Abstracts

Rationale: We have previously demonstrated a robust effect of electrodermal biofeedback in reducing seizure frequency in patients with drug resistant epilepsy (Nagai et al., 2004a, 2009, 2014). A randomized controlled trial showed more than half of patients in a biofeedback group reporting reduction in seizure frequency of 50% or more. The result was recently replicated in the independent group confirming the effect of electrodermal biofeedback in a similar manner (Micoulaud-Franchi, 2014). The establishment of EDA biofeedback treatment was based on a series of investigations into the relationship between peripheral and central arousal (Nagai 2004b, Nagai 2004c). Increasing skin conductance elicits a reduction in electroencephalographic signatures of cortical excitability (slow cortical potential) in patients with epilepsy (Nagai et al., 2009), an effect that can be observed with specific anti-epileptic drugs (Birbaumer et al, 2001). It was also identified that changes in the activity of ventromedial prefrontal cortex (VMPFC) /orbitofrontal cortex (OFC) during EDA biofeedback correlate inversely with skin conductance level (Nagai et al., 2004c). In the current study, we quantified how functional neural connectivity is altered following the EDA biofeedback treatment, to investigate whether there is a normalization of abnormal resting state network connectivity in patients with epilepsy. In particular, we were interested in the role of VMPFC/OFC within network modulation through EDA biofeedback.Methods: Eight patients with temporal lobe epilepsy went through a clinical trial of EDA biofeedback, undergoing EDA biofeedback training three times a week for four weeks (see Nagai 2004a for more detail). Scanning (1.5T fMRI) was performed at the first session and at the last session. Resting state fMRI was processed using SPM8. Seed-based analysis was conducted with VMPFC/OFC (ROI selected from the previous finding; Nagai et al., 2004c).Results: Biofeedback significantly reduced the patients’ seizure frequency after a month of training (p = 0.001). The reduction was observed in all patients with an average reduction of 39.6% (Fig.1). There were increases in functional connectivity between the seed VMPFC/OFC region and inferior temporal gyrus, anterior cingulate, putamen and precentral gyrus (Fig.2). Decreased connectivity was observed with superior temporal gyrus and angular gyrus.Conclusions: EDA biofeedback represents an accessible and side-effect free non-pharmacological treatment for patients with drug resistant epilepsy. We show a characteristic shift in resting state network organization that is consistent with a treatment-related enhancement connectivity between regions implicated in affective arousal and cortical control. Our neuroimaging study provides insight into potential mechanisms underlying the efficacy and validity of EDA biofeedback treatment for epilepsy.