Grey matter demyelination is an important facet of multiple sclerosis (MS) physiopathology, which has been associated with cognitive disabilities¹. Although MRI is the gold standard for the diagnosis and monitoring of MS, the vast majority of cortical alterations present in MS patients still fail to be detected. Development of ultra-short echo time (UTE) has enabled the detection of myelin sheats² and thus hold great potential for the non-invasive detection of demyelinating lesions in both white and grey matter.

The cuprizone (CPZ) mouse model is a well-established model of central nervous system (CNS) toxic demyelination and spontaneous remyelination³. CPZ administration results in variable spatial and temporal patterns of demyelination in white and grey matter within the CNS, specifically the corpus callosum (CC). This model allows the study of several degree of demyelination and associated neuroinflammation within the same animal.

In the present work, we examined the temporal changes of UTE measures following acute CPZ-induced inflammation/demyelination and evaluated the sensitivity of in vivo UTE measurements for the detection of cortical grey matter and white matter alterations. Importantly, we compared the sensitivity of UTE with that of conventional T₂-weighted (T₂w) imaging on a clinical 7T system. We performed longitudinal UTE and T₂w MRI at key time points of white matter inflammation/demyelination, cortical demyelination and subsequent remyelination. We showed that UTE was more specific to myelin integrity than conventional T₂w imaging, thus demonstrating its potential clinical value for diagnosis and monitoring of MS patients.

Animals: C57BL/6J mice (n=19) received a 0.2% CPZ diet for 6 weeks to induce demyelination and neuroinflammation. Animals were returned to a standard rodent chow to allow for remyelination. A total of six mice were imaged longitudinally prior (W0) and after 4 and 6 weeks of CPZ (W4, W6) and after 6 weeks after returning to a normal diet (W6+6), (Fig.1A).

MR acquisitions: 2D multi-slice T2w images were acquired on a GE 7T system (GE Healthcare, Waukesha, WI) using the following parameters: TR = 2 s, TE = 13.6 ms , NA = 3, Voxel size = 250*250µm2 , slice thickness = 1 cm. 3D UTE with off-resonance saturation contrast (OSC) images were acquired using: TR = 20ms , TE=,76µs NA = 4, Voxel sixe = 200*200*500µm3, saturation pulse = 90 degrees; -1800 Hz off-resonance.

Data analysis: UTE off-resonance saturation ratio (OSR) images were reconstructed using the following formula: MUTE-OSR = (|Mno sat| - |Msat|) / |Mnosat|. Regions of interest (ROI) were manually delineated with AMIRA software (Mercury Computer systems, San Diego, USA) on T₂w images. Regional average signal was normalized to the mean cerebral spinal fluid signal of the third ventricle. In a similar manner, ROIs were delineated on UTE-OSR images. White matter regions include the splenium and the genu of the CC and grey matter regions the somatosensory cortices and caudate-putamen (Fig.1B). Results are presented as mean signal intensity ±SD and statistical significance between time points was evaluated using repeated measures (IBM SPSS Statistics v.23, *p<0.05, **p<0.01, ***p<0.005).

Immunofluorescence (IF): IF analyses were performed for myelin interity (MBP).

Both mean UTE-OSR and T₂w signal intensity were able to detect demyelination induced by CPZ after 4 and 6 weeks of CPZ administration at the level of the splenium (Fig.2A-B). However, only UTE-OSR could distinguish the spontaneous remyelination that occurs between W4 and W6 of CPZ (p=0.0207).

Additionally, UTE-OSR was sensitive to long-lasting impairment of myelination after remyelination has occurred (p=0.044). In the somatosensory cortices, maximal demyelination usually occurs after 6W of CPZ, and is followed by spontaneous remyelination once CPZ is withdrawn from the diet. While T₂w imaging failed to detect any cortical alterations (Fig.2C), it is interesting to note that UTE-OSR was able to determine long lasting effects related to a demyelinating event, and therefore suggests that UTE-OSR may also be indicative of additional processes linked to tissue remodelling.

Similarly, UTE-OSR detected demyelination and long-lasting changes in the caudate putamen (Fig.3A) and the genu of the CC, a region mildly affect by CPZ diet (Fig.3B). Further IF analyses are currently ongoing to evaluate myelin integrity at each time points and associated neuroinflammatory events including microgliosis and astrogliosis.

In conclusion, our results demonstrate that UTE-OSR shows higher sensitivity to myelin integrity in grey and white matter as compared to conventional T2w imaging. This method could thus improve the diagnosis and monitoring of MS patients on clinical 7T MR systems.