Neuromyelitis optica spectrum disorder (NMOSD) is an inflammatory demyelinating disease mediated by inflammation against the aquaporin-4 (AQP4) water channel on astrocytes predominantly attacked optic nerves and spinal cord. For over half of NMOSD patients, optic neuritis (ON), characterized by decreased vision, pain during eye movement, and visual field defects (Vabanesi et al., 2019; Wu et al., 2019), is the first clinical feature (Oertel et al., 2018), which can trigger severe visual dysfunction accompanied by poor visual outcomes (Sotirchos et al., 2020). It has been reported that about three-fifths of NMOSD patients in the United States (Wingerchuk et al., 1999) were functionally blind in at least one eye, and nearly one-fifth of cases in the United Kingdom and Japan (Kitley et al., 2012) developed to permanent bilateral visual deficits.

Most neuroimaging studies of NMOSD have reported structural damage predominantly restricted to white matter, especially for optic radiation which is characterized by decreased FA and increased MD using diffusion tensor imaging (DTI; Yu et al., 2008; Jeantroux et al., 2012; Liu et al., 2012; Rueda et al., 2012; Zhao et al., 2012). The above findings are widely accepted as the mechanism of axonal degeneration secondary to ON. Subsequently, increasing studies have shown subtle morphological alteration in the visual cortex beyond the optic radiation, featured by reduced gray matter volume (Pichiecchio et al., 2012; Von Glehn et al., 2014; Liu et al., 2018) and cortical atrophy (Calabrese et al., 2012; Liu et al., 2014) in NMOSD patients. At present, anterograde trans-synaptic degeneration through the lateral geniculate nuclei has been considered as the leading mechanism to explain the reduced volume or thickness of the primary visual cortex in NMOSD patients with ON (Von Glehn et al., 2014; Manogaran et al., 2016; Tian et al., 2018), but the evidence for this mechanism remains inadequate.

Motivated by previous work, we performed diffusion kurtosis imaging (DKI) to explore the microstructural alterations of cerebral gray matter (cortex and/or subcortex) that likely precede detectable volumetric changes at the macroscopic level (Gong et al., 2017). On the one hand, DKI is a higher-order extension of DTI and provides a more accurate description of non-Gaussian diffusion of water molecules in heterogeneous and complex brain tissue than DTI that is a fundamental limitation given its assumption of Gaussian diffusion of biological water (Lu et al., 2006; Jensen and Helpern, 2010; Bester et al., 2015). On the other hand, DKI is sensitive to capturing the microstructural characteristics in isotropic tissues (gray matter) (Jensen et al., 2005), another limitation of DTI (Bester et al., 2015; McKenna et al., 2019). To date, DKI has been widely used in multiple diseases, such as schizophrenia (McKenna et al., 2019), Alzheimer’s disease (Xue et al., 2019), and normal tension glaucoma (Li et al., 2020). These findings imply a key role of DKI to investigate brain alteration.

Diffusion kurtosis imaging additionally provides three diffusion kurtosis tensor derived metrics, including MK, RK, and AK (Jensen and Helpern, 2010; Bester et al., 2015). MK serves as a measure of the overall kurtosis, suggesting the microstructural complexity and heterogeneity (Hui et al., 2008; Wu and Cheung, 2010). RK and AK are usually used to characterize the compactness of myelin and axon (Cheung et al., 2009; Kamagata et al., 2017). FA and MD, two commonly quoted metrics that also can be derived from DKI, are specific to assessing the integrity of myelin. Li et al. (2020) have reported that MK, AK, and FA were reduced in bilateral visual cortices in patients with normal tension glaucoma. Therefore, DKI may pave a new sight to deeper understand the neuropathology of NMOSD. In this study, we conducted a DKI study in NMOSD patients only with ON to characterize the altered microstructure in cerebral gray matter and its correlations with clinical features.

Compared with NCs, NMOSD-ON patients showed significantly decreased MK and RK both in the left ITG, and decreased AK in the bilateral CAL. While the abnormalities were found with the increased MD involving the left FFG, right CAL, and right HIP/PHG. There was no significant difference in FA. Moreover, correlation analysis showed that the mean deviation was negatively correlated with AK values of bilateral CAL and positively correlated with MD values of right CAL, showing the clinical implications of these findings in NMOSD-ON.

In the present study, significantly decreased kurtosis metrics (MK, RK, and AK) and increased MD were detected in the visual cortices of NMOSD-ON, including bilateral CAL, left ITG, and left FFG. Structurally, the visual pathway consists of the retina, optic nerve, chiasm, lateral geniculate nucleus, optic radiation, and visual cortex. The CAL of the occipital visual cortex is a part of Brodmann’s area 17 (BA17), BA18, and BA19. BA17, the primary visual cortex, and is densely connected with the optic nerve. BA18 and BA19 serve as the secondary visual cortex to integrate visual information received from BA17 and transmit it to the ventral and dorsal stream beyond the occipital lobe (Goodale and Milner, 1992). The ITG and FFG are the portions of the ventral stream where the decreased MK and RK and the increased MD suggested disability in visual stimuli processing, object recognition, and memory recall (Ishai et al., 1999; Haxby et al., 2000). In summary, the abnormal DKI metrics in these areas thus suggested that both primary and higher visual cortices are impaired. Several studies also have demonstrated the reduced volume or thickness in primary and higher visual cortices by voxel-based morphometry (VBM) and cortical analyses (Freesurfer) (Calabrese et al., 2012; Duan et al., 2012; Von Glehn et al., 2014; Wang et al., 2015; Tian et al., 2018). Beyond the abovementioned volumetric changes at the macroscopic level, we found cerebral cortical damage at the microscopic level. Decreased MK represents the reduced complexity of microstructure, resulting from neuronal edema, membrane structural damage, and myelin injury. Decreased RK and AK represent decreased compactness of myelination and fiber, while increased MD reflects the injury of myelin. As trans-synaptic degeneration is considered as the underlying mechanism of NMOSD (Von Glehn et al., 2014; Manogaran et al., 2016; Tian et al., 2018), we speculate that the microstructural damage in visual cortices could be attributed to it. Histologically, cortical lesions in NMOSD patients are characterized by the substantial loss of astrocytes, with secondary loss of either oligodendrocytes or neurons owing to astrocyte dysfunction and/or inflammatory damage (Kawachi and Lassmann, 2017; Jarius et al., 2020). Therefore, these DKI derived metrics characterized the microstructural damage likely attribute to secondary damage of the oligodendrocytes, and further support for the mechanism of trans-synaptic degeneration.

The HIP/PHG were found with the increased MD, consistent with the previous study (Liu et al., 2018). The HIP/PHG are subcortical structures that are parts of the limbic system and play a crucial role in cognitive function (Bush et al., 2000; Botvinick et al., 2004). Liu et al. (2015) have reported that the gray matter volume reductions in these regions may affect cognitive performance. As we didn’t perform neuropsychological assessment, there was no quantitative evidence to evaluate their cognitive problems. The microstructural alteration in these cognitive-related regions that are not directly connected with the visual cortex at least cannot be explained by trans-synaptic degeneration. We suspected that the bottom-up (from visual regions to cognitive-related regions) or top-down (from cognitive-related regions to visual regions) process could be the involved mechanism (McMains and Kastner, 2011). But the detail of the specific contribution needs to be clarified by future studies.

Mean deviation derived from Humphrey visual field is a commonly used index to evaluate visual field defects. We found that it was negatively correlated with AK values in the bilateral CAL and positively correlated with MD in the right CAL. The altered microstructure in these visual cortices in NMOSD-ON patients was mainly caused by anterograde trans-synaptic degeneration secondary to the damage of the optic nerve. These data thus suggest a potential role of microstructural damage in the visual cortices in predicting disability in NMOSD-ON patients.

Some limitations are present in this study. First, our findings are based on few patients. NMOSD is a rare disease, and it is difficult to perform large-scale studies, especially for patients with ON alone. Second, our findings are from a single imaging modality obtained in this sample. Recent studies have highlighted the importance of multimodal integration rather than the use of a single modality in revealing neural substrates underlying various brain disorders. In the future, we will conduct a multimodal study to further explore the alterations of cerebral gray matter among patients with ON, myelitis, or both ON and myelitis by cross-sectional designs. Finally, it is difficult to confirm whether these abnormalities are correlated with cognitive performance due to the lack of neuropsychological tests. We believe that this issue will be paid more attention.

In summary, microstructural damage in the occipital visual cortex, correlating with visual deficits in NMOSD-ON patients, potentially acts as a substrate to explore the mechanisms underlying NMOSD-specific neurodegeneration. Moreover, the HIP/PHG beyond the visual cortex were involved, which may provide a novel insight into exploring its relationship with the loss of visual-related input.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Beijing Tongren hospital committee. The patients/participants provided their written informed consent to participate in this study.

HZ: conceptualization, methodology, formal analysis, investigation, writing original draft, writing review and editing, visualization, and funding acquisition. QL: data curation, conceptualization, methodology, formal analysis, investigation, visualization, and funding acquisition. LL: resources, data curation, conceptualization, and investigation. XQ: methodology, writing – review and editing, and visualization. QW: writing – review and editing and visualization. BY: visualization. JX: conceptualization, methodology, resources, supervision, project administration, and funding acquisition. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.