5273

Surface-based Quantitative Susceptibility Mapping of Cortical Pathology in Multiple Sclerosis
Marco Castellaro1,2, Roberta Magliozzi2,3, Alessandro Palombit1, Stefania Montemezzi4, Francesca Benedetta Pizzini4, Alessandra Bertoldo1, and Massimiliano Calabrese2

1Department of Information Engineering, University of Padova, Padova, Italy, 2Department of Neurological, Biomedical and Movement Sciences, University of Verona, Verona, Italy, 3Faculty of Medicine, Imperial College London, London, United Kingdom, 4University of Verona, Verona, Italy

Synopsis

Multiple Sclerosis has been showed to be characterized by extent cortical sub-pial demyelination and iron alterations. Moreover, a “surface-in” gradient of pathology has been showed to be present in MS. In this study, we used QSM to investigate iron and/or myelin changes in the whole cortex. Moreover, we exploit surface-based methods to clarify the presence of a laminar specific changes in cortical susceptibility of Relapsing-Remitting MS patients.

Purpose

Histopathological examinations of multiple sclerosis MS brains indicate that sub-pial demyelinating lesions appeared topographically related to focal meningeal inflammation1. It has been hypothesized that cortical demyelination in multiple sclerosis may be driven by meningeal inflammation, accompanied by a decreasing gradient of demyelination from the pial surface. This fact has also been linked to alterations in iron content. The hypothesis that we want to test is the possible presence of changes in cortical iron/myelin content. Quantitative Susceptibility Mapping (QSM) has been used to study mainly the iron dis-regulation in Deep Grey Matter of MS patients. In this study, we extend the study of the susceptibility to the whole cortex of MS patient exploiting a high-resolution 3D Echo Planar Imaging sequence2,3, a surface-mapping method based on the Human Connectome Pipeline4 and non-parametric permutation testing5.

Methods

Subjects: A group of 30 RRMS patients (40.5±10.3 years mean±std; m:f 10:20; EDSS median 2, average disease duration 9.2±6.3 years) and 40 age- and sex-matched healthy controls (HC; 39.5±10.5 years; m:f 16:24) was examined in the current study.

Data acquisition: Participants were imaged at 3T (Philips Achieva TX R3.2.3.2) using a 3D single-echo segmented echo-planar-imaging (EPI) sequence1 (0.55x0x0.55x0.55 mm, TE/TR=29ms/51ms, FA=20°) 3D T1-weighted Turbo Field Echo sequence (1x1x1 mm, TE/TR=4.2ms/8.3ms, FA=8°) 3D FLAIR (1x1x1 mm, TE/TR=268/5500ms, TI=1850ms).

Analysis: We reconstructed quantitative susceptibility maps (χ) from the single-echo 3D-EPI using a Total Generalized Variation (TGV) algorithm which incorporates individual steps of phase unwrapping, background field removal and dipole inversion in a single iteration3,6. 3D FLAIR was used to semi-automatically segment MS lesions that were then used to perform lesion filling on the 3D T1. Filled images were then analysed with Freesurfer to reconstruct WM/GM and pial surface boundaries. Surfaces were manually checked and corrected for topological defects. These surfaces were used to compute equi-volume cortical surface layers by sampling the surface vertices in steps of 25% of cortical volume (0%: depth 1 - WM/GM, 100%: depth 5 - pial surface) across the entire cortex (Figure 1 and 2). The resulting surfaces were re-sampled on the original 3D-EPI space to extract χ values for each vertex and each layer depth. To perform group comparison, we registered individual subject’s native-mesh surfaces to Conte69 population-average surfaces with L-R correspondence7. Group-wise χ differences in the cortex were analysed, layer-by-layer, using a permutation-based test with PALM5 employing threshold-free cluster enhancement and differences were considered statistically significant with p-value<0.05 family-wise corrected to address multiple comparisons. We excluded both the WM/GM surface and the pial surface from the statistical analysis due to the low reliability of the QSM map in those areas. The cortex was parcellated in ROI (Braincolor8 http://braincolor.mindboggle.info/protocols/index.html) and the percentage of the statistically significant vertices was calculated as a measure of extension of χ changes.

Results

A “surface-in” gradient, driven from the pial surface, of increased susceptibility can be seen in several regions of the cortex of the RRMS group (Figure 3 – left). Moreover, we observed an increased extent of changes in χ over the cortex depending on the considered layer (Figure 4). Permutation tests revealed significant differences between the two groups in extent regions of the cortex (p-value reported in Figure 3 - right). Most of the differences showed an increased χ in the RRMS group compared to the HC group (χ changes - Figure 3 - left). Figure 4 reports the maximum p-value associated with each ROI and the percentage of the extension of the area associated to a change in χ.

Discussion

This study reveals the presence of an extended change in χ of cortical RRMS patients compared to HC. Remarkably, the RRMS group showed always an increased χ, which could be linked to iron accumulation, since iron is the more abundant paramagnetic compound in the GM tissue. However, we cannot exclude that the presence of demyelinated tissue or atrophy9,10 could have contributed to this increase. The most interesting finding is that this change is depth-dependent, showing more spread differences in the more superficial layers of the cortex compared to the inner ones. Further studies are necessary to better disentangle the contribution of cortical demyelination and/or cell alterations8 that requires the estimation of χ, R2*, R1 and atrophy.

Conclusion

The presence of a depth-mediated change of χ in the cortex of RRMS patients is reported. The differences are more extended in the more superficial layers of the cortical ribbon than the inner ones, providing a support for the hypothesis of a possible gradient “surface-in” of iron alterations accompanied with cortical demyelination in cortical areas of RRMS patients.

Acknowledgements

No acknowledgement found.

References

1. Magliozzi, R., Howell, O. W., Reeves, C., Roncaroli, F., Nicholas, R., Serafini, B., ... & Reynolds, R. (2010). A gradient of neuronal loss and meningeal inflammation in multiple sclerosis. Annals of neurology, 68(4), 477-493.

2. Sati, P., Thomasson, D. M., Li, N., Pham, D. L., Biassou, N. M., Reich, D. S., & Butman, J. A. (2014). Rapid, high-resolution, whole-brain, susceptibility-based MRI of multiple sclerosis. Multiple Sclerosis Journal, 20(11), 1464-1470.

3. Langkammer, C., Bredies, K., Poser, B. A., Barth, M., Reishofer, G., Fan, A. P., ... & Ropele, S. (2015). Fast quantitative susceptibility mapping using 3D EPI and total generalized variation. Neuroimage, 111, 622-630.

4. Glasser, M. F., Sotiropoulos, S. N., Wilson, J. A., Coalson, T. S., Fischl, B., Andersson, J. L., ... & Van Essen, D. C. (2013). The minimal preprocessing pipelines for the Human Connectome Project. Neuroimage, 80, 105-124.

5. Winkler, A. M., Ridgway, G. R., Webster, M. A., Smith, S. M., & Nichols, T. E. (2014). Permutation inference for the general linear model. Neuroimage, 92, 381-397.

6. Castellaro, M., Magliozzi, R., Palombit, A., Pitteri, M., Silvestri, E., Camera, V., ... & Calabrese, M. (2017). Heterogeneity of Cortical Lesion Susceptibility Mapping in Multiple Sclerosis. AJNR. American journal of neuroradiology. In press.

7. Van Essen, David C. "Cortical cartography and Caret software." Neuroimage 62.2 (2012): 757-764.

8. Klein, A., & Tourville, J. (2012). 101 labeled brain images and a consistent human cortical labeling protocol. Frontiers in neuroscience, 6.

9. Marques, J. P., Khabipova, D., & Gruetter, R. (2017). Studying cyto and myeloarchitecture of the human cortex at ultra-high field with quantitative imaging: R 1, R 2* and magnetic susceptibility. NeuroImage, 147, 152-163.

10. Schweser, F., Martins, A. L. R. D., Hagemeier, J., Lin, F., Hanspach, J., Weinstock-Guttman, B., ... & Zivadinov, R. (2017). Mapping of thalamic magnetic susceptibility in multiple sclerosis indicates decreasing iron with disease duration: A proposed mechanistic relationship between inflammation and oligodendrocyte vitality. NeuroImage, In press.

Figures

Figure 1. Representative example of a QSM map obtained from a 3D-EPI acquisition obtained with the Total Generalized Variation Algorithm (TGV) in the three canonical views. Surfaces of Pial (Yellow) and WM/GM (Red) boundaries have been obtained with freesurfer from a 3D-T1 with previously filled hypointensities (mask of lesion obatined from 3D Flair). Human connectome project workbench have been employed to delineate intermediate surfaces between WM/GM and pial surface are reported, respectively at 25% distance from WM (violet) 50% (cyan) 75% (green).

Figure 2. Average value of susceptibility obtained with a Total Generalized Variation algorithm. The left column shows the Multiple Sclerosis group whereas the right column the Healthy Controls group. Each row represents the different cortical depth from the white matter (WM) up to the pial surface.

Figure 3. Left Panel) Inflated cortical surfaces with superimposed the differences in susceptibility between the two groups (Hc-MS) Right Panel) Correspondent p-value resulting from the permutation based t-test for group-level susceptibility differences.

Figure 4. Regions of interest (ROI) that showed a change in susceptibility, for each ROI and each layer depth is reported the maximum p-value and the percentage of vertices that showed different magnetic susceptibility values between HC and RRMS subjects.



Proc. Intl. Soc. Mag. Reson. Med. 26 (2018)
5273