528. Alterations of the event-related potential parameter P300 in antidepressant vagus nerve stimulation: preliminary results of a longitudinal prospective study

Abstract

Abstract Background Major depressive disorder (MMD) is one of the leading causes of global disease burden worldwide. Around 40 percent of patients suffering from MDD are considered treatment resistant and about 10% have chronic depression who do not achieve remission for years. Vagus nerve stimulation (VNS) is an invasive neuromodulatory treatment to manage these chronic forms of depression. Around 30 % of patients respond to antidepressant VNS. For optimal patient selection, it would be crucial to stratify responders based on predictive markers. Electroencephalography (EEG) and event-related potentials (ERP), especially the P300 component is associated with selective attention processing whereas decreases in P300 amplitude and increases in P300 latency were shown in major depressive disorder (MMD). Furthermore, a previous study demonstrated an enhanced amplitude in patients with MDD who respond to antidepressant VNS. Aims & Objectives The aim of this ongoing prospective study is to assess alterations in P300 amplitude and latency in patients that receive VNS for treatment-resistant depression. Method In this longitudinal prospective study, EEG measurements were conducted first at baseline and subsequently at day 1, day 30 and 180 after VNS stimulation onset. To capture ERPs, an auditory oddball paradigm was implemented, using a 64-channel EEG cap. To analyze preliminary data 26 measurements of 7 patients were included. Data analysis was done via the MATLAB toolbox EEGLAB. 2 datasets had to be excluded because of poor recording quality. After importing the raw EEG data, it was re-referenced to average reference. Subsequently, all datasets were down sampled to 256 Hz and bandpass filtered from 0.1 to 50 Hz using the basic FIR filter. By using clean rawdata and ASR function as well as independent component analysis, the data was cleaned and artifacts like eye-blinks and muscle artifacts were detect and removed. P300 amplitude and latency from Pz channel was extracted using the ERPLAB plugin. For statistical analysis, a linear mixed model was applied via the MATLAB Statistics and Machine Learning toolbox. Results Mean latency of P300 at baseline was 377.4 ms (± 32.5). No significant changes were found at day 1 (+30.8ms, p=0.47) and day 30 (+38.9 ms, p=0.35). However, at day 180 there was a significant decrease in latency compared to baseline (-190.0 ms, p=0.0025, 95% CI: 302.5 to -77.5 ms). Mean P300 amplitude at baseline was 2.42 μV (±0.68). There were no significant changes in amplitude at day 1 (+0.09 μV, p=0.92), day 30 (+1.23 μV, p=0.16) or day 180 (+0.51 μV, p=0.64). Due to the low N, we did not investigate the relationship to antidepressant response yet. Discussion & Conclusions These preliminary results suggest a possible reduction of P300 latency after 180 days of antidepressant VNS compared to baseline. Regarding amplitude, no significant effect was found, although a slight increase at day 30 was assessed. A limitation of this study is the small sample size and missing data points because of poor recording quality. Therefore, the results should be interpreted as preliminary and explorative. Analyzing data of a larger patient population with an additional measurement timepoints after one year is part of the ongoing study.