Analysis of RR variability in drug-resistant epilepsy patients chronically treated with vagus nerve stimulation.

Vagus nerve stimulation (VNS) has been suggested as an adjunctive treatment for drug-resistant epilepsy when surgery is inadvisable. The overall safety profile of VNS seems to be favorable as only minor adverse effects have been described. The purpose of this study was to determine if cardiac vagal tone is eventually modified by short- and long-term VNS. The effects of short- and long-term VNS were evaluated in seven subjects with intractable epilepsy. Autonomic cardiac function has been carried out by means of a 24-h analysis of RR variability at baseline (t(0)), 1 month (t(1), short-term VNS) and 36 months after VNS initiation (t(2), long-term VNS). Frequency- and time-domain parameters were calculated. Periodic cardiological and neurological evaluations were performed.Clinically relevant cardiac effects were not observed throughout the study. Despite the limited number of patients and the variety of data among them, for all the patients, a common trend towards a nocturnal decrease in the high-frequency (HF) component of the spectrum was observed after long-term VNS (mean+/-S.D.: 40+/-18 normalized units (nu) at t(0), 38+/-17 nu at t(1), 18+/-10 nu at t(2); p<0.05 of t(2) vs. either t(0) or t(1)). The day-to-night changes in the power of low-frequency (LF) and HF components were significantly blunted after long-term VNS (LF day-to-night change: +16+/-13 nu at t(0) and +15+/-8 nu at t(1) vs. +3+/-13 nu at t(2), p<0.02; HF day-to-night change: -18+/-13 nu at t(0) and -13+/-11 nu at t(1) vs. +3+/-12 nu at t(2), p<0.003). No significant changes were observed with regard to the time-domain parameters of the heart rate variability. Throughout the neurological follow-up, one subject became seizure-free, three experienced a seizure reduction of >50%, two patients of <50% and one had no changes in his seizure frequency. Our findings suggest that long-term VNS might slightly affect cardiac autonomic function with a reduction of the HF component of the spectrum during night and a flattening of sympathovagal circadian changes, not inducing, however, clinically relevant cardiac side effects.