Multimodal Characterization of Grey Matter Alterations in Neuromyelitis Optica
Yaou Liu1, Yunyun Duan2, Huiqing Dong2, Tianyi Qian2, Frederik Barkhof3, Jinhui Wang4, and Kuncheng Li2

1Xuanwu Hospital,Capital Medical University, Beijing, China, People's Republic of, 2Beijing, China, People's Republic of, 3Amsterdam, Netherlands, 4Hanzhou, China, People's Republic of

Synopsis

Combining double-inversion-recovery (DIR), high-resolution structural MRI, diffusion tensor imaging (DTI) and resting-state functional MRI (rs-fMRI), this study systematically investigated structural and functional alterations in grey-matter (GM) structures in thirty-five neuromyelitis optica (NMO) patients compared with healthy controls. We demonstrated that NMO exhibits both structural and functional alterations of GM in the cerebrum and cerebellum. Multimodal MRI techniques complementary worked to capture NMO-related GM abnormalities. GM alterations, especially diffusion abnormalities, correlated with cognitive impairment in NMO. These findings have important implications for understanding the roles of GM damage and also for highlighting multimodal MRI techniques as objective biomarkers in NMO.

PURPOSE

To systematically investigate structural and functional alterations of grey matter (GM) and examine their clinical relevance in patients with neuromyelitis optica (NMO) using multimodal MRI techniques.

METHODS

Thirty-five NMO patients and 36 age- and gender-matched healthy controls (HC) were recruited in this study. All MRI scans were performed on a MAGNETOM Trio Tim 3T MR system (Siemens Healthcare, Erlangen, Germany) . The scan protocol included conventional sequences (axial T2WI and FLAIR), double inversion recovery (DIR), diffusion tensor imaging, and rs-fMRI. Based on the multimodal MRI data acquired above, we analyzed WM lesions from T2-weighted and FLAIR images, GM lesions from DIR images, GM volumes from T1 images, microstructural integrity (fractional anisotropy, FA and mean diffusivity, MD) from diffusion images, and intrinsic functional architecture including amplitude of low-frequency fluctuation, ALFF (local oscillation power)1 and weighted functional connectivity strength, wFCS (inter-regional functional connectivity over the whole brain )2, 3 from rs-fMRI. The rs-fMRI data were post-processed using REST toolbox and in combination with an in-house pipeline developed in MATLAB. For the multimodal MRI measures, a voxel-wise general linear model was used to infer differences between groups. The Dice coefficient and partial correlation analyses were used to estimate cross-modality spatial overlaps and quantitative relationships. Partial correlation was also used to estimate clinical-MRI relationships.

RESULTS

WM and GM lesions 20 out of the 35 NMO patients showed no WM lesions on cerebral MRI and 15 had non-specific WM lesions. The mean T2LV was 5.12 (± 3.86) mL for the patients. No cortical lesions were identified in the NMO patients based on DIR. Morphological and microstructural organizational differences between groups Compared with the HC, the NMO patients showed significantly decreased GMV in the bilateral thalamus, insula, middle temporal gyrus and anterior cingulate, the right caudate and inferior frontal gyrus, and the left calcarine sulcus (p < 0.05, corrected) (Figure 1A). For diffusion-based MD and FA, widespread brain regions in both the cerebrum and cerebellum were abnormal compared with the HC (Figure 1B and 1C). Functional measurement differences between groups Limited and regionally specific NMO-related alterations were found in the functional scans. Specifically, for the ALFF, significant decreases were mainly observed in sensorimotor and visual cortex (p < 0.05, corrected) (Figure 1D), while for the wFCS, NMO-related decreases were primarily observed in visual and cognition-related areas and the cerebellum (p < 0.05, corrected) (Figure 1E). Cross-modality and MRI-clinical relationship Figure 2 shows spatial overlaps among the five multimodal MRI-based measures in revealing NMO-related GM alterations. For the NMO patients, the scatter plots of all clinical correlations are shown in Figure 3.

DISCUSSION

Structural and functional alterations were observed in GM in NMO. Specifically, the VBM analysis revealed regional GM atrophy in multiple regions. This atrophy could be caused by primary cortical pathology such as an inflammatory-degenerative process induced by Aquaporin-4 (AQP4) or secondary trans-synaptic degeneration. Our study showed that diffusion changes are widely distributed in brain areas implicated in visual, sensorimotor and cognitive functions. Given that a serum immunoglobulin G autoantibody (NMO-IgG) selectively binds to AQP4 4 in NMO, we speculate the disruption of water homeostasis may be one key pathophysiological basis for the observed diffusion changes. For functional measures, our results showed subtle functional alterations, mainly in motor, visual and cognitive related areas, corresponding to the symptoms in NMO, and consistent with previous findings.5 We observed that most of the brain areas with decreased GMV in NMO are those also harboring diffusion changes, and this is supported by the correlation analysis, which revealed significant associations between regional FA/MD and GMV in the patients. Subtle functional MRI changes were limited to areas with diffusion abnormalities and correlated with FA, implying microstructural alteration, even without or prior to cortical atrophy development inducing resting-state functional damages. Significant correlation with cognitive scores of the patients was found for GM structural changes, particularly microstructural abnormalities. This implies that GM structural changes may serve as potential biomarkers for assessing and monitoring cognitive impairment in NMO.

CONCLUSION

Despite the lack of focal cortical lesions, patients with NMO exhibit both structural and functional alterations of GM in cerebrum and cerebellum that predominantly involve deep GM, specifically in visual, motor and cognitive regions. Complementary multimodal MRI techniques were used to capture NMO-related GM abnormalities, but DTI changes were most widespread. GM alterations, especially diffusion abnormalities, are associated with both disease duration and number of relapses, and correlate with cognitive impairment but not physical disability in NMO.

Acknowledgements

This work was supported by the ECTRIMS-MAGNMIS Fellowship from ECTRIMS (Y.L), the National Science Foundation of China (Nos. 81101038, 30930029, 81471221 and 81230028), the National Basic Research Program of China (2013CB966900) the Beijing Natural Science fund (No.7133244), and the Beijing Nova Programme (xx2013045).

References

1. Zang YF, He Y, Zhu CZ, et al. Altered baseline brain activity in children with ADHD revealed by resting-state functional MRI. Brain Dev 2007;29:83-91.

2. Cole MW, Pathak S, Schneider W. Identifying the brain's most globally connected regions. Neuroimage 2010;49:3132-3148.

3. Tomasi D, Volkow ND. Functional connectivity density mapping. Proc Natl Acad Sci U S A 2010;107:9885-9890.

4. Lennon VA, Wingerchuk DM, Kryzer TJ, et al. A serum autoantibody marker of neuromyelitis optica: distinction from multiple sclerosis. Lancet 2004;364:2106-2112.

5. Liu Y, Liang P, Duan Y, et al. Abnormal baseline brain activity in patients with neuromyelitis optica: a resting-state fMRI study. Eur J Radiol 2011;80:407-411.

Figures

Figure1. Differences between groups in GMV (A), FA (B), MD (C), ALFF (D) and wFCS (E) in the cerebrum. (ALFF = amplitude of low-frequency fluctuation; GMV= grey matter volume; FA = fractional anisotropy; GMV = gray-matter volume; MD = mean diffusivity; wFCS = weighted functional connectivity strength)

Figure 2. Cross-modality relationships. A, spatial overlap map among the five measures in revealing NMO-related alterations. B, pairwise Dice coefficients in the spatial extent. C, pairwise spatial overlap maps. D, pairwise correlations in the patient group. *, P < 0.05. **, P < 0.01. ***, P < 0.005.

Figure 3. Relationships between multimodal MRI-based measures and clinical variables. ( CAL= calcarine; FA = fractional anisotropy; GMV = grey-matter volume; HIP=hippocampus; MD = mean diffusivity; NR=number of relapse; PASAT = paced auditory serial addition test; .PHG= parahippocampus; STG= superior temporary gyrus; THA=thalamus)



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4074