Longitudinal RS fMRI & DTI to monitor recovery after stroke

Annually, 15 million people worldwide suffer from stroke, leaving about one third permanently disabled. The majority of patients (around 80%) is afflicted with motor impairments. Rehabilitation of motor impairments aims at recovering functions by promoting neuroplasticity and supporting cortical reorganization. Findings so far suggest that the combination of clinical measures and neuroimaging tools (e.g. diffusion tensor imaging (DTI), functional MRI) bears great potential to monitor functional mechanisms of recovery and to improve prediction of recovery. This would have implications for the choice and optimal timing of interventions and patient stratification based on most likely outcome. However, results obtained by task-related fMRI are highly dependent on patient effort and adherence during the performance of specific tasks and only reflect changes in particular neuronal systems (e.g. motor network). In contrast, resting state-fMRI (RS fMRI) is a newly evolving method, allowing identification of functional networks during rest based on coherent patterns of spontaneous fluctuations in the BOLD signal. RS-fMRI provides a great asset to explore how functional connectivity of the lesioned area and distant brain regions might be related to impaired motor behavior, obviating many limitations of task- related fMRI. Currently, the majority of studies investigated structural or functional reorganization in chronic stroke patients only in a cross-sectional manner, not taking dynamic changes into account. Hence, longitudinal and multimodal examination of these changes is essential. This study thus seeks to evaluate mechanisms of motor recovery by repeated functional MRI (task- related & resting state) and DTI in a minimum of 32 acute stroke patients (1-2 weeks after stroke and after 3 and 6 months). To control for reproducibility and stability, 14 healthy controls will be investigated. Our goal is to achieve better understanding of the complex basic mechanisms of structural and functional recovery (i.e. rearrangement, compensation, remapping, reorganization or reactivation) over the first six months after a stroke, and to assess the incremental value of RS data for prediction and observation of individual patient trends over and above clinical measures and task- related fMRI. We expect this to provide useful information on mechanisms of reorganization after stroke.

Annually, 15 million people worldwide suffer from stroke, leaving about one third permanently disabled. Motor rehabilitation aims at recovering functions by supporting cortical reorganization and promoting neuroplasticity. The combination of clinical measures (e.g. stroke severity scores) and neuroimaging tools (e.g. diffusion tensor imaging (DTI), functional MRI (fMRI)) bears great potential to monitor functional mechanisms of recovery and to improve prediction of recovery. Currently, the majority of studies are cross-sectional and investigated structural and/or functional reorganization in chronic stroke patients (weeks to months post-stroke), not taking early dynamic changes into account. Hence, longitudinal and multimodal examination of these changes is essential. The brains capacity for reorganization is highest within the first weeks post-stroke. We therefore evaluated mechanisms of motor recovery by repeated MRI (DTI, task-related & resting state) in 43 acute stroke patients (24 hours after stroke and after 3 months; mean age= 66 years, 40% female). Resting state-fMRI allows identifying functional networks during rest based on coherent patterns of spontaneous fluctuations in the BOLD signal. To control for reproducibility and stability, 15 healthy controls were investigated. We found that white matter integrity decreased within three months in stroke patients and these decreases were associated with worse motor outcome. Furthermore, prediction of motor function improved by 23%, when adding information of frontal white matter integrity in addition to baseline motor function and lesion size. In addition, functional changes within the motor and default-mode network were observed in patients, which were related to motor function. The clinical importance of resting-state fMRI was underlined by the high applicability in acute patients (100%), as opposed to only 60% of patients which were able to participate in task-related fMRI. An incremental improvement of the above mentioned prediction model by simple mean network scores, could not be observed. However, further regional analyses are currently running. This study adds to our knowledge about the dynamic structural and functional brain mechanisms associated with motor recovery after stroke within the first three months.