

The anterior cingulate is the front portion of the cingulate cortex which is like a “collar” encircling the anterior corpus callosum, comprises of a wide swath of neural territory along the frontal midline of the cerebrum. The anterior cingulate is broadly identified to play a role in the interaction between cognitive control[43-44]and emotional response[45-46]. The negative relationship between anterior cingulate and the HADS we found in the current study may due to the anxiety emotion caused by the vision loss (Figure 6). A recent study has reported that damage to the anterior cingulate cortex is involved in deficits in visual function[47]. And lower ReHo was identified at the BAC cortex in patients with optic neuritis[48].The DC decline in the BAC exhibited in the present research may indicate functional impairment in RVO group, which offered additional proof that the RVO might contribute to the disfunction of the anterior cingulate.

Notwithstanding all this, the present investigation still has some limitations. First, the total number of enlisted patients and controls were relatively small, and the small sample size might influence the reliability. Moreover, different types of RVO were involved, which should be classified when analyzed. Nevertheless, no significant differences were discovered between CRVO and BRVO (Table 3), this could be a result of the sample size. Further investigation is required to study the cerebral functional alterations more precisely.

In conclusion, the current research illustrated that patients with RVO had aberrant intrinsic activities in distinct cerebral areas, which provided discernment into the neural alteration in RVO subjects, and aided to reveal the latent mechanisms of RVO.

Table 4 DC technique applied in ocular diseases

ACKNOWLEDGEMENTS

Conflicts of Interest: Dong WJ,None;Su T,None;Li CQ,None;Shu YQ,None;Shi WQ,None;Min YL,None;Yuan Q,None;Zhu PW,None;Liu KC,None;Yi JL,None;Shao Y,None.