Impact of Cardiac-Based Vagus Nerve Stimulation Closed-Loop Stimulation on the Seizure Outcome of Patients With Generalized Epilepsy: A Prospective, Individual-Control Study.

OBJECTIVES We designed a prospective, individual-controlled study to evaluate the effect of cardiac-based VNS (cbVNS) in a cohort of patients with generalized epilepsy (GE). MATERIALS AND METHODS Twenty patients were included. They were followed up for six months under regular VNS (rVNS) and subsequently for six months during cbVNS. Stimulation parameters were 500 μsec, 30 Hz, and up to 2.5 mA. Seizure frequency was documented after two, four, and six months during the rVNS and cbVNS phases. Patients with at least 50% seizure frequency reduction were considered responders. The total and relative amount of stimulation cycles generated by both rVNS and cbVNS activation were documented. Findings during rVNS were compared to baseline and cbVNS data were compared to those during rVNS. RESULTS There was a significant decrease in mean seizure frequency (61% [95% CI, 48-74]; p < 0.001) during the rVNS phase compared to baseline. There was no additional significant (16% [95% CI, 4-35]; p = 0.097) mean seizure frequency reduction during cbVNS compared to the rVNS phase. Fifteen patients (75%) were considered responders after rVNS. Four patients (20%) were considered responders after six months of cbVNS. During the cbVNS phase, the mean total number of cycles/day was 346, 354, and 333 for months two, four, and six, respectively; the cycles generated by rVNS were 142, 138, and 146 for months two, four, and six, respectively; and cycles generated by cbVNS were 204, 215, and 186 for months two, four, and six, respectively. There was no relationship between the mean total number of cycles (-6[95% CI, -85 to 72]; p = 0.431), the mean number of auto-stimulation cycles (27[95% CI,-112 to 166]; p = 0.139), the mean number of regular cycles (-33[95% CI,-123 to 57]; p = 0.122), or the mean percentage of auto-stimulation cycles (13[95% CI,19- 45]; p = 0.109) and outcome during the cbVNS phase. Eight patients showed some decrease in seizure frequency during cbVNS. CONCLUSIONS rVNS was effective in reducing seizure frequency in patients with generalized epilepsy, but activation of the cbVNS feature did not add significantly to rVNS efficacy. On the other hand, although not statistically significant, 40% of the patients showed some reduction in seizure frequency, which might prove useful at an individual level.