Neurophysiologic Characteristics of the Anterior Nucleus of the Thalamus during Deep Brain Stimulation Surgery for Epilepsy

Abstract

<b><i>Introduction:</i></b> Anterior nucleus of the thalamus (ANT) deep brain stimulation (DBS) is an increasingly promising treatment option for refractory epilepsy. Optimal therapeutic benefit has been associated with stimulation at the junction of ANT and the mammillothalamic tract (mtt), but electrophysiologic markers of this target are lacking. The present study examined microelectrode recordings (MER) during DBS to identify unique electrophysiologic characteristics of ANT and the ANT-mtt junction. <b><i>Methods:</i></b> Ten patients with medically refractory epilepsy underwent MER during ANT-DBS implantation under general anesthesia. MER locations were determined based on coregistration of preoperative MRI, postoperative CT, and a stereotactic atlas of the thalamus (Morel atlas). Several neurophysiological parameters including single unit spiking rate, bursting properties, theta and alpha power and cerebrospinal fluid (CSF)-normalized root mean square (NRMS) of multiunit activity were characterized at recording depths and compared to anatomic boundaries. <b><i>Results:</i></b> From sixteen hemispheres, 485 recordings locations were collected from a mean of 30.3 (15.64 ± 5.0 mm) recording spans. Three-hundred and ninety-four of these recording locations were utilized further for analysis of spiking and bursting rates, after excluding recordings that were more than 8 mm above the putative ventral ANT border. The ANT region exhibited discernible features including: (1) mean spiking rate (7.52 Hz ± 6.9 Hz; one-way analysis of variance test, <i>p</i> = 0.014 when compared to mediodorsal nucleus of the thalamus [MD], mtt, and CSF), (2) the presence of bursting activity with 40% of ANT locations (<i>N</i> = 59) exhibited bursting versus 24% the mtt (χ²; <i>p</i> &lt; 0.001), and 32% in the MD (<i>p</i> = 0.38), (3) CSF-NRMS, a proxy for neuronal density, exhibited well demarcated changes near the entry and exit of ANT (linear regression, <i>R</i> = −0.33, <i>p</i> &lt; 0.001). Finally, in the ANT, both theta (4–8 Hz) and alpha band power (9–12 Hz) were negatively correlated with distance to the ventral ANT border (linear regression, <i>p</i> &lt; 0.001 for both). The proportion of recordings with spiking and bursting activity was consistently highest 0–2 mm above the ventral ANT border with the mtt. <b><i>Conclusion:</i></b> We observed several electrophysiological markers demarcating the ANT superior and inferior borders including multiple single cell and local field potential features. A local maximum in neural activity just above the ANT-mtt junction was consistent with the previously described optimal target for seizure reduction. These features may be useful for successful targeting of ANT-DBS for epilepsy.