Gabriel Mangeat1,2, Russell Ouellette2,3, Constantina Andrada Treaba2,3, Tobias Granberg2,3, Elena Herranz2,3, Celine Louapre2,3, Nikola Stikov1,4, Jacob A. Sloane3,5, Eric C. Klawiter2,3,6, Caterina Mainero2,3, and Julien Cohen-Adad1,7
1Polytechnique Montreal, Montreal, QC, Canada, 2Athinoula A. Martinos Center for Biomedical Imaging, MGH, Charlestown, MA, United States, 3Harvard Medical School, Boston, MA, United States, 4Montreal Health Institute, Montreal, QC, Canada, 5Beth Israel Deaconess Medical Center, Boston, MA, United States, 6Department of Neurology, MGH, Boston, MA, United States, 7CRIUGM, Functional Neuroimaging Unit, UniversiteĢ de MontreĢal, Montreal, QC, United States
Synopsis
Multiple sclerosis (MS) is a chronic disorder of the central nervous system characterized by diffuse abnormalities along white matter tracts and demyelination, including cortical lesions. In this study, we explored the interplay between cortical and brain structural networks integrity in a cohort of early MS subjects by combining quantitative mapping of T2* and T1 relaxation rates from 7T MRI acquisitions to measure cortical demyelination with diffusion imaging and graph theory to assess the structural brain architecture. Results suggest that motor, premotor and anterior cingulate cortices are affected simultaneously by cortical demyelination and connectomics alterations, at a very early stage of MS.Purpose
Multiple sclerosis (MS) is a chronic disorder of the central nervous system typically characterized by white matter (WM) lesions and diffuse abnormalities along WM tracts1,2. Histopathological examinations also reported the presence of cortical demyelinating lesions3,4, which can occur early in the course of the disease but are hardly seen at conventional field strength MRI due to low contrast.
In this study, we explored the interplay between cortical and brain structural networks integrity in a cohort of early MS subjects by combining multivariate statistics of quantitative mapping of T2* and T1 relaxation rates from 7T MRI acquisitions to measure cortical demyelination5 with diffusion imaging and graph theory to assess the structural brain architecture2,6 (the so-called connectomics). Specifically, we investigated whether i) cortical demyelination and abnormal structural connectomics could be detected from the very early MS stages, ii) disrupted connectomics and cortical integrity could be two interrelated events in the MS pathophysiology.
Methods
Data acquisition. Thirteen early MS patients (38±9years; 10 females; disease duration ≤3years, median/range EDSS=1/[0-3]) and 11 healthy controls (HC, 30±9years, 6 females) were scanned on a 7T whole-body scanner (Siemens Healthcare, 32 channels head coil) to obtain high resolution quantitative T2* (TR/TE=3680/3.12+3.32*[1..6]ms, resolution=0.5x0.5x0.5mm3, acquisition time (TA): 40min) and T1 images (dual magnetization-prepared rapid gradient echo, MP2RAGE)7, TR/TE/TI=5000/2.93/[900 3200]ms, resolution=0.75x0.75x0.75mm3, TA=10min); and on a 3T whole-body scanner (Siemens Healthcare, MAGNETOM CONNECTOME, gradient strength=300mT/m, 64 channels head coil) to obtain measures of connectomics from diffusion-weighted images (DWI, TE/TR=57.0/8800ms, δ=12.9ms, Δ=21.8ms, 3 b-value shells: 1k/5k/10k s/mm2, number of directions: 64/128/128, resolution=1.5x1.5x1.5mm3 total TA=53 min).
Processing. Combining quantitative T1 and T2* to estimate a robust myelin map: Figure1A. For each subject, T1 and T2* were registered to individual cortical surfaces, sampled at the mid-cortical distance and registered to a common template surface (fsaverage). Cortical thickness was computed to correct both metrics for partial-volume-effect. Then a spatial Independent Component Analysis was used to extract the shared myelin related signal in T1 and T2* maps thus creating the Combined Myelin Estimation (CME), a metric more specific to myelin than T1 or T2* separately5,8. For this first processing part, 2 HC and 1 MS were removed due to strong subject motion or susceptibility artifacts. Areas of cortical demyelination: A general linear model, including age and gender as adjustment factors, was used to compare CME in MS patients vs. healthy controls in all Brodmann areas (BA). DWI processing: Figure1B. Images were corrected for gradient nonlinearity warping, Eddy-current distortions, and motion. DSI-Studio (http://dsi-studio.labsolver.org) was employed for whole-brain tractography with quantitative anisotropy9 threshold of 0.035 and maximum angle of 60°. Then the structural connectivity matrix of each participant was extracted using Brodmann Areas (BA) and MNI10 regions as ROIs. Graph theoretical and network-based statistic: The Brain Connectivity Toolbox (BCT, http://www.brain-connectivity-toolbox.net/) was used to investigate the structural connectomics organisation. The following graph-theory metrics were measured: clustering coefficients, global/local efficiency, strength, and modularity. BrainNet-viewer11 was used networks visualisation.
Results
Figure2A shows the myelin estimated maps (CME) averaged across HC and MS groups. We can visually observe a loss of myelin around the motor, visual and auditory cortices. In the whole cortex, CME was decreased, T1 and T2* were increased while no cortical thinning was observed in MS vs HC. Figure2B shows the averaged structural networks of HC and MS. In the whole graph, an increase of the modularity was observed (p<0.05), but no significant changes in global-efficiency.
Figure3 illustrates areas of significant demyelination as well as nodes (ROIs) presenting significantly altered topological measures in early MS vs HC. Interestingly, motor, premotor, frontal eye field and anterior cingulate cortices (BA4, BA6, BA8 & BA32-33) are affected simultaneously by cortical demyelination and connectomics alterations: namely, decrease in strength and local-efficiency (p<0.05). Moreover, a cortical demyelination in the visual cortex (BA17, 18, 19, non-significant) was observed along with significant alterations in connectomics, namely an increased clustering (p<0.05).
Discussion/Conclusion
Using CME, a multivariate statistical framework that combines quantitative T
1 and T
2* for increased specificity to myelin content, we could detect areas suggestive of cortical demyelination in a small cohort of early MS patients, even before the appearance of cortical thinning. Interestingly, these cortical areas where part of networks for which abnormal structural connectomics, by means of increased clustering, was demonstrated relative to HC. This suggests that early in the disease diminished structural integration between separate structural modules can occur along with cortical demyelination. Future studies will investigate whether there is a causal-effect relation between these two processes and whether the use of these metrics can improve correlation with clinical outcome, particularly cognition, compared to conventional MR metrics.
Acknowledgements
We would like to thank Qiuyun Fan (A.A. Martinos Center for Biomedical Imaging) and Benjamin De Leener (Polytechnique Montreal) for helpful discussions. We would like to mention that both last authors contributed equally to this work. This study was supported by the National Institute of Health [NIH R01NS078322-01-A1], the Canadian Institute of Health Research (CIHR FDN-143263), the Sensorimotor Rehabilitation Research Team (SMRRT), the Fonds de Recherche du Québec - Santé (FRQS 28826), the Fonds de Recherche du Québec - Nature et Technologies (FRQNT 2015-PR-182754), Quebec Bio-Imaging Network (QBIN) and the Natural Sciences and Engineering research Council of Canada (NSERC) and the Polytechnique MEDITIS program.References
1. Filippi M, van den Heuvel MP, Fornito A, et al. Assessment of system dysfunction in the brain through MRI-based connectomics. Lancet Neurol. 2013;12(12):1189–1199.
2. Shu N, Liu Y, Li K, et al. Diffusion tensor tractography reveals disrupted topological efficiency in white matter structural networks in multiple sclerosis. Cereb. Cortex 2011;21(11):2565–2577.
3. Geurts JJG, Bö L, Pouwels PJW, et al. Cortical lesions in multiple sclerosis: combined postmortem MR imaging and histopathology. AJNR Am. J. Neuroradiol. 2005;26(3):572–577.
4. Geurts JJG, Barkhof F. Grey matter pathology in multiple sclerosis. Lancet Neurol. 2008;7(9):841–851.
5. Mangeat G, Govindarajan ST, Mainero C, Cohen-Adad J. Multivariate combination of magnetization transfer, T2* and B0 orientation to study the myelo-architecture of the in vivo human cortex. Neuroimage 2015;119:89–102.
6. Li Y, Jewells V, Kim M, et al. Diffusion tensor imaging based network analysis detects alterations of neuroconnectivity in patients with clinically early relapsing-remitting multiple sclerosis. Hum. Brain Mapp. 2013;34(12):3376–3391.
7. Marques JP, Kober T, Krueger G, et al. MP2RAGE, a self bias-field corrected sequence for improved segmentation and T1-mapping at high field. Neuroimage 2010;49(2):1271–1281.
8. Mangeat G, Louapre C, Herranz E, et al. Multivariate combination of quantitative T2* and T1 at 7T MRI detects in vivo subpial demyelination in the early stages of MS. In: Mult Scler. 2015 p. 485.
9. Fang-Cheng Y, Van JW, Wen-Yih IT. Generalized Q-Sampling Imaging. IEEE Trans. Med. Imaging 2010;27:1415–1424.
10. Fonov V, Evans AC, Botteron K, et al. Unbiased average age-appropriate atlases for pediatric studies. Neuroimage 2011;54(1):313–327.
11. Xia M, Wang J, He Y. BrainNet Viewer: a network visualization tool for human brain connectomics. PLoS One 2013;8(7):e68910.