Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2018 Oct 13;40(3):976–986. doi: 10.1002/hbm.24426

Altered neurovascular coupling in neuromyelitis optica

Xi Guo 1, Jiajia Zhu 2, Ningnannan Zhang 1, Linjie Zhang 3, Yuan Qi 3, Huanhuan Cai 4, Xue Zhang 1, Jie Sun 1, Qiuhui Wang 1, Li Yang 3, Fu‐Dong Shi 3, Chunshui Yu 1,5,✉
PMCID: PMC6865682  PMID: 30315685

Abstract

Neurovascular coupling reflects the close relationship between neuronal activity and cerebral blood flow (CBF), providing a new mechanistic insight into health and disease. Neuromyelitis optica (NMO) is an autoimmune inflammatory demyelinating disease of the central nervous system and shows cognitive decline‐related brain gray matter abnormalities besides the damage of optic nerve and spinal cord. We aimed to investigate neurovascular coupling alteration and its clinical significance in NMO by using regional homogeneity (ReHo) to measure neuronal activity and CBF to measure vascular response. ReHo was calculated from functional MRI and CBF was computed from arterial spin labeling (ASL) in 56 patients with NMO and 63 healthy controls. Global neurovascular coupling was assessed by across‐voxel CBF‐ReHo correlations and regional neurovascular coupling was evaluated by CBF/ReHo ratio. Correlations between CBF/ReHo ratio and clinical variables were explored in patients with NMO. Global CBF‐ReHo coupling was decreased in patients with NMO relative to healthy controls (p = .009). Patients with NMO showed decreased CBF/ReHo ratio (10.9%–17.3% reduction) in the parietal and occipital regions and increased CBF/ReHo ratio (8.0%–13.3% increase) in the insular, sensorimotor, temporal and prefrontal regions. Some of these abnormalities cannot be identified by a single CBF or ReHo analysis. Both abnormally decreased and increased CBF/ReHo ratios were correlated with more severe clinical impairments and cognitive decline in patients with NMO. These findings suggested that patients with NMO show abnormal neurovascular coupling, which is associated with disease severity and cognitive impairments.

Keywords: cerebral blood flow, demyelinating autoimmune disease, functional magnetic resonance imaging, neuromyelitis optica, neurovascular coupling, resting state

1. INTRODUCTION

Although occupying only 2% of body weight, the human brain consumes nearly 20% energy of the whole body, most of which is used to support resting‐state neuronal activity (Phillips, Chan, Zheng, Krassioukov, & Ainslie, 2016). This is facilitated by a tight coupling between neuronal activity and blood supply. Elevated neuronal activity tends to have greater metabolic demand, resulting in increased perfusion (Venkat, Chopp, & Chen, 2016). The structural basis for neurovascular coupling is the neurovascular unit (NVU) which is mainly composed of neurons, astrocytes and vessels (Muoio, Persson, & Sendeski, 2014). The impairment of any component of the NVU can affect its function and lead to abnormal neurovascular coupling (Zlokovic, 2010). Astrocytes can relay information exchange between neurons and vessels, and thus act as a hub in the NVU for neurovascular coupling (Filosa, Morrison, Iddings, Du, & Kim, 2016; Stobart & Anderson, 2013).

Resting‐state functional MRI (fMRI) can reveal spontaneous neuronal activity by analyzing blood‐oxygen‐level‐dependent (BOLD) signals (Fox & Raichle, 2007). Simultaneously, blood supply can be assessed by resting‐state cerebral blood flow (CBF). As expected, several studies have revealed a close correlation between brain activity and CBF, which has been used to investigate neurovascular coupling in the normal human brain, as well as abnormal neurovascular coupling in pathological conditions of schizophrenia and Alzheimer's disease (Merlini, Davalos, & Akassoglou, 2012; Phillips et al., 2016; Tarantini, Tran, Gordon, Ungvari, & Csiszar, 2016; Zhu et al., 2017).

Neuromyelitis optica (NMO) is an autoimmune inflammatory demyelinating disease of the central nervous system (Wingerchuk, Lennon, Lucchinetti, Pittock, & Weinshenker, 2007). Besides the involvement of optic nerve and spinal cord, gray matter (GM) abnormalities have been frequently reported in NMO. NMO patients have shown astrocyte damage and neuron loss in cerebral cortex in pathologic studies (Kawachi & Lassmann, 2017); altered neuronal activities in resting‐state fMRI studies (Cai et al., 2017; Liang et al., 2011; Liu et al., 2011; Liu et al., 2017; Lopes et al., 2015); and reduced gray matter volume (GMV) in structural MRI (Kim et al., 2017; Liu et al., 2015; Masuda et al., 2017; Wang, Zhang et al., 2015). Several of these alterations have been correlated with cognitive impairment in patients with NMO. Based on the theory of neurovascular coupling, GM damage in NMO can be understood from a new perspective. Astrocyte damage is thought to be a key pathological characteristic of NMO (Kawachi & Lassmann, 2017; Takeshita et al., 2017), which may disrupt the NVU and eventually result in abnormal neurovascular coupling. In addition, the previously reported neuronal (Duan et al., 2012; Liu et al., 2015; Pichiecchio et al., 2012; Saji et al., 2013; von Glehn et al., 2014; Wang, Zhang et al., 2015) and vascular impairments (Sanchez‐Catasus et al., 2013; Takeshita et al., 2017) could also affect neurovascular coupling in NMO. Based on the evidence of astrocyte damage, neuronal loss, and vascular impairments in NMO, we hypothesized that patients with NMO would show alterations in neurovascular coupling.

In the present study, we used regional homogeneity (ReHo) to represent neuronal activity, CBF to measure vascular response, across‐voxel CBF‐ReHo correlation to reflect global neurovascular coupling, and CBF/ReHo ratio to estimate regional neurovascular coupling (Liang, Zou, He, & Yang, 2013). The ReHo quantifies the similarity of the BOLD signal of a given voxel to those of its nearest neighbors in a voxel‐wise manner (Jiang & Zuo, 2016; Zang, Jiang, Lu, He, & Tian, 2004) and both decreased and increased ReHo have been found in NMO (Liang et al., 2011). The CBF changes and their associations with the number of optic neuritis attacks were also found in NMO patients (Sanchez‐Catasus et al., 2013). However, neither global (reflecting the consistency of spatial distribution between cerebral blood flow and neuronal activity) nor regional (measuring the amount of blood supply per unit of neuronal activity) neurovascular coupling changes have been investigated in NMO. Here, we combined the ReHo and CBF for the first time to investigate neurovascular coupling changes and its clinical significance in NMO, expecting to provide further information on the neuropathological mechanisms of NMO from a new perspective.

2. MATERIALS AND METHODS

2.1. Participants

The experimental protocol was approved by the local medical research ethics committee and written informed consent was obtained from all participants. A total of 56 patients with NMO and 63 sex‐matched, age‐matched and education‐matched healthy controls were recruited in our study. Inclusion criteria were age (18–70 years) and right‐handedness. All patients fulfilled the revised Wingerchuk diagnostic criteria for NMO (Wingerchuk, Lennon, Pittock, Lucchinetti, & Weinshenker, 2006), including two absolute criteria of optic neuritis and myelitis and at least two of the following three supportive criteria: brain MR imaging negative or non‐diagnostic for multiple sclerosis at onset, MR imaging evidence of a spinal cord lesion involving more than three vertebral segments, and a positive serological test for NMO antibodies. Serological test for the water channel aquaporin‐4 (AQP4) antibody was conducted by the Neuroimmunology laboratory of Tianjin Neurological Institute, using a cell‐based array by the quantitative flow cytometry method. The exclusion criteria were MR imaging contraindications, history of head trauma or other neuropsychiatric diseases, other autoimmune diseases, and poor image quality. The disease severity was assessed by the Expanded Disability Status Scale (EDSS) scores. The detailed demographic data for these participants are shown in Table 1.

Table 1.

Demographic characteristics of participants

Characteristics NMO (n = 56) Healthy controls (n = 63) Statistics p value
Sex, F/M 48/8 53/10 χ2 = 0.058 .809
Age, y 48 (18–68) 49 (25–67) T = 0.676 .501
Education, y 11 (0–16) 12 (6–19) T = 1.568 .120
EDSS score 3.94 (0–8.5) NA NA NA
Open in a new tab

All values are expressed as mean with ranges (minimum to maximum values). Abbreviations: EDSS = expanded disability status scale; NA = not applicable; NMO = neuromyelitis optica.

2.2. Data acquisition

MR imaging data were acquired using a 3.0‐Tesla MR scanner (Discovery MR750, General Electric, Milwaukee, WI). Tight but comfortable foam padding was used to minimize head motion, and earplugs were used to reduce scanner noise. All subjects were instructed to keep their eyes closed, relax, move as little as possible, think of nothing in particular, and stay awake during the scans. Resting‐state BOLD images were acquired using a gradient‐echo single‐short echo planar imaging sequence: TR/TE = 2000/45 ms; FOV = 220 mm × 220 mm; matrix = 64 × 64; FA = 90°; slice thickness = 4 mm; gap = 0.5 mm; 32 interleaved transverse slices; and 180 volumes. The resting‐state perfusion imaging was performed using a pseudo‐continuous ASL (pcASL) sequence with a 3D fast spin‐echo acquisition and background suppression (TR/TE = 4886/10.5 ms; post‐label delay = 2025 ms; spiral in readout of eight arms with 512 sample points; FA = 111°; FOV = 240 mm × 240 mm; reconstruction matrix = 128 × 128; slice thickness = 4 mm, no gap; 40 axial slices; number of excitation = 3; and acquisition time = 284 s). The label and control whole‐brain image volumes required eight TRs, respectively. A total of three pairs of label and control volumes were acquired. Sagittal 3D T1‐weighted images were acquired by a brain volume sequence with the following parameters: repetition time (TR) = 8.2 ms; echo time (TE) = 3.2 ms; inversion time (TI) = 450 ms; flip angle (FA) = 12°; field of view (FOV) = 256 mm × 256 mm; matrix = 256 × 256; slice thickness = 1 mm, no gap; and 188 sagittal slices. For each subject, all images were visually inspected during the acquisition of MRI data to ensure that no visible artifacts were found. If artifact appeared in any MR image, the relevant sequence will be re‐scanned in time to obtain qualified MR images of each subject.

2.3. Cerebral blood flow calculation

An ASL difference image was calculated after subtracting the label image from the control image. The three ASL difference images were averaged to calculate the CBF maps in combination with the proton‐density‐weighted reference images (Xu et al., 2010). SPM8 software was used to normalize the CBF images to the Montreal Neurological Institute (MNI) space. Specifically, individual ASL difference images were nonlinearly coregistered to the PET‐perfusion template in MNI space and then they were averaged to generate a study‐specific template of ASL difference images. Individual ASL difference images were non‐linearly coregistered to the study‐specific template. The deformation maps were used to warp individual CBF images to the MNI space. Then, each co‐registered CBF map was removed of non‐brain tissue and divided by the subject's global mean CBF value of the GM. Finally, standardized maps were spatially smoothed with a Gaussian kernel of 6 mm × 6 mm × 6 mm full‐width at half maximum (FWHM).

2.4. fMRI data preprocessing

The SPM8 (http://www.fil.ion.ucl.ac.uk/spm) was used to preprocessing the fMRI data. The first 10 volumes of each subject were discarded to allow the signal to reach equilibrium and the participant to adapt to the scanning noise. The remaining volumes were corrected for the acquisition time delay between slices. Then, realignment was performed to correct for head motion between time points. All subjects' BOLD data were within the defined motion thresholds (i.e., translational or rotational motion parameters less than 2 mm or 2°). We also calculated the frame‐wise displacement (FD), which indexes the volume‐to‐volume changes in head position. Several nuisance covariates (six motion parameters, their first time derivations, and average BOLD signals of the ventricular, white matter and whole brain) were regressed out from the data. A recent study has reported that the signal spike caused by head motion significantly contaminated the final resting‐state fMRI results even after regressing out the linear motion parameters (Power, Barnes, Snyder, Schlaggar, & Petersen, 2012). Therefore, we further regressed out spike volumes when the FD of the specific volume exceeded 0.5. The functional images were then band‐pass filtered with a frequency range of 0.01–0.08 Hz. In the normalization step, individual structural images were linearly co‐registered with the mean functional image; the structural images were then nonlinearly transformed to MNI space. Finally, the transformation parameters were applied to the functional images. The functional images were then resampled into a 3 × 3× 3 mm3 voxel.

2.5. Regional homogeneity calculation

Kendall's coefficient concordance (KCC) was used to measure regional homogeneity of a given voxel with its nearest 26 neighbor voxels in a voxel‐wise manner within the gray matter mask (34,911 voxels) (Zang et al., 2004). An individual ReHo map was obtained for each subject and was then divided by the global mean KCC value. Standardized maps were then spatially smoothed with a 6 mm × 6 mm × 6 mm FWHM Gaussian kernel.

2.6. Global CBF‐ReHo coupling analysis

To quantitatively evaluate the global coupling between CBF and ReHo, correlation analyses across voxels were performed for each participant (Liang et al., 2013). Then, a two sample t‐test was used to compare the differences in CBF‐ReHo correlation coefficients between the two groups, while controlling for the effects of age, sex, and education.

2.7. CBF/ReHo ratio analysis

To quantify the regional neurovascular coupling, we computed the CBF/ReHo ratio of each voxel. The CBF/ReHo ratio of each voxel for each participant was divided by the global mean value to improve normality. Voxel‐wise comparisons were then performed to identify brain regions with significant group differences in CBF/ReHo ratio using a two sample t‐test with age, sex, and education as the nuisance variables. Multiple comparisons were corrected by a voxel‐wise false discovery rate (FDR) method (Benjamini & Hochberg, 1995; Genovese, A Lazar, & Nichols, 2002) with a corrected threshold of q < 0.05.

2.8. Voxel‐wise comparisons in CBF and ReHo

Voxel‐wise comparisons were performed to identify the CBF or ReHo differences between the two groups controlling for age, sex, and education. Multiple comparisons were also corrected using a voxel‐wise FDR method (q < 0.05).

2.9. Clinical and neuropsychological assessments

A series of neuropsychological tests were performed by a professional psychologist (L.Z. with 5 years of experience) for all participants. Both the Mini‐Mental State Examination (MMSE) and https://www.ncbi.nlm.nih.gov/pubmed/25751471 (MoCA) were used as screening instruments for rough assessment of cognitive function (Saczynski et al., 2015). The California Verbal Learning Test–Second Edition (CVLT‐II) was used to test verbal learning and memory function, including immediate (immediate recall of trails 1–5, IR1‐5), short‐delayed (short‐delay cued recall, SDCR; short‐delay free recall, SDFR), and long‐delayed (long‐delay cued recall, LDCR; long‐delay free recall, LDFR) verbal memory (Blanco‐Rojas et al., 2013). The Brief Visuospatial Memory Test‐Revised (BVMT‐R) was used to measure visuospatial learning and memory (Tam & Schmitter‐Edgecombe, 2013) The Symbol Digit Technique Test (SDMT) was administered to measure information processing speed and working memory (Blanc et al., 2008).

For each significant cluster derived from voxel‐wise two sample t‐test between the patient and control groups, the mean CBF/ReHo ratio of voxels in this cluster was extracted. Spearman correlation coefficients were used to evaluate correlations between CBF/ReHo ratios of these clusters and clinical scores in NMO patients. Multiple comparisons were also corrected using the FDR method with a corrected threshold of q < 0.05. Several subjects were excluded from correlation analyses due to the failure of completing some items of these tests. The number of participants ultimately included in the cognitive‐related analyses and the scores of these tests are showed in Table 3.

Table 3.

Neuropsychological tests for patients with NMO and healthy controls

Neuropsychological tests NMO Healthy controls p value
MoCA 22.55 ± 4.61 (n = 42) 25.21 ± 2.76 (n = 43) .002*
MMSE 26.76 ± 3.29 (n = 46) 28.56 ± 1.58 (n = 43) .001*
BVMT‐R
Total learning 21.40 ± 7.43 (n = 44) 27.61 ± 13.69 (n = 43) .010*
Delayed recall 8.14 ± 3.00 (n = 44) 9.07 ± 2.99 (n = 43) .150
SDMT 41.87 ± 12.80 (n = 38) 49.40 ± 14.78 (n = 43) .017*
CVLT‐II
IR 1–5 46.65 ± 12.08 (n = 48) 50.47 ± 8.20 (n = 43) .079
SDFR 10.54 ± 2.97 (n = 48) 11.61 ± 2.34 (n = 43) .064
SDCR 10.35 ± 2.83 (n = 48) 11.74 ± 2.27 (n = 43) .012*
LDFR 10.52 ± 3.16 (n = 48) 12.19 ± 2.16 (n = 43) .004*
LDCR 10.74 ± 2.99 (n = 48) 11.88 ± 2.17 (n = 43) .041*
Open in a new tab

All values are expressed as mean ± SD.

*

Significant for p < .05.

Abbreviations: BVMT‐R = The Brief Visuospatial Memory Test – Revised; CVLT‐II = California Verbal Learning Test–Second Edition; IR1‐5 = immediate recall of trail 1–5; LDCR = long‐delay cued recall; LDFR = Long‐delay free recall; MMSE = mini‐mental state examination; MoCA = Montreal cognitive assessment; SDCR = short‐delay cued recall; SDFR = short‐delay free recall; SDMT = Symbol Digit Modalities Test.

2.10. Reproducibility validation

Both primary and secondary neuronal impairments in GM have been reported in NMO, which manifesting as changes in GMV (Duan et al., 2012; Liu et al., 2015; Pichiecchio et al., 2012; Saji et al., 2013; von Glehn et al., 2014; Wang, Zhang et al., 2015). To at least partially reduce the effects of GMV changes, we repeated the global CBF‐ReHo coupling comparison with the mean GMV of each subject as an additional covariate of no interest, and also re‐compared the voxel‐wise CBF/ReHo ratio with the GMV of each voxel as an additional covariate of no interest.

We also used amplitude flow frequency fluctuation (ALFF) from fMRI data to reflect neuronal activity. The ALFF was calculated using REST software (http://www.restfmri.net/). The preprocessing steps included slice timing, realignment, regression, normalization, and smoothing using the same parameters as the ReHo preprocessing. Then the preprocessed time series were transformed to a frequency domain using fast Fourier transform (FFT), and the power spectrum was then obtained. The square root of the power spectrum was calculated at each frequency and averaged across 0.01–0.08 Hz for each voxel. This averaged square root was taken as the ALFF (Zang et al., 2007). For standardization purposes, the ALFF of each voxel was divided by the global mean ALFF within the GM mask. Then, the global CBF‐ALFF coupling and the voxel‐wise CBF/ALFF ratio were calculated and compared between the two groups controlling for age, sex, and education. For the voxel‐wise CBF/ALFF ratio analysis, we used an uncorrected threshold of p < .005.

2.11. Statistical analysis

All demographic and clinical variables but the sex were examined with two sample t‐test using the Statistical Package for the Social Sciences version 22.0 (SPSS, Chicago, IL). Sex data were analyzed with a χ2 test.

3. RESULTS

3.1. Spatial distribution of the CBF/ReHo ratio

The CBF, ReHo, and CBF/ReHo ratio maps of patients with NMO and healthy controls are shown in Figure 1. Despite of subtle differences, both groups showed higher CBF, ReHo, and CBF/ReHo ratio in the precuneus, medial and lateral prefrontal cortex, inferior parietal lobule (IPL) and anterior cingulate cortex (ACC).

Figure 1.

Figure 1

Open in a new tab

Spatial distribution maps of CBF, ReHo, and CBF/ReHo ratio in healthy controls and patients with NMO. Despite of subtle differences, both groups show consistent spatial distribution (brain regions and magnitude) in these measures. Abbreviations: CBF = cerebral blood flow; HC = healthy controls; L = left; NMO = neuromyelitis optica; R = right; ReHo = regional homogeneity [Color figure can be viewed at https://wileyonlinelibrary.com]

3.2. Global CBF‐ReHo coupling changes in patients with NMO

Significant across‐voxel correlations between CBF and ReHo were found in each participant (including all patients with NMO and all healthy controls). Two representative correlation maps from one NMO patient and one healthy control are shown in Figure 2a. Nevertheless, patients with NMO showed a significant reduced global CBF‐ReHo coupling (T = 2.673, p = .009) compared to healthy controls at the group level (Figure 2b).

Figure 2.

Figure 2

Open in a new tab

Reduced global CBF‐ReHo coupling in patients with NMO. (a) Examples of the spatial correlation across voxels between CBF and ReHo in a control subject (blue; r = 0.9752) and a patient with NMO (red; r = 0.9085). (b) At the group level, the mean global CBF‐ReHo coupling is significantly lower (p = .009) in patients with NMO than in healthy controls. Error bars represent the standard error. Abbreviations: CBF = cerebral blood flow; HC = healthy controls; NMO = neuromyelitis optica; ReHo = regional homogeneity [Color figure can be viewed at https://wileyonlinelibrary.com]

3.3. CBF/ReHo ratio changes in patients with NMO

Compared with healthy controls, patients with NMO exhibited decreased CBF/ReHo ratio in the IPL bilaterally (10.9% and 14.2%), the right superior parietal lobule (SPL) (14.5%) and lingual gyrus (17.3%), as well as increased CBF/ReHo ratio in the bilateral insular cortices (10.3% and 8.6%) and postcentral gyri (10.2% and 9.2%), the right superior temporal gyrus(STG)(10.3%) and ACC (8.2%), and the left medial (MFG)(13.3%) and inferior frontal gyri(IFG)(8.0%) (q < 0.05, FDR corrected) (Figure 3 and Table 2).

Figure 3.

Figure 3

Open in a new tab

Group differences in CBF/ReHo ratio between patients with NMO and healthy controls (p < .05, FDR corrected). Compared to healthy controls, patients with NMO exhibited decreased CBF/ReHo ratio in the posterior brain regions, but increased CBF/ReHo ratio in the anterior brain regions. Abbreviations: CBF = cerebral blood flow; HC = healthy controls; L = left; ReHo = regional homogeneity; NMO = neuromyelitis optica; R = right [Color figure can be viewed at https://wileyonlinelibrary.com]

Table 2.

Brain regions with significant group differences in CBF/ReHo ratio

Regions Brodmann areas Cluster size (voxels) Peak t values Coordinates in MNI (x, y,z)
NMO < healthy controls
Right lingual gyrus 18 22 −3.9 18,–96,–15
Left inferior parietal lobule 40 236 −4.9 −27,–60,42
Right superior parietal lobule 7 221 −5.1 21,–72,48
Right inferior parietal lobule 40 19 −3.6 36,–48,45
NMO > healthy controls
Right insula 13 238 4.5 39,0,0
Right superior temporal gyrus 21, 22 52 4.4 57,0,−6
Left medial frontal gyrus 10 97 5.4 –6,60,3
Left insula 13 38 3.7 −36,3,–3
Right postcentral gyrus 3 270 5.0 63,–9,18
Right anterior cingulate cortex 24 35 3.7 12,51,12
Left inferior frontal gyrus 45 46 4.1 −45,18,18
Left postcentral gyrus 3 102 4.9 −51,–12,24
Open in a new tab

Abbreviations: CBF = cerebral blood flow; ReHo = regional homogeneity; MNI = montreal neurological institute.

3.4. CBF and ReHo changes in patients with NMO

Compared with healthy controls, patients with NMO showed decreased CBF in the calcarine and SPL bilaterally, the right IPL and olfactory lobe as well as increased CBF in the putamen and thalamus bilaterally, the right STG, supplementary motor area and middle cingulate cortex (MCC), and the left MFG and postcentral gyrus (q < 0.05, FDR corrected) (Figure 4). However, no significant difference was found in ReHo between the two groups using the same correction method.

Figure 4.

Figure 4

Open in a new tab

Group differences in CBF between patients with NMO and healthy controls (p < .05, FDR corrected). Compared to healthy controls, patients with NMO exhibited decreased CBF in the posterior regions, but increased CBF in the anterior regions. Abbreviations: CBF = cerebral blood flow; L = left; R = right [Color figure can be viewed at https://wileyonlinelibrary.com]

The relationships between CBF/ReHo ratio and CBF changes in patients with NMO are shown in Figure 5. Although some brain regions (violet and cyan) showed consistent changes between these two measures, several brain regions (red and green) only showed significant changes in CBF/ReHo ratio (q < 0.05, FDR corrected).

Figure 5.

Figure 5

Open in a new tab

The overlaps of CBF/ReHo ratio and CBF changes in NMO. The CBF changes mimic but not completely consistent with the CBF/ReHo ratio changes in NMO. Abbreviations: CBF = cerebral blood flow; L = left; NMO = neuromyelitis optica; ReHo = regional homogeneity; R = right [Color figure can be viewed at https://wileyonlinelibrary.com]

3.5. Clinical correlation analysis

Using an uncorrected threshold of p < .05, patients with NMO performed significantly worse on the Mini‐Mental State Examination (MMSE) (p = .001), https://www.ncbi.nlm.nih.gov/pubmed/25751471 (MoCA) (p = .002), Brief Visuospatial Memory Test‐Revised (BVMT‐R) (p = .010), Symbol Digit Technique Test (SDMT) (p = .017), short‐delay cued recall (SDCR) (p = .012), long‐delay free recall(LDFR) (p = .004) and long‐delay cued recall (LDCR) (p = .041) of the California Verbal Learning Test–Second Edition (CVLT‐II).

Spearman correlations between the CBF/ReHo ratio of each significant cluster and the clinical and neuropsychological scores in NMO are shown in Table 4. After correcting for multiple comparisons (q < 0.05, FDR corrected), only the abnormally reduced CBF/ReHo ratio in the right SPL was significantly correlated with MoCA (q = 0.036, FDR corrected). In the following part, we also show correlations using an uncorrected threshold of p < .05. We found significant positive correlations between the EDSS scores and the CBF/ReHo ratios of left SPL (rho = 0.265, p = .048) and postcentral gyrus (rho = 0.280, p = .037). For the neuropsychological tests, both abnormally decreased and increased CBF/ReHo ratios were related to cognitive decline in patients with NMO. Specifically, the abnormally reduced CBF/ReHo ratio in the right SPL was positively associated with MMSE (rho = 0.397, p = .006), BVMT‐R total learning (rho = 0.322, p = .033), delayed recall (rho = 0.393, p = .008) and SDMT (rho = 0.358, p = .027); the decreased CBF/ReHo ratio in the right IPL was also positively associated with MoCA (rho = 0.429, p = .005), MMSE (rho = 0.307, p = .038) and SDMT (rho = 0.385, p = .017). In contrast, the abnormally increased CBF/ReHo ratio in the right insula was negatively correlated with MoCA (rho = −0.314, p = .043); the abnormally increased CBF/ReHo ratio in the right STG was negatively correlated with MoCA (rho = −0.314, p = .043), MMSE (rho = −0.379, p = .009) and BVMT‐R total learning (rho = −0.303, p = .046); the abnormally increased CBF/ReHo ratio in the left insula was negatively associated with MoCA (rho = −0.385, p = .012), BVMT‐R delayed recall (rho = −0.337, p = .025), SDMT (rho = −0.370, p = .022), CVLT IR1‐5 (rho = −0.423, p = .003), SDFR (rho = −0.351, p = .014), SDCR (rho = −0.331, p = .021), and LDFR (rho = −0.350, p = .015); the abnormally increased CBF/ReHo ratio in the right postcentral gyrus was negatively associated with MoCA (rho = −0.309, p = .047), MMSE (rho = −0.314, p = .033), and BVMT‐R total learning (rho = −0.428, p = .004); the abnormally increased CBF/ReHo ratio in the left IFG was negatively associated with MoCA (rho = −0.374, p = .015), MMSE (rho = −0.356, p = .015), SDMT (rho = −0.391, p = .015), SDFR (rho = −0.348, p = .015) and SDCR (rho = −0.381, p = .008); and the abnormally increased CBF/ReHo ratio in the left postcentral gyrus was negatively associated with MoCA (rho = −0.309, p = .047), MMSE (rho = −0.323, p = .029), and BVMT‐R delayed recall (rho = −0.423, p = .004).

Table 4.

Correlations between CBF/ReHo ratio and clinical variables in NMO

Regions EDSS MoCA MMSE Total learning (BVMT‐R) Delayed recall (BVMT‐R) SDMT IR1‐5 (CVLT‐II) SDFR (CVLT‐II) SDCR (CVLT‐II) LDFR (CVLT‐II) LDCR (CVLT‐II)
NMO < healthy controls
R lingual gyrus −0.199 (0.140) −0.010 (0.950) 0.077 (0.609) −0.038 (0.808) 0.011 (0.945) 0.290 (0.078) 0.137 (0.351) 0.035 (0.812) 0.094 (0.526) −0.019 (0.898) 0.068 (0.650)
L inferior parietal lobule 0.162 (0.232) 0.117 (0.460) 0.144 (0.341) −0.093 (0.548) −0.146 (0.343) −0.103 (0.540) −0.060 (0.685) −0.041 (0.780) −0.025 (0.867) 0.065 (0.659) −0.023 (0.878)
R superior parietal lobule 0.053 (0.695) 0.533 (<0.001) * 0.397 (0.006) * 0.322 (0.033) * 0.393 (0.008) * 0.358 (0.027) * 0.229 (0.117) 0.245 (0.093) 0.252 (0.084) 0.134 (0.365) 0.091 (0.542)
R inferior parietal lobule 0.069 (0.613) 0.429 (0.005) * 0.307 (0.038) * 0.237 (0.121) 0.252 (0.099) 0.385 (0.017) * 0.240 (0.100) 0.176 (0.231) 0.172 (0.242) 0.079 (0.592) 0.007 (0.964)
NMO > healthy controls
R insula 0.181 (0.181) −0.314 (0.043) * −0.268 (0.072) 0.044 (0.776) −0.128 (0.406) −0.234 (0.158) −0.223 (0.127) −0.119 (0.420) −0.136 (0.356) −0.098 (0.507) 0.024 (0.872)
R superior temporal gyrus 0.051 (0.711) −0.425 (0.005) * −0.379 (0.009) * −0.303 (0.046) * −0.236 (0.123) −0.318 (0.052) −0.187 (0.202) −0.173 (0.241) −0.136 (0.356) −0.033 (0.822) 0.062 (0.678)
L medial frontal gyrus 0.037 (0.789) 0.004 (0.980) 0.124 (0.412) 0.026 (0.868) 0.061 (0.694) 0.232 (0.162) −0.063 (0.202) −0.028 (0.851) −0.068 (0.646) −0.003 (0.986) −0.139 (0.353)
L insula −0.112 (0.410) −0.385 (0.012) * −0.138 (0.360) −0.233 (0.128) −0.337 (0.025) * −0.370 (0.022) a −0.423 (0.003) * −0.351 (0.014) * −0.331 (0.021) −0.350 (0.015) * −0.178 (0.231)
R postcentral gyrus −0.014 (0.920) −0.309 (0.047) * −0.314 (0.033) * −0.428 (0.004) * −0.248 (0.105) −0.315 (0.054) −0.105 (0.479) −0.111 (0.453) −0.136 (0.356) −0.088 (0.554) 0.015 (0.920)
R anterior cingulate cortex −0.031 (0.818) 0.017 (0.917) −0.168 (0.264) 0.213 (0.165) 0.158 (0.307) −0.026 (0.878) −0.196 (0.183) −0.127 (0.391) −0.157 (0.287) −0.145 (0.326) −0.062 (0.681)
L inferior frontal gyrus 0.265 (0.048) * −0.374 (0.015) * −0.356 (0.015) * −0.094 (0.542) −0.189 (0.220) −0.391 (0.015) * −0.237 (0.104) −0.348 (0.015) * −0.381 (0.008) * −0.239(0.102) −0.273 (0.064)
L postcentral gyrus 0.280 (0.037) * −0.309 (0.046) * −0.323 (0.029) * −0.257 (0.092) −0.423 (0.004) * −0.154 (0.355) −0.111 (0.454) −0.128 (0.385) −0.134 (0.363) −0.004 (0.980) −0.080 (0.594)
Open in a new tab

Spearman correlation coefficients were used to evaluate correlations between clinical scores and CBF/ReHo ratios of each significant cluster derived from voxel‐wise two sample t‐test between the two groups.

All values are expressed as the Spearman's rho (p value).

*

Significant for p < .05 and all the significant results are shown in bold. Abbreviations: BVMT‐R = The Brief Visuospatial Memory Test – Revised; CVLT‐II = California Verbal Learning Test–Second Edition; EDSS = expanded disability status scale; IR1‐5 = immediate recall of trail 1–5; L = left; LDCR = long‐delay cued recall; LDFR = long‐delay free recall; MMSE = mini‐mental State Examination; MoCA = Montreal cognitive assessment; R = right; SDCR = short‐delay cued recall; SDFR = short‐delay free recall; SDMT = Symbol Digit Modalities Test.

3.6. Validation analyses

Because GMV changes may affect global and regional neurovascular coupling changes in NMO, we repeated intergroup comparisons while further controlling for the GMV. For global CBF‐ReHo coupling, there was still a significant difference (T = 2.574, p = .011) between patients and healthy controls after controlling for the global GMV of each subject. For CBF/ReHo ratio, the spatial distribution of brain regions with altered CBF/ReHo ratio after GMV correction was similar to that without GMV correction (Figure 6). These findings suggest that the altered neurovascular coupling in NMO is independent of GMV changes.

Figure 6.

Figure 6

Open in a new tab

Group differences in CBF/ReHo ratio between patients with NMO and healthy controls after correction for GMV (p < .05, FDR corrected). The spatial distribution of brain regions with altered CBF/ReHo ratio after GMV correction was similar to that without GMV correction. Abbreviations: CBF = cerebral blood flow; GMV = gray matter volume; HC = healthy controls; L = left; NMO = neuromyelitis optica; R = right; ReHo = regional homogeneity [Color figure can be viewed at https://wileyonlinelibrary.com]

To validate the reproducibility of our findings, we also used the ALFF to replace the ReHo to assess neuronal activity. Although nonsignificant difference (T = 1.313, p = .192) in global CBF‐ALFF coupling between the two groups, in the informative regional neurovascular coupling analysis, brain regions with altered CBF/ALFF ratio (p < .005, uncorrected) (Figure 7) were similar to those with altered CBF/ReHo ratio (q < 0.05, FDR corrected), indicating that the reported neurovascular coupling alterations in NMO are reproducible.

Figure 7.

Figure 7

Open in a new tab

Group differences in CBF/ALFF ratio between patients with NMO and healthy controls (p < .005, uncorrected). The spatial distribution of brain regions with altered CBF/ALFF ratio was similar to that with altered CBF/ReHo ratio. Abbreviations: ALFF = amplitude flow frequency fluctuation; CBF = cerebral blood flow; HC = healthy controls; L = left; NMO = neuromyelitis optica; R = right [Color figure can be viewed at https://wileyonlinelibrary.com]

To avoid the bias caused by disease severity, we excluded seven most disabled patients with an EDSS score greater than 7.0 and repeated our analyses (n = 49). In the global CBF‐ReHo coupling analysis, there was still a significant difference (T = 2.863, p = .005) between patients and healthy controls. In the CBF/ReHo ratio analysis, the spatial distribution of significant brain regions (p < .005, uncorrected) (Figure 8) was similar to the main results (q < 0.05, FDR corrected) (Figure 3). These findings suggest that the altered neurovascular coupling is a characteristic which independent of the extent of disease disability in NMO.

Figure 8.

Figure 8

Open in a new tab

Group differences in CBF/ReHo ratio between the NMO patients with EDSS scores<7.0 (n = 49) and healthy controls (p < .005, uncorrected). The spatial distribution of brain regions is independent of disease severity. Abbreviations: CBF = cerebral blood flow; HC = healthy controls; L = left; NMO = neuromyelitis optica; R = right; ReHo = regional homogeneity [Color figure can be viewed at https://wileyonlinelibrary.com]

4. DISCUSSION

To our knowledge, this is the first study to investigate neurovascular coupling changes in NMO by combining BOLD and ASL techniques. Patients with NMO showed reduced global CBF‐ReHo coupling and abnormal CBF/ReHo ratio. Both abnormally decreased and increased CBF/ReHo ratios were associated with clinical impairments and cognitive decline in NMO. These findings may improve our understanding of the neural mechanisms of NMO from the perspective of neurovascular coupling.

In consistent with previous studies (Liang et al., 2013; Zhu et al., 2017), a significant across‐voxel correlation between CBF and ReHo was found in healthy controls, indicating the importance of the normal neurovascular coupling in physiology of the human brain. Although across‐voxel correlation between CBF and ReHo was also found in NMO, it was lower than that in healthy controls at the group level, indicating reduced global neurovascular coupling in NMO.

Although the CBF‐ReHo correlation could only roughly assess the neurovascular coupling, any changes in neurovascular coupling should be interpreted in terms of its structural basis (neurovascular unit). The neurovascular unit is composed of neurons, astrocytes and vessels (Muoio et al., 2014). The single or combined impairment of the NVU components can affect its function and lead to abnormal neurovascular coupling (Zlokovic, 2010). Astrocytes act as key intermediaries between neurons and vessels in NVU (Howarth, 2014; Stobart & Anderson, 2013). In NMO, the AQP4, the target antigen of autoimmunity, is concentrated in the foot processes of astrocytes; and a disease‐specific circulating autoantibody (NMO‐IgG) directed against this antigen results in astrocyte damage (Misu et al., 2007; Popescu et al., 2010; Saji et al., 2013). The impaired astrocytes may lose their role in information exchange between neurons and vessels, leading to the compromised coordination between neuronal activity and blood supply (Venkat et al., 2016). Thus the astrocyte damage may be one cause for the reduced neurovascular coupling in NMO. Neurons are also important components of NVU and are thought to be the driving force behind neurovascular coupling due to their high energy demand. Volume reduction and neuronal loss in GM have also been observed in NMO, which is partly due to the axonal degeneration secondary to optic neuritis and myelitis or may be a direct consequence of neuronal damage in GM (Duan et al., 2012; Kawachi & Lassmann, 2017; Liu et al., 2015; Pichiecchio et al., 2012; Saji et al., 2013; von Glehn et al., 2014; Wang, Zhang et al., 2015). The observed neuron damage may also account for the reduced neurovascular coupling in NMO. Any changes in vascular components may also affect the integrity of NVU. Dysfunction of endothelial cells has been observed in NMO (Takeshita et al., 2017), which may lead to the regulatory dysfunction of perfusion (Sanchez‐Catasus et al., 2013) and eventually result in the reduced neurovascular coupling. However, the relationship between vascular damage and reduced neurovascular coupling in NMO needs to be further studied.

Compared to the CBF‐ReHo correlation that renders a comprehensive change of neurovascular coupling of the whole brain in NMO, the CBF/ReHo ratio could provide more detailed information on the regional neurovascular coupling changes in this disorder. The regional neurovascular coupling (CBF/ReHo ratio) is like a balance with two ends. One end is neuronal activity (ReHo) and the other end is vascular response (CBF). The CBF/ReHo ratio keeps balance in healthy brain. In NMO, the deviation from the balance (abnormal neurovascular coupling) may result in either increased or decreased CBF/ReHo ratio. The former indicates redundant blood supply per unit of neuronal activity, whereas the latter denotes inadequate blood supply per unit of neuronal activity. The devastating effects of the CBF/ReHo ratio changes are confirmed by clinical correlation analyses, in which both abnormally decreased and increased CBF/ReHo ratios were found to be associated with more severe clinical impairments and cognitive decline in patients with NMO. More importantly, the analysis of CBF/ReHo ratio could identify abnormal regions without significant changes in both CBF and ReHo. In this situation, the balance may be disrupted if CBF and ReHo change in opposite direction. For example, the subtle increased CBF and decreased ReHo may result in significantly increased CBF/ReHo ratio in NMO, and the subtle decreased CBF and increased ReHo may result in significantly decreased CBF/ReHo ratio. These may explain why several regions with significant intergroup differences in CBF/ReHo ratio but not in CBF or ReHo.

The CBF/ReHo ratio is a quite novel imaging marker to assess regional neurovascular coupling changes in brain disorders. In the validation analyses, we confirmed that the CBF/ReHo changes are independent of GMV atrophy and disease severity. In the CBF/ALFF ratio analysis, we reproduced the pattern of CBF/ReHo changes and thus confirmed that the abnormal neurovascular coupling is a reliable finding in NMO. With functional connectivity strength (FCS) to assess neuronal activity, a previous study has used CBF/FCS ratio to explore neurovascular changes in schizophrenia (Zhu et al., 2017). The authors reported a completely different pattern of CBF/FCS changes in schizophrenia and correlations of this measure with clinical severity in schizophrenia patients (Zhu et al., 2017). These findings suggest that the CBF/BOLD ratio is a clinically meaningful imaging marker to assess regional neurovascular coupling in brain disorders.

In this study, we identified several brain regions whose altered neurovascular coupling was correlated with impairments in comprehensive or specific cognitive domains. All of these regions are important for cognitive processing, impairments of which have been associated with compromised performance in specific cognitive domains in diseased conditions (Loitfelder et al., 2014). For example, the parietal regions have been claimed to be related to visuospatial attention and working memory (Wang, Yang et al., 2015); and the abnormally decreased CBF/ReHo ratios in the parietal regions were really correlated with deficits in these specific cognitive domains in NMO. The insular and IFG have been found involved in language processing (Augustine, 1996; Keller, Crow, Foundas, Amunts, & Roberts, 2009); and the abnormally increased CBF/ReHo ratios in these regions were expectedly correlated with impairments in verbal learning and memory in NMO. Similarly, the association between abnormally increased CBF/ReHo ratio in the postcentral gyrus and deficits in visuospatial learning and memory in NMO is consistent with the role of the postcentral gyrus in visuospatial attention (Balslev, Odoj, & Karnath, 2013). Unexpectedly, the ACC is an important region for cognitive control (Bush, Luu, & Posner, 2000) and the lingual gyrus plays a role in visual memory (Bogousslavsky, Miklossy, Deruaz, Assal, & Regli, 1987); however, we failed to find any significant correlations between the abnormal CBF/ReHo ratio of these brain regions and cognitive scores. The underlying mechanisms should be investigated in future.

There are several limitations in this study. First, a slice thickness of 4 mm was used in resting‐state MR sequences to improve signal to noise ratio (SNR) because of the relatively low SNR in these sequences. The technical limit may reduce the precision in the calculation of ReHo and CBF. Second, the CBF‐ReHo correlation and CBF/ReHo ratio indirectly reflect rather than accurately measure neurovascular coupling and thus we cannot exactly figure out the specific pathophysiological mechanisms that lead to alterations of neurovascular coupling in NMO. Third, several subjects in our study were excluded from the correlation analyses because of the failure to complete the cognitive assessments. However, the smallest group still included 38 patients with NMO; we believe that our findings for these correlations are reliable. Finally, there are many statistical tests for the clinical correlation analyses, but only one of them could pass the FDR correction. A larger sample is needed to validate these nominal significant findings.

5. CONCLUSION

In conclusion, patients with NMO show the disrupted neurovascular coupling, which is associated with disease severity and cognitive impairments. Specifically, we found reduced CBF‐ReHo coupling in NMO, indicating a new neuropathological underpinning. We further revealed regional changes in neurovascular coupling and their relationship with disease severity and cognitive impairments in NMO, providing potential imaging markers for assessing clinical disability and cognitive impairments in patients with NMO. The exact biological mechanisms underlying altered neurovascular coupling in NMO should be investigated in the future.

ACKNOWLEDGMENTS

All the authors report no disclosures. The authors thank the patients as well as healthy volunteers for participating in this study, and members of the Departments of Radiology and Neurology for patient recruitment and collection of clinical data. Natural Science Foundation of China (81425013); National Key Research and Development Program of China (2018YFC1314300); Tianjin Key Technology R&D Program (17ZXMFSY00090).

Guo X, Zhu J, Zhang N, et al. Altered neurovascular coupling in neuromyelitis optica. Hum Brain Mapp. 2019;40:976–986. 10.1002/hbm.24426

Funding information: National Natural Science Foundation of China, Grant/Award Number: 81425013; Tianjin Key Technology R&D Program, Grant/Award Number: 17ZXMFSY00090; National Key Research and Development Program of China, Grant/Award Number: 2018YFC1314300; Natural Science Foundation of China, Grant/Award Number: 81425013

REFERENCES

  1. Augustine, J. R. (1996). Circuitry and functional aspects of the insular lobe in primates including humans. Brain Research. Brain Research Reviews, 22, 229–244. [DOI] [PubMed] [Google Scholar]
  2. Balslev, D. , Odoj, B. , & Karnath, H. O. (2013). Role of somatosensory cortex in visuospatial attention. The Journal of Neuroscience, 33, 18311–18318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Benjamini, Y. , & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society: Series B: Methodological, 57(1), 289–300. [Google Scholar]
  4. Blanc, F. , Zephir, H. , Lebrun, C. , Labauge, P. , Castelnovo, G. , Fleury, M. , … de Seze, J. (2008). Cognitive functions in neuromyelitis optica. Archives of Neurology, 65, 84–88. [DOI] [PubMed] [Google Scholar]
  5. Blanco‐Rojas, L. , Arboix, A. , Canovas, D. , Grau‐Olivares, M. , Oliva Morera, J. C. , & Parra, O. (2013). Cognitive profile in patients with a first‐ever lacunar infarct with and without silent lacunes: A comparative study. BMC Neurology, 13, 203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bogousslavsky, J. , Miklossy, J. , Deruaz, J. P. , Assal, G. , & Regli, F. (1987). Lingual and fusiform gyri in visual processing: A clinico‐pathologic study of superior altitudinal hemianopia. Journal of Neurology, Neurosurgery, and Psychiatry, 50, 607–614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bush, G. , Luu, P. , & Posner, M. I. (2000). Cognitive and emotional influences in anterior cingulate cortex. Trends in Cognitive Sciences, 4, 215–222. [DOI] [PubMed] [Google Scholar]
  8. Cai, H. , Zhu, J. , Zhang, N. , Wang, Q. , Zhang, C. , Yang, C. , … Yu, C. (2017). Subregional structural and connectivity damage in the visual cortex in neuromyelitis optica. Scientific Reports, 7, 41914. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Duan, Y. , Liu, Y. , Liang, P. , Jia, X. , Yu, C. , Qin, W. , … Li, K. (2012). Comparison of grey matter atrophy between patients with neuromyelitis optica and multiple sclerosis: A voxel‐based morphometry study. European Journal of Radiology, 81, e110–e114. [DOI] [PubMed] [Google Scholar]
  10. Filosa, J. A. , Morrison, H. W. , Iddings, J. A. , Du, W. , & Kim, K. J. (2016). Beyond neurovascular coupling, role of astrocytes in the regulation of vascular tone. Neuroscience, 323, 96–109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Fox, M. D. , & Raichle, M. E. (2007). Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging. Nature Reviews. Neuroscience, 8, 700–711. [DOI] [PubMed] [Google Scholar]
  12. Genovese, C. , A Lazar, N. , & Nichols, T. (2002). Thresholding of statistical maps in functional neuroimaging using the false discovery rate. NeuroImage, 15(4), 870–878. [DOI] [PubMed] [Google Scholar]
  13. Howarth, C. (2014). The contribution of astrocytes to the regulation of cerebral blood flow. Frontiers in Neuroscience, 8, 103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Jiang, L. , & Zuo, X. N. (2016). Regional homogeneity: A multimodal, multiscale neuroimaging marker of the human connectome. The Neuroscientist, 22, 486–505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kawachi, I. , & Lassmann, H. (2017). Neurodegeneration in multiple sclerosis and neuromyelitis optica. Journal of Neurology, Neurosurgery, and Psychiatry, 88, 137–145. [DOI] [PubMed] [Google Scholar]
  16. Keller, S. S. , Crow, T. , Foundas, A. , Amunts, K. , & Roberts, N. (2009). Broca's area: Nomenclature, anatomy, typology and asymmetry. Brain and Language, 109, 29–48. [DOI] [PubMed] [Google Scholar]
  17. Kim, S. H. , Park, E. Y. , Park, B. , Hyun, J. W. , Park, N. Y. , Joung, A. , … Kim, H. J. (2017). Multimodal magnetic resonance imaging in relation to cognitive impairment in neuromyelitis optica spectrum disorder. Scientific Reports, 7, 9180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Liang, P. , Liu, Y. , Jia, X. , Duan, Y. , Yu, C. , Qin, W. , … Li, K. (2011). Regional homogeneity changes in patients with neuromyelitis optica revealed by resting‐state functional MRI. Clinical Neurophysiology, 122, 121–127. [DOI] [PubMed] [Google Scholar]
  19. Liang, X. , Zou, Q. , He, Y. , & Yang, Y. (2013). Coupling of functional connectivity and regional cerebral blood flow reveals a physiological basis for network hubs of the human brain. Proceedings of the National Academy of Sciences of the United States of America, 110, 1929–1934. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liu, Y. , Fu, Y. , Schoonheim, M. M. , Zhang, N. , Fan, M. , Su, L. , … Shi, F. D. (2015). Structural MRI substrates of cognitive impairment in neuromyelitis optica. Neurology, 85, 1491–1499. [DOI] [PubMed] [Google Scholar]
  21. Liu, Y. , Jiang, X. , Butzkueven, H. , Duan, Y. , Huang, J. , Ren, Z. , … Wang, J. (2017). Multimodal characterization of gray matter alterations in neuromyelitis optica. Multiple Sclerosis Journal, 24, 1308–1316. [DOI] [PubMed] [Google Scholar]
  22. Liu, Y. , Liang, P. , Duan, Y. , Jia, X. , Wang, F. , Yu, C. , … Li, K. (2011). Abnormal baseline brain activity in patients with neuromyelitis optica: A resting‐state fMRI study. European Journal of Radiology, 80, 407–411. [DOI] [PubMed] [Google Scholar]
  23. Loitfelder, M. , Fazekas, F. , Koschutnig, K. , Fuchs, S. , Petrovic, K. , Ropele, S. , … Enzinger, C. (2014). Brain activity changes in cognitive networks in relapsing‐remitting multiple sclerosis ‐ insights from a longitudinal FMRI study. PLoS One, 9, e93715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lopes, F. C. , Alves‐Leon, S. V. , Godoy, J. M. , de Souza Batista Scherpenhuijzen, S. , Fezer, L. , & Gasparetto, E. L. (2015). Optic neuritis and the visual pathway: Evaluation of neuromyelitis Optica Spectrum by resting‐state fMRI and diffusion tensor MRI. Journal of Neuroimaging, 25, 807–812. [DOI] [PubMed] [Google Scholar]
  25. Masuda, H. , Hirano, S. , Takahashi, N. , Hatsugano, E. , Uzawa, A. , Uchida, T. , … Mori, M. (2017). Comparison of cognitive and brain grey matter volume profiles between multiple sclerosis and neuromyelitis optica spectrum disorder. PLoS One, 12, e0184012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Merlini, M. , Davalos, D. , & Akassoglou, K. (2012). In vivo imaging of the neurovascular unit in CNS disease. Intravital, 1, 87–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Misu, T. , Fujihara, K. , Kakita, A. , Konno, H. , Nakamura, M. , Watanabe, S. , … Itoyama, Y. (2007). Loss of aquaporin 4 in lesions of neuromyelitis optica: Distinction from multiple sclerosis. Brain, 130, 1224–1234. [DOI] [PubMed] [Google Scholar]
  28. Muoio, V. , Persson, P. B. , & Sendeski, M. M. (2014). The neurovascular unit ‐ concept review. Acta Physiologica (Oxford, England), 210, 790–798. [DOI] [PubMed] [Google Scholar]
  29. Phillips, A. A. , Chan, F. H. , Zheng, M. M. , Krassioukov, A. V. , & Ainslie, P. N. (2016). Neurovascular coupling in humans: Physiology, methodological advances and clinical implications. Journal of Cerebral Blood Flow and Metabolism, 36, 647–664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pichiecchio, A. , Tavazzi, E. , Poloni, G. , Ponzio, M. , Palesi, F. , Pasin, M. , … Bastianello, S. (2012). Advanced magnetic resonance imaging of neuromyelitis optica: A multiparametric approach. Multiple Sclerosis, 18, 817–824. [DOI] [PubMed] [Google Scholar]
  31. Popescu, B. F. , Parisi, J. E. , Cabrera‐Gomez, J. A. , Newell, K. , Mandler, R. N. , Pittock, S. J. , … Lucchinetti, C. F. (2010). Absence of cortical demyelination in neuromyelitis optica. Neurology, 75, 2103–2109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Power, J. D. , Barnes, K. A. , Snyder, A. Z. , Schlaggar, B. L. , & Petersen, S. E. (2012). Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. NeuroImage, 59, 2142–2154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Saczynski, J. S. , Inouye, S. K. , Guess, J. , Jones, R. N. , Fong, T. G. , Nemeth, E. , … Marcantonio, E. R. (2015). The Montreal cognitive assessment: Creating a crosswalk with the mini‐mental state examination. Journal of the American Geriatrics Society, 63, 2370–2374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Saji, E. , Arakawa, M. , Yanagawa, K. , Toyoshima, Y. , Yokoseki, A. , Okamoto, K. , … Kawachi, I. (2013). Cognitive impairment and cortical degeneration in neuromyelitis optica. Annals of Neurology, 73, 65–76. [DOI] [PubMed] [Google Scholar]
  35. Sanchez‐Catasus, C. A. , Cabrera‐Gomez, J. , Almaguer Melian, W. , Giroud Benitez, J. L. , Rodriguez Rojas, R. , Bayard, J. B. , … Valdes‐Sosa, P. (2013). Brain tissue volumes and perfusion change with the number of optic neuritis attacks in relapsing neuromyelitis optica: A voxel‐based correlation study. PLoS One, 8, e66271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Stobart, J. L. , & Anderson, C. M. (2013). Multifunctional role of astrocytes as gatekeepers of neuronal energy supply. Frontiers in Cellular Neuroscience, 7, 38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Takeshita, Y. , Obermeier, B. , Cotleur, A. C. , Spampinato, S. F. , Shimizu, F. , Yamamoto, E. , … Ransohoff, R. M. (2017). Effects of neuromyelitis optica‐IgG at the blood‐brain barrier in vitro. Neurology Neuroimmunology & Neuroinflammtion, 4, e311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Tam, J. W. , & Schmitter‐Edgecombe, M. (2013). The role of processing speed in the brief Visuospatial memory test ‐ revised. The Clinical Neuropsychologist, 27, 962–972. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tarantini, S. , Tran, C. H. , Gordon, G. R. , Ungvari, Z. , & Csiszar, A. (2016). Impaired neurovascular coupling in aging and Alzheimer's disease: Contribution of astrocyte dysfunction and endothelial impairment to cognitive decline. Experimental Gerontology, 94, 52–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Venkat, P. , Chopp, M. , & Chen, J. (2016). New insights into coupling and uncoupling of cerebral blood flow and metabolism in the brain. Croatian Medical Journal, 57, 223–228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. von Glehn, F. , Jarius, S. , Cavalcanti Lira, R. P. , Alves Ferreira, M. C. , von Glehn, F. H. , Costa, E. C. S. M. , … Yasuda, C. L. (2014). Structural brain abnormalities are related to retinal nerve fiber layer thinning and disease duration in neuromyelitis optica spectrum disorders. Multiple Sclerosis, 20, 1189–1197. [DOI] [PubMed] [Google Scholar]
  42. Wang, J. , Yang, Y. , Fan, L. , Xu, J. , Li, C. , Liu, Y. , … Jiang, T. (2015). Convergent functional architecture of the superior parietal lobule unraveled with multimodal neuroimaging approaches. Human Brain Mapping, 36, 238–257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wang, Q. , Zhang, N. , Qin, W. , Li, Y. , Fu, Y. , Li, T. , … Yu, C. (2015). Gray matter volume reduction is associated with cognitive impairment in neuromyelitis optica. AJNR. American Journal of Neuroradiology, 36, 1822–1829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Wingerchuk, D. M. , Lennon, V. A. , Lucchinetti, C. F. , Pittock, S. J. , & Weinshenker, B. G. (2007). The spectrum of neuromyelitis optica. Lancet Neurology, 6, 805–815. [DOI] [PubMed] [Google Scholar]
  45. Wingerchuk, D. M. , Lennon, V. A. , Pittock, S. J. , Lucchinetti, C. F. , & Weinshenker, B. G. (2006). Revised diagnostic criteria for neuromyelitis optica. Neurology, 66, 1485–1489. [DOI] [PubMed] [Google Scholar]
  46. Xu, G. , Rowley, H. A. , Wu, G. , Alsop, D. C. , Shankaranarayanan, A. , Dowling, M. , … Johnson, S. C. (2010). Reliability and precision of pseudo‐continuous arterial spin labeling perfusion MRI on 3.0 T and comparison with 15O‐water PET in elderly subjects at risk for Alzheimer's disease. NMR in Biomedicine, 23, 286–293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zang, Y. , Jiang, T. , Lu, Y. , He, Y. , & Tian, L. (2004). Regional homogeneity approach to fMRI data analysis. NeuroImage, 22, 394–400. [DOI] [PubMed] [Google Scholar]
  48. Zang, Y. F. , He, Y. , Zhu, C. Z. , Cao, Q. J. , Sui, M. Q. , Liang, M. , … Wang, Y. F. (2007). Altered baseline brain activity in children with ADHD revealed by resting‐state functional MRI. Brain Dev, 29, 83–91. [DOI] [PubMed] [Google Scholar]
  49. Zhu, J. , Zhuo, C. , Xu, L. , Liu, F. , Qin, W. , & Yu, C. (2017). Altered coupling between resting‐state cerebral blood flow and functional connectivity in schizophrenia. Schizophrenia Bulletin, 43, 1363–1374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Zlokovic, B. V. (2010). Neurodegeneration and the neurovascular unit. Nature Medicine, 16, 1370–1371. [DOI] [PubMed] [Google Scholar]

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES