Commentary: Integrating electrodermal biofeedback into pharmacologic treatment of grand mal seizures

摘要

Thirty percent of patients with epilepsy experience seizures despite optimal anticonvulsant drug treatment. Stress is frequently identified by patients with epilepsy as a precipitant of seizures (Spector et al., 2000 ; Ferlisi and Shorvon, 2014 ). Patients also often report using countermeasures to control the seizure onset. These are typically spontaneous and individualized such as increasing arousal by walking, breathing, standing, focusing attention, changing way of thinking, and more rarely inducing relaxation (Lee and No, 2005 ; Hether et al., 2013 ). In parallel, behavioral and psychological interventions, complementing conventional therapeutic methods for the management of epileptic seizures, have gained greater clinical attention over the past decade. Among these, Biofeedback (BFK) represents a noninvasive biobehavioral treatment that enables a patient to gain volitional control over a specific physiological process. BFK has already shown its value when applied to patients with epilepsy (Sterman and Friar, 1972 ; Rockstroh et al., 1993 ; Nagai et al., 2004a ; Nagai, 2011 ; Micoulaud-Franchi et al., 2014a , b ). Scrimali et al. ( 2015 ) have rightly pointed out the potential usefulness of electrodermal biofeedback in the management of refractory epilepsy. In a single case study, they report an effect of electrodermal activity (EDA) relaxation biofeedback in reducing seizures in a patient treated for 2 years. This case study supports the necessity to expand clinical armamentarium for treatment-resistant patients with few alternatives. However, in this commentary we would like to specify that contrary to what it is claimed by the authors, decreasing arousal, using relaxation biofeedback is not the method proposed by Nagai et al. ( 2004a ). On the contrary, these authors used a protocol based on increasing arousal. The EDA biofeedback treatment designed by Nagai et al. ( 2004a ) is aimed at increasing the tonic levels of peripheral sympathetic arousal rather than at decreasing EDA level. In addition, these authors (Scrimali et al., 2015 ) cite Nagai et al. ( 2004a ) when they claim that “Grand Mal” seizures are characterized by increased sympathetic arousal. However, there is no mention of such a relationship in the paper of Nagai et al. that rather highlights an inverse relationship between sympathetic arousal and an electroencephalographic index of cortical arousal that is linked with seizure activity. The motivation for believing that EDA relaxation biofeedback may be beneficial for patients with epilepsy comes is understandable given the relationship between stress and increased skin conductance level. It is known that stress and anxiety are commonly reported triggers of seizures (Fenwick, 1991 ; Antebi and Bird, 1993 ; Spector et al., 2001 ; Nakken et al., 2005 ; Petitmengin et al., 2006 ; Pinikahana and Dono, 2009 ). Thus, inducing relaxation in order to reduce the impact of stress has theoretical justification for patients with epileptic seizure triggered by stress. EDA reflects the action of sympathetic autonomic nerves on eccrine sweat glands and is a sensitive indicator of central (involuntary and voluntary induced) changes in arousal, associated with emotion, attention, and physical activity (Neumann and Blanton, 1970 ). In fact, EDA can be increased not only by stress, but also by preparation of and engagement with intellectual activity, during emotion regulation or even when experiencing happiness (Gross and Levenson, 1993 ; Reynaud et al., 2012 ). Moreover, regarding the relationship between autonomic physiology and seizures in epilepsy, a state of enhanced relaxation/low arousal can potentially worsen seizures through lowering the seizure threshold (Nagai et al., 2004b ). The theoretical rational behind the development of EDA biofeedback originated with a neurofeedback protocol focusing on Slow Cortical Potentials (SCP), an index of thalamocortical excitability. In SCP neurofeedback, patients are trained to reduce and reverse the negative amplitude of SCPs (Kotchoubey et al., 2001 ). This neurofeedback method has been shown to elicit a significant decrease in seizure frequency (Rockstroh et al., 1993 ) and can produce a sustained benefit that is apparent ten years after the end of treatment (Strehl et al., 2014 ). The amplitude of the Contingent Negative Variation (CNV) (a task-evoked event-related potential considered as a SCP and hence an index of cortical excitation) is greater during states of reduced autonomic arousal (Nagai et al., 2004b , 2009 ). Peripheral sympathetic activity (autonomic arousal), measured using EDA, can be increased either involuntarily by aversive auditory stimulation, or voluntarily by using biofeedback arousal. In both cases, the increase is associated with a reduction in the amplitude of the CNV. This observation indicated that the measures of peripheral sympathetic activity and thalamocortical excitation (as measured by the CNV) are inversely related (Nagai