Structural Neuroimaging to Generate Biomarkers of Response to Deep Brain Stimulation

Deep Brain Stimulation (DBS) is the use of continuous, typically high-frequency direct electrical current, delivered through neurosurgically implanted electrodes, aimed at altering activity within dysfunctional brain circuits. DBS is now an established therapy for several neurological conditions such as Parkinson’s Disease, dystonia, and tremor, and is emerging as a viable therapy in several others including depression and disorders of memory. However, the response to DBS is highly variable across individuals with the same disease, a fact often attributed to clinical factors such as age or illness severity, or to technical factors such as DBS electrode position. The central hypothesis underlying this thesis is that brain structure, as evaluated by structural magnetic resonance imaging (MRI), is also a key factor influencing clinical response to DBS therapy across various neurological and psychiatric diseases. The specific aims of the thesis project were to determine: 1) whether pre-DBS brain structure relates to, or predicts, eventual clinical response to DBS; and 2) whether ongoing DBS therapy produces detectable structural changes in the brain that may influence clinical response. In study #1, we demonstrate that in patients undergoing DBS of the subcallosal cingulate region (SCG) for depression, future response to DBS is predicted by pre-treatment variations in brain structure, especially volume of the SCG, thalamus, and the total volume of cortical grey matter. Studies #2 and #3 provide preliminary, first-in-human evidence, in two different patient populations (Parkinson’s Disease and Alzheimer’s Disease), that DBS is associated with structural neuroplasticity demonstrable on post-operative MRI scans, and may even result in the reversal of focal atrophy in selected patients. Study #4, however, illustrates the difficulties inherent in replicating these results on a larger scale because of methodological concerns related to the reliability and concordance of automated algorithms currently used to analyze structural neuroimaging data. Despite these concerns, this thesis illustrates that structural neuroimaging with MRI indeed has the potential to generate clinically useful biomarkers of response to DBS therapy, which may ultimately influence patient selection for DBS, guide DBS programming, and inform prognosis after DBS implantation.Ph.D