Alternating cycles of vagal nerve stimulation (VNS) and rest periods are used to treat intractable epilepsy. The left vagus carrying mostly afferent nerves is used and stimulation of depressor nerve afferents from baroreceptors is avoided by placing electrodes distal to the separation of the inferior cervical cardiac branch from the main nerve trunk. Plasticity of the central nervous system to sensory inputs is well known (Mendell et al. 2001) but the effects of activating a large number of brainstem afferents on normal function is not well studied. We measured a non-invasive index of the output of the nucleus ambiguus (NA) continuously during VNS and in the periods immediately following the stimuli.

Six patients (3 males) on VNS treatment were enrolled in the study after informed consent and approval of the local Ethics Committee. The VNS system consists of a stimulator (NeuroCybernetic Prosthesis Cyberonics Inc., Texas, USA) placed in the left chest and an electrode wound around the cervical left vagus nerve and tunnelled under the skin to the stimulator. The NeuroScope^TM system (MediFit Diagnostics Ltd, London) measured a non-invasive index of cardiac vagal tone (CVT) continuously (Julu et al. 1999). Constant current pulses of 500 µs duration (250 µs in two patients) were applied at 30 Hz in three cycles of 60 s stimulation and 66 s rest in supine position. Stimulus strengths were set relative to the patient’s therapeutic level (Ir) (in mA) as Ir – 0.75, Ir – 0.5, Ir – 0.25 and Ir. Mean CVT within 60 s of stimulation and 66 s of rest were calculated and the averages of three cycles were used for respective stimulus strengths.

The CVT increased in steady ramps, ended by habituation to the stimuli. Abrupt end of pulses provoked rebound increases in CVT from the habituated levels. Baseline CVT was (means ± S.E.M., units of a linear vagal scale) (Julu, 1992) 5.5 ± 1.2 and it increased to a significant level of 12 ± 6.5 at Ir – 0.25 mA (P < 0.03, Wilcoxon signed rank test). The CVT during the intervening rest periods also increased to a significant level of 6.4 ± 1.2 at Ir – 0.25 mA (P < 0.02). There were no significant differences in the mean CVT measured during stimulation compared with the intervening rest periods for all stimulus strengths except Ir.

The results show evidence of a sustained increase in CVT during VNS and immediately after the stimuli, implying brainstem plasticity to electrical stimulation. There were signs of habituation of NA to current pulses and rebound activation when the pulses stopped. These were in spite of excluding baroreceptor afferents from the depressor nerve.

Julu, P.O.O. (1992). J. Auton. Pharmacol. 12, 109-115.

Julu, P.O.O., Kerr, A.M., Hansen, S., Witt-Engerstrom, I., Al-Rawas, S., Engerstrom, L., Apartopoulos, F. & Jamal, G.A. (1999). J. Physiol. 523.P, 254P.

Mendell, L.M., Munson, J.B. & Arvavian, V.L. (2001). J. Physiol. 533, 91-97. abstract