Vagus nerve stimulation decreases hippocampal and prefrontal EEG power in freely moving rats: a biomarker for effective stimulation?

• ¹ Ghent University, Laboratory for Clinical and Experimental Neurophysiology, Neurobiology, and Neuropsychology, Belgium
• ² Ghent University, Department of Engineering, MEDISIP, Belgium
• ³ University of Amsterdam, Swammerdam Institute of Life Sciences, Department of Neurobiology, Netherlands
• ⁴ Vrije Universiteit Brussel, Center for Neurosciences, Department of Pharmaceutical Chemistry, Belgium

BACKGROUND: Vagus nerve stimulation (VNS) is a neuromodulatory therapy approved for the treatment of drug resistant epilepsy and depression. While more than 100.000 patients world-wide have been treated with VNS, parameters predicting effectively delivered stimulation are still lacking. Based on the knowledge that VNS causes wide spread changes in cortical and subcortical activity, the present study aimed to assess EEG as a biomarker for effective VNS. METHODS: Eight male Sprague Dawley rats (weighing 320-370g) underwent implantation of a cuff-electrode around the left vagus nerve and stereotactic implantation of bipolar depth electrodes in the left hippocampus and medial prefrontal cortex (mPFC). One month after surgery, effects of VNS on EEG were assessed both during day and night to control for effects of behavioral state. VNS was applied with a rapid cycle, 7 seconds ON/18 seconds OFF, 30 Hz, 1 mA , 250 μs pulse width. Ten seconds of EEG were recorded in each OFF period during a two hour stimulation period, preceded by a one hour baseline period and followed by a one hour post stimulation period. Two corresponding SHAM sessions, serving as control, were conducted without active stimulation. EEG spectral bands were defined as delta = 1-4 Hz, theta = 4-8Hz, alpha = 8-16 Hz, beta = 16-30 Hz, low-gamma = 30-50 Hz and high-gamma = 50-100 Hz. For statistical analysis, EEG power was normalized to baseline mean and subsequently averaged over 20-minute time periods. RESULTS: Seeing as no significant differences were observed between day and night conditions, only night data was considered for further analysis. Application of VNS led to a suppression of power both in the hippocampus and medial prefrontal cortex (Figure 1). Statistically significant suppression of hippocampal power was observed in all spectral bands compared to baseline and corresponding SHAM conditions. Similarly, VNS significantly suppressed mPFC theta, alpha, beta and high-gamma power. Onset of effects was immediate and outlasted the VNS period for at least 20 minutes. One rat, with high VNS electrode impedance, had no effects of VNS on EEG. CONCLUSIONS: VNS decreased absolute power of hippocampal and prefrontal EEG. EEG is easily obtainable and could constitute a feasible biomarker for effectively delivered VNS. Further, the obtained marker may be used to test and optimize stimulation paradigms.