Longitudinal changes of structural and functional connectivity in multiple sclerosis in a natural history cohort and during an exercise trial
Final Report Abstract
Magnetic resonance imaging (MRI) allows investigating changes in the brain network topology by means of functional connectivity (resting state) and structural connectivity (diffusion weighted). Multiple Sclerosis (MS) can be understood as a network disease where recurrent inflammation and progressive neurodegeneration affect brain networks. Such changes might reflect the hierarchical organisation of neural networks. Aerobic exercises are considered to promote neuroprotective or neuroregenerative mechanisms in healthy individuals and neurological diseases including MS and might serve as a framework to study neuroplasticity of brain networks. Here, I aimed to investigate functional and structural connectivity in a natural history cohort of early MS (n=65, mean follow-up 4.2 years) and their changes in a randomised waiting group controlled exercise trial over three months (n=57). Healthy controls served for both cohorts as a reference (n=59 respectively n=30). For both data sets we applied the same methods to guarantee comparability: Functional networks were reconstructed based on wavelet correlations from the BOLD time series and mean fractional anisotropy of tracks derived from probabilistic tractography was used for structural networks. We observed a distinctive pattern of structural network alterations in patients with MS compared to controls: Already at beginning of the disease, hub or management regions of the brain show a pronounced loss of connectivity, which can be observed in controls within the age of 40 to 50. Moreover, a physiological adaption seems to take place as these structurally worse connected nodes show an increased functional connectivity. However, the adaption alone does not seem to enhance sufficiently neuroplasticity as the waiting group in the exercise trial did not show any changes from baseline. In contrast, we observed an increased structural and functional connectivity in hub regions in the exercise group. Summarized I could proof the feasibility of the adapted hub disruption index as an outcome of neurodegeneration and repair in a long-term natural history cohort and a randomized controlled trial. Moreover, the index was more sensitive than global graph metrics or standard clinical and MRI outcomes such as EDSS or T2-lesions in monitoring diseases progression in mildly disabled patients. Analysing functional and structural networks with regards to their topology is a promising outcome to investigate neurodegeneration and neuroplasticity in MS. In addition, we could add further evidence that exercising is currently the most promising option to enhance neuronal repair in MS.
