The Cerebellum Posterior Lobe Associates with the Exophthalmos of Primary Hyperthyroidism: A Resting-State fMRI Study

Background Exophthalmos occurs in patients with primary hyperthyroidism. There were few studies about the changes of brain functional networks of patients with exophthalmos of primary hyperthyroidism (EOPH). However, differences in spontaneous brain activity in patients with EOPH remain unclear. Objective This study explored alterations in the brain functional networks of patients with EOPH using a voxel-wise degree centrality (DC) method. Methods A total of 20 patients with EOPH (8 men and 12 women) were enrolled. In addition, 20 patients with primary hyperthyroidism without exophthalmos, matched in age, sex, and education status, were enrolled as a control group. The Hospital Anxiety and Depression Scale was used to assess the anxiety and depression status of participants. All participants were examined using resting-state functional MRI. Changes in spontaneous brain activity were investigated using the DC method. To distinguish between the DC values of the patients with EOPH and those of the control group, we analyzed the receiver operating characteristic (ROC) curve. The interrelationships between the DC values and clinical variables in the patients with EOPH were evaluated using Pearson's correlation coefficient. Results Patients with EOPH exhibited notably lower DC values in the cerebellum posterior lobe than the control group. In addition, there were negative correlations between the anxiety scores (AS) and the depression scores (DS) and DC values of the cerebellum posterior lobe. The ROC curve analysis of the cerebellum posterior lobe demonstrated that the area under the curve method had a high diagnostic accuracy. Conclusions Our study was the first, to our knowledge, to explore changes in the brains of patients with EOPH using the DC method. The DC value was significantly different in the cerebellum posterior lobe in patients with EOPH, indicating that the cerebellum posterior lobe is associated with EOPH.


Introduction
Primary hyperthyroidism is a common clinical syndrome, which is caused by excessive synthesis and release of thyroid hormones. Moreover, patients usually experience high metabolic and sympathetic excitation [1,2]. Many patients with primary hyperthyroidism will develop typical eye symptoms, including exophthalmos, eyelid edema, and visual loss, of which exophthalmos is one of the most significant and specific signs [3,4]. Figure 1 shows exophthalmos and nonprominent eyes in patients with primary hyperthyroidism.
Exophthalmos of primary hyperthyroidism (EOPH) is characterized by extraocular muscle hypertrophy, an increase in muscle tension, and ocular movement disorder [5]. e ocular protrusion of EOPH is greater than 18 mm, and EOPH often progresses bilaterally [6]. Because patients with EOPH cannot close their eyelids, the conjunctiva and cornea are exposed to air, easily leading to conjunctivitis and keratitis, and even blindness. Previous studies have shown that EOPH might be associated with extraocular muscle hypertrophy, retrobulbar tissue increase, and the invasion of lymphocytes [7]. However, rarely research had focused on the intrinsic brain activity variations of EOPH. e function and contribution of brain tissue are not yet fully determined. Fortunately, brain image scanning might promote noninvasive exploration of alterations in the brain functional networks of patients with EOPH.
In recent years, the development of neuroimaging technology has provided more possibilities for studying the mechanism of neuropsychiatry-related diseases. At the same time, the rapid development of fMRI has made it possible to provide solid evidence that changes in brain regions are closely related to several ophthalmological diseases. Based on the different magnetic properties of oxyhemoglobin and deoxyhemoglobin, blood-oxygen-level dependent signals of different brain regions are obtained, while subjects remain relaxed and awake with closed eyes [8,9]. fMRI can be used to indirectly search for specific brain activation areas of subjects and is a useful tool in determining the functional connectivity of the human brain. Moreover, the fMRI data can also serve as the physiological basis for information processing and mental representation.
Voxel-wise degree centrality is one method of restingstate fMRI, which measures the topological structure of brain functional connectors at the voxel level and detects the functional relationships between a brain region and the rest of the brain [10]. e DC value represents the number of direct connections of a given voxel in the functional connector, and a high DC value represents a node with substantial direct connections to other nodes [11]. us, DC values can objectively and comprehensively reflect the abnormalities of brain functional areas. erefore, DC is a better network measurement than other methods. Moreover, DC has been utilized to assess the altered brain activity in patients with epilepsy, Parkinson's disease, type-two diabetes, attention deficit hyperactivity disorder, and Alzheimer's disease [12][13][14][15][16]. Accordingly, the DC method is a dependable rs-fMRI technology, which has not yet been used in the study of EOPH. In this study, the DC method is used to further understand converted spontaneous brain activities in patients with EOPH in comparison with patients with primary hyperthyroidism without exophthalmos.

Subjects.
A total of 20 patients with EOPH (8 men and 12 women) were recruited from the First Affiliated Hospital of Nanchang University, China. Inclusion criteria were as follows: (1) met the diagnostic criteria for primary hyperthyroidism; (2) exophthalmos confirmed by imaging examination; (3) bilateral exophthalmos; (4) men were 18-60 years old, while women were 18-50 years old; (5) righthandedness. Exclusion criteria were as follows: (1) a history of severe craniocerebral trauma; (2) a history of epileptic seizures, schizophrenia, bipolar disorder, or other mental disorders; (3) a history of alcohol or drug abuse; (4) other endocrine diseases and autoimmune diseases besides hyperthyroidism; (5) pregnancy; (6) claustrophobia or other contraindications for magnetic resonance imaging.
Twenty patients with primary hyperthyroidism without exophthalmos (8 men and 12 women), matched for age, sex, and education status, were also recruited as control groups for this research. Inclusion criteria were as follows: (1) primary hyperthyroidism, (2) no symptoms of exophthalmos, (3) no other endocrine diseases or autoimmune diseases, (4) no history of alcohol or drug abuse, (5) no history of neurological or psychiatric disorders, (6) not pregnant, (7) righthandedness, and (8) no MRI scanning contraindications. e study received approval from the Medical Ethics Committee of the First Affiliated Hospital of Nanchang University and abided by the Declaration of Helsinki. All participants were informed of the objectives, contents, potential risks, and signed informed consent forms.

Data
Processing for fMRI. MRIcro (http://www.MRIcro. com), Statistical Parametric Mapping software (SPM8, http://www.fil.ion.ucl.ac.uk/spm), the Data Processing Assistant for rs-fMRI advanced edition (DPARSFA, http:// rfmri.org/DPARSF) software, and the Resting-State Data Analysis Toolkit (REST, http://www.restfmri.net) were utilized to prefilter the data. e procedure of preprocessing included the following steps. Firstly, the first 10 volumes of each field are abandoned to achieve balance of the signal and adaptation of the participants. Secondly, the translation (in millimeters) and rotation (in degrees) of each subject is estimated after head movement correction, and volumes with movements in the x, y, or z direction greater than 2 mm en, fMRI images were detrended and bandpass filtered (0.01-0.08 Hz) to reduce effects of lowfrequency drift and physiological high-frequency respiratory and cardiac noise. en, the fMRI images are standardized to the Montreal Neurological Institute (MNI) using standard echo planar imaging templates and resampling at a resolution of 3 mm × 3 mm × 3 mm.

Degree Centrality.
On the basis of the individual voxelwise functional network, we calculated the DC value by counting the number of degrees of the binarized adjacency matrix or significant suprathreshold correlations between the subjects and used the following equation to convert the voxel DC diagram of each individual to a z-score map: where z i refers to the z-score of the ith voxel, DC i is the DC value of the ith voxel, and std all refers to the standard deviation [17].

Correlation
Analysis. Graphpad prism 7 (GraphPad Software Inc., San Diego, CA, USA) was then used to analyze the correlation between DC mean values in altered brain areas and the clinical behaviors (p < 0.05 was taken to indicate significant difference).

Statistical
Analysis. e demographic and clinical variables between the EOPH and control groups were compared using SPSS20.0 software (SPSS, Chicago, IL, USA) with the independent samples t-test. And p < 0.05 was regarded as statistically significant. e statistical threshold of the voxel level for multiple comparisons according to the Gaussian random field (GRF) theory was set at a level of p < 0.05. e general linear model analysis generated by the SPM8 toolkit was used to investigate the difference in DC values between patients with EOPH and the control group. e receiver operating characteristic (ROC) curve method was used to distinguish the mean DC values in brain areas of patients with EOPH from those of control groups. e correlations between the DC values of cerebrum areas and the clinical features in patients with EOPH were investigated using Pearson's correlation coefficient.

Demographics and Behavioral
Results. Twenty patients with EOPH (8 men and 12 women) and 20 control group patients (8 men and 12 women) were involved in this study.
ere was no statistically significant difference in patients' age, sex, or weight or in intraocular pressure between the two groups. However, significant differences existed in the duration of hyperthyroidism (p � 0.046), the best-corrected right-eye visual acuity (p � 0.009), the best-corrected lefteye visual acuity (p � 0.012), and the anxiety (p < 0.001) and depression scores (p < 0.001) in the two groups. Details are shown in Table 1.

Difference in DC.
e DC value clearly decreased in the cerebellum posterior lobe in the EOPH group (Figures 2(a) and 2(b) and Table 2). e mean DC between the two groups is shown in Figure 2(c).

Discussion
Resting-state fMRI can effectively display spatial patterns of temporal correlation. e DC method is a new and reliable rs-fMRI method for revealing changes in brain functional connectivity and detecting and quantifying activation sites in the brain. is method has already been applied effectively in the study of various ocular diseases, including glaucoma, comitant exotropia strabismus, late monocular blindness, and open globe injury ( Figure 5) [18][19][20][21]. As far as we know, this is the first time that the DC method has been used to identify changes in the endogenous functional connectivity on the basis of voxel-based global brain correlation analysis in individuals with EOPH.
In this study, we demonstrated altered intrinsic brain connectivity patterns of the cerebellum posterior lobe in individuals with EOPH. e EOPH group exhibited significantly decreased DC signal values in the cerebellum posterior lobe. In addition, anxiety and depression scores were negatively correlated with DC values of the cerebellum posterior lobe (Figure 6). e cerebellum posterior lobe lies between the cerebellar fissure and the posterolateral fissure and occupies the greater part of the cerebellum in human beings. e function of the cerebellum posterior lobe is to influence the initiation, planning, and coordination of movement and determine the strength, direction, and scope of the movement [22,23]. In addition, studies have consistently revealed that the cerebellum posterior lobe is also responsible for nonmotor functions [24]. e role of the cerebellum posterior lobe in eye movement control and adaptation is well known [25]. When performing microstimulations on the cerebellums of macaque monkeys, Fujikado and Noda [26] found that the cerebellum posterior lobe was responsible for saccadic eye movements. In another previous study, Lisberger [27] used the concept of internal models, which provided a conceptual structure, to help in understanding how the cerebellum controls eye movement. In addition, Colnaqhi and her International Journal of Endocrinology  associates [28] applied transcranial magnetic stimulation to test cerebellar control of reflexive and voluntary eye movements, proving that the cerebellum posterior lobe plays a key role in reflexive and voluntary eye movements. Furthermore, in research on concomitant strabismus patients, the mean diffusivity values in bilateral posterior cerebellar lobes were found, using digital transducer interface technology, to be significantly decreased [29]. In addition, Tan et al. [20] found that the DC value of the right posterior cerebellar lobe was decreased in patients with concomitant strabismus, indicating that dysfunction of the posterior cerebellar lobe was related to abnormal coordination of the extraocular muscles. All these facts prove that the cerebellum posterior lobe is closely associated with eye movement regulation. Consistent with these findings, we demonstrated decreased DC values in the cerebellum posterior lobe in this study; this finding might reflect the reduction of the role and position of the cerebellum posterior lobe in the whole-brain network in patients with EOPH. Furthermore, we speculatively suggest that the changed functional connectivity in the cerebellum posterior lobe might be associated with eye movement abnormalities in patients with EOPH.
However, the cerebellum posterior lobe also plays a key role in emotion [30,31]. Lou et al. [32] found that the DC value in the cerebellum posterior lobe was decreased in depressed patients. erefore, it was speculated that dysfunction of the cerebellar posterior lobe might be related to depression. In another study, patients with obstructive sleep apnea exhibited complex and abnormal DC values in the cerebellum and other brain areas at rest [33]. It has been reported that obstructive sleep apnea patients also    experience depression and anxiety [34].
us, the brain activity change in the cerebellum posterior lobe might also be a basis of anxiety and depression. Furthermore, Göbel et al. [35] reported that experimental thyrotoxicity could lead to an increase in gray matter volumes in the left and right cerebellum posterior lobe. Moreover, the increase in gray matter volumes in the cerebellum posterior lobe was thought to be compensatory for depression, working memory, and movement coordination. In this study, we used the Hospital Anxiety and Depression Scale to assess the anxiety and depression status of participants. Our results showed that patients with EOPH had higher anxiety and depression scores than the control group. In addition, anxiety, and depression scores were negatively correlated with DC values of the cerebellum posterior lobe. It is likely that the alteration of brain network centrality is also associated with impairment of the emotion procession in patients with EOPH. e AUC is often utilized to reflect diagnostic accuracy. e ROC result revealed that the AUC of the cerebellum posterior lobe was 1.000, which suggested that there were significant differences in DC values between the EOPH and control groups. Consequently, we inferred that the DC value of the cerebellum posterior lobe might be a potential diagnostic marker for patients with EOPH.
Our research still has some limitations. First of all, our study population is not very large. More participants are needed to further verify the results. Furthermore, the molecular mechanism between the variation of cerebellum posterior lobe activity and exophthalmos in patients with hyperthyroidism needs further investigation.

Conclusion
To sum up, this study demonstrated that patients with EOPH had abnormal networks in their cerebellum posterior lobe, which indicates that the cerebellum posterior lobe is associated with exophthalmos and provides insight into the neural variation in patients with EOPH.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this article.