Abstracts

Rationale: Preoperative thalamic targeting methods have historically relied on indirect targeting mainly because individual thalamic nuclei are not well resolved with traditional neuroimaging methods. Standard target points can be identified in reference to universally applied coordinates to the anterior-posterior commissure line by using atlas-based methods or stereotaxy-based approaches. Importantly, these preoperative techniques do not fully account for variations in functional anatomy between patients, or account for thalamic atrophy that occurs in patients with epilepsy. These interindividual variations, in part, contribute to the reported variability in patient outcomes with thalamic deep brain stimulation (DBS) for epilepsy. This study aims to address the variability noted between traditional indirect and direct targeting methods for the anterior nucleus of the thalamus (ANT) in patients whom were ultimately implanted with ANT DBS for epilepsy. Methods: As part of their pre-operative assessment, patients underwent brain MRI. Patients were preferentially imaged at 3 Tesla unless existing contraindications (e.g. VNS, pacemaker, etc.), in which case scans were performed at 1.5 Tesla. Imaging sequences utilized for this study consisted of non-contrast high-resolution T1-weighted MP-RAGE and high-resolution fast gray matter acquisition T1 inversion recovery (FGATIR), as previously reported.1 Thalamic volumes were calculated for each patient from the MP-RAGE images using automated methods, as implemented in FreeSurfer software (http://surfer.nmr.mgh.harvard.edu). Degree of thalamic volume loss was assessed by comparison of thalamic volumes to matched healthy controls2 to calculate a z-score. An asymmetry index between the left and right thalamus was also calculated as the ration of the left thalamic volume and the right thalamic volume. Indirect targeting coordinates were 5.5 mm lateral to the AC-PC line, 1 mm anterior to the midcommissural point, and 11 mm superior to the AC-PC line. The direct target was the posterior superior aspect of the insertion of the mammillothalamic tract (MMT) as visualized on FGATIR imaging.Statistical analysis was performed using Prism v. 8.1.1 (GraphPad Software, San Diego, CA, USA). Continuous variables were expressed as mean and standard deviation. Spearman correlation was utilized to assess the relationship between thalamic volume and the difference of the direct and indirect target coordinates in each axis, as well as the difference in Euclidean distance. Mann-Whitney test was used to compare differences between direct and indirect targets in each of three principal axes. A p-value of <.05 was considered statistically significant.  Results: Fifteen consecutive patients with 30 leads were analyzed for this study. There was a significant variation between the direct and indirect targets in the y-axis and z-axis on both sides. On the left, the direct target was located at a mean of y = 2 ± 1.3 mm and z = 9.3 ± 1.8 mm (both p=0.02). On the right, the direct target was located at y = 2.9 ± 1.8 and z = 9.2 ± 2 mm (both p=<0.0003). There was no significant difference in the x-coordinate on either side (-5.7 ± 0.9 on the left and 5.8 ± 1.1 on the right; p>0.5). Additionally, there was a correlation between thalamic volume and difference between direct and indirect targets in both the y and z axis (Figure 1, 2). Conclusions: Our data suggest that there is a significant difference in direct and indirect targeting in the anterior-posterior and rostro-caudal axes when targeting the ANT for DBS for epilepsy. Variation between targeting methods is also significantly correlated with thalamic volume. Such dependence of target location on thalamic volume highlights the potential value of direct targeting in ANT DBS since thalamic atrophy is a common finding in epilepsy. This study further supports the role of direct targeting of the ANT when implanting DBS for medically refractory epilepsy. Funding: No funding