Erik Middlebrooks

Parkinson disease is a relatively common and debilitating neurologic disorder. One of the most widely used indications for deep brain stimulation (DBS) worldwide has been in Parkinson disease, in which it has been shown to not only be highly effective in symptom control, but also allow the reduction of other medications that have debilitating side effect profiles. While there has been continued debate on the optimal target, primarily between the subthalamic nucleus and globus pallidus internus (GPi), well-controlled comparison trials have shown no substantial difference in patients’ quality of life between the 2 targets. It turns out that each target may have a slightly different benefit profile and both are likely important targets based on the patient’s phenotype; however, interest in GPi DBS has declined worldwide. The decline is in part driven by the added difficulty of targeting the GPi. The GPi is a relatively large nucleus on the scale of DBS and has no clearly demarcated internal boundaries between functional zones. Historically, this led to some difficulty with optimal targeting in programming, often necessitating the use of higher voltages and reduction in battery life.

In this study, we leverage connectomics in an attempt to define a more specific functional zone within the GPi to enable greater precision and targeting of the most effective subsection within the GPi nucleus. To this end, we employed MR tractography to designate voxels to specific networks based on greatest connectivity, which were then correlated with optimal stimulation settings. Unsurprisingly, the voxels with greatest connectivity to the primary motor network accounted for the most significant improvement in the Unified Parkinson Disease Rating Scale. This area corresponded to the more ventral portion of the posterior GPi, which corroborates previous evidence on the suspected area of maximal tremor benefit.

In summary, this study was the first of its kind to apply such a connectomic analysis to GPi DBS and establish the potential utility of this technique in optimizing GPi DBS targeting and programming. More recent advances in MR tractography, including MRI systems with high gradient performance, multiband acceleration allowing improved sampling of q-space and higher spatial resolution, as well as the increasing advancement of ultra-high-field MRI, stand to significantly benefit the accuracy and applicability of this approach. Since this publication, we have implemented such advances in connectivity analyses into our clinical arsenal, including routine 7T MRI in presurgical DBS cases. These improved techniques offer the potential to more accurately model small, subcortical connections and refine models of pallidal connectivity, creating more accurate biomarkers for DBS.

Read this article at AJNR.org …