Abstracts

Rationale: TNS has shown promise as a treatment for epilepsy and depression in phase I clinical trials (1). TNS is hypothesized to have a mechanism of action similar to vagus nerve stimulation (VNS). To test this hypothesis and to better understand the potential mechanism of action of TNS, brain [O-15] water PET scans were obtained during TNS. In functional imaging studies during VNS, decreases in activation of medial temporal regions are consistently seen, as are increases in activation of the bilateral anterior cingulate gyri and dorsomedial and dorsolateral prefrontal cortices (2-5). Methods: Five subjects with major depressive disorder enrolled in a phase I clinical trial of TNS as a depression therapy underwent six PET scans prior to starting the trial. Three scans were acquired while receiving 60 seconds of TNS at maximum tolerated intensity. Three were acquired with TNS turned on but current set to zero. For each scan, TNS was initiated at the time of [O-15] water IV injection. Clinical response to TNS was assessed by the Hamilton Depression Rating Scale (HAM-D). Regional brain activity was analyzed by statistical parametric mapping (spm). Correlations were examined between change in HAM-D (score at week 8 minus score at baseline) and PET data analyzed using the standardized volume of interest (sVOI) method. Results: The most significant activations with TNS by spm analyses occurred in the left inferior frontal gyrus, right medial and middle frontal gyrus, bilateral parietotemporal cortex and bilateral anterior cingulate gyri. The most significant deactivations with TNS by spm analyses occurred in the left parahippocampal gyrus, right sensorimotor, right superior parietal area, bilateral temporo-occipital cortex and bilateral visual cortex. The right superior parietal region showed the strongest negative correlation, with stronger deactivation being associated with less significant improvement in HAM-D. Conversely, stronger deactivation in the left caudate was associated with better clinical response (r=0.93, p=0.023). The patient with the least clinical response (HAM-D change of -4) showed activation in this left caudate area. Conclusions: TNS activations and deactivations occur in regions important for epilepsy and depression. These activations and deactivations show some overlap with findings in VNS functional neuroimaging studies. This suggests potential similarities in mechanism of action between TNS and VNS. The correlation of PET data with clinical response may enable prediction of clinical response early in the course of treatment. REFERENCES: 1) DeGiorgio CM et al. Neurology 2009;72:936-38. 2) Chae J-H et al. J Psychiatr Res 2003;37:443-455. 3) Kraus T et al. J Neural Transm 2007;114:1485-93. 4) Zobel A et al. Psychiatry Res 2005;139:165-79. 5) Conway CR et al. Psychiatry Res 2006;146:179-84.