PROJECT SUMMARY: Using brain lesions and deep brain stimulation to identify an epilepsy circuit Focal epilepsy is common in patients with brain lesions such as stroke, trauma, and tumors. Why some patients with brain lesions develop epilepsy while others do not is unknown. Deep brain stimulation (DBS) offers new therapeutic promise for patients with focal epilepsy, but seizure freedom is rare. Why some patients improve after DBS while others do not is also unknown. Current dogma focuses on the lesion location or DBS site alone. We hypothesize that connectivity of lesion locations and DBS sites to remote nodes distant from the lesion or DBS sites itself can explain variance in lesional epilepsy and DBS response. New techniques developed by the Fox lab (PI) can identify remote nodes connected to lesions causing and DBS sites treating brain diseases. These techniques combine the location of the lesion or stimulation site with a normative atlas of human brain connectivity termed the connectome. As such, they do not require connectivity data from the patients themselves and can be applied to almost any clinical lesion or DBS dataset. This approach has successfully mapped circuits involved in parkinsonism, amnesia, depression, and over 20 other brain diseases. Moreover, identifying circuit nodes connected to both lesions causing and DBS sites treating the disease can identify new or improved therapeutic targets. Our preliminary data in ischemic stroke suggests there is a common brain circuit involved in epilepsy. Lesion locations that cause epilepsy (n = 76) are more connected to the cerebellum and basal ganglia (‘Ce-BG’) compared to control lesions (n = 625). Connectivity of anterior thalamic DBS sites to these same circuit nodes is correlated with seizure reduction after DBS (n=30). While epilepsy is often considered a cortical disease, these subcortical Ce-BG nodes have previously been implicated in the modulation of seizures and cortical excitability. Although promising, further work is needed to determine if these results generalize across lesion types (Aim 1, n = 2,700), prospectively predict epilepsy risk (Aim 2, n = 6,000), and correlate with DBS response (Aim 3, n = 198). For each aim, we will focus on our Ce-BG nodes as an a priori hypothesis, but also perform data-driven (unbiased whole-brain) analyses to minimize risk of false positives or negatives. We will perform all analyses using a functional connectome (primary outcome) and structural connectome. Completion of these aims will determine whether connectivity of lesion locations and DBS sites to a common set of subcortical nodes can explain variance in lesional epilepsy and DBS response. Identifying this circuit could identify patients at high risk for epilepsy, guide DBS programming, and serve as a target for future brain stimulation trials. Collectively, these results could facilitate a shift in focus from the seizure-onset zone itself, which differs in every patient, to a brain circ...