- Top of page
Summary: Purpose: Early animal studies of the therapeutic mechanisms of vagus nerve stimulation (VNS) suggested that seizure suppression requires maximal activation of small, unmyelinated vagal C fibers. However, effective therapeutic stimulation parameters appear to be subthreshold for these fibers in humans, and there are no clinical reports of the autonomic side effects that would be expected if these fibers were maximally activated. We report here that selective destruction of C fibers with capsaicin does not affect VNS-induced seizure suppression in rats.
Methods: Rats were pretreated with capsaicin or vehicle in three injections over a 2-day period. A cuff electrode was later implanted on the left cervical vagus nerve. Two days after surgery, VNS was given to half of the capsaicin- and vehicle-treated rats. The remaining rats were connected to the stimulator but did not receive VNS. Thirty seconds after VNS onset, seizures were induced by pentylenetetrazol (PTZ), and seizure severity was measured. Two days later, the reciprocal VNS treatment was given, and PTZ-induced seizure severity was again measured.
Results: VNS effectively reduced seizure severity in both capsaicin- and vehicle-treated rats as compared with their non-VNS baselines.
Conclusions: These results indicate that activation of vagal C fibers is not necessary for VNS-induced seizure suppression.
Vagus nerve stimulation (VNS) is a novel anticonvulsant therapy now gaining prominence for the treatment of refractory epilepsy. Compared with conventional epilepsy neurosurgery, this technique involves a lower-risk surgery and results in fewer complications (1). The surgery in humans involves the placement of spiral electrodes on the left cervical vagus nerve, with intermittent stimulation provided by a neurocybernetic prosthesis (NCP) (Cyberonics, Inc., Houston, TX, U.S.A.) implanted subcutaneously in the upper chest. Clinical trials have found the device to reduce the incidence of complex partial seizures in the majority of patients tested, with 20–40% of patients achieving a >50% reduction in seizure frequency (2). Stimulation parameters are adjusted to produce a minimum of side effects. Some side effects, such as a tingling sensation in the neck and temporary hoarseness during stimulation, do persist; however, significant gastrointestinal, cardiac, or respiratory effects are not apparent (2).
At the cervical level, the afferent component of the vagus nerve is a composite of myelinated A and B fibers, and unmyelinated C fibers, conveying sensory information from a number of organs, including the heart, lungs, liver, stomach, and intestines. The afferent C fibers are the most numerous, accounting for ∼65–80% of the fibers in the cervical vagal trunk (3). Early animal studies on the therapeutic mechanisms of VNS suggested that seizure suppression requires the activation of vagal C fibers (4). However, effective therapeutic stimulation parameters appear to be subthreshold for vagal C fibers in humans, and there are no clinical reports of autonomic side effects. These autonomic side effects would be expected if C fibers were maximally activated (5).
Identification of the vagal fiber types involved in VNS-induced seizure suppression is important for a number of reasons. C-fiber activation requires higher stimulation amplitudes and longer pulse durations than does activation of myelinated A and B fibers, thus shortening NCP battery life. If C-fiber activation is not necessary for seizure suppression, clinical stimulation parameters could be tailored to activate only the A and B fibers to prolong battery life and to reduce the need for surgical NCP replacement. Identification of the fiber types involved also would aid in the elucidation of the VNS therapeutic mechanisms and the development of rational cotherapies.
To determine whether C-fiber activation is necessary for VNS-induced seizure suppression, we examined the effects of capsaicin pretreatment on the ability of VNS to suppress pentylenetetrazol (PTZ) seizures in rats. Capsaicin is the active ingredient in many hot peppers and, when administered peripherally to rats in high doses, produces a long-lasting and selective excitotoxic lesion of afferent C fibers, especially when given neonatally (see ref. 6 for a review). It was predicted that selective destruction of C fibers with capsaicin would have no effect on VNS-induced seizure suppression.
- Top of page
This study demonstrates that selective destruction of capsaicin-sensitive C fibers does not affect VNS-induced seizure suppression in rats. Left cervical VNS was just as effective in capsaicin-treated rats as it was in vehicle-treated controls. These data lend support to the growing clinical consensus that vagal C fibers are not recruited in significant numbers with clinically relevant VNS parameters as evidenced by the types of side effects produced (12).
Systemic capsaicin is not selective to vagal afferent C fibers; other capsaicin-sensitive fibers, such as the splanchnic nerve also were undoubtedly lesioned. This would not have skewed the results, however, as VNS-induced seizure suppression is an event mediated by vagus nerve afferents (13). Baseline seizure severities also were measured after capsaicin administration, so the nonspecific capsaicin effects on other nerves containing C fibers were taken into account before VNS was given.
If capsaicin-sensitive C fibers are not necessary for VNS-induced seizure suppression, several alternative hypotheses may account for therapeutic VNS effects: (a) capsaicin-insensitive C fibers may be responsible for seizure suppression, (b) ablation of individual vagal fiber groups is not sufficient to prevent VNS-induced seizure suppression, or (c) vagal A and/or B fibers are responsible for seizure suppression.
The first hypothesis cannot be ruled out by the present results. A small population of vagal C fibers is not sensitive to the effects of capsaicin, such as those activated by gastric distention (12,14). One way to maximize C-fiber lesions is to administer capsaicin during the neonatal period in rats (6,12). This was done in the present study, and the results did not differ from those obtained in adults.
The second hypothesis assumes that a critical mass of fiber recruitment must be obtained to suppress seizures, regardless of the fiber type. That is, all vagal fibers may contribute to the therapeutic effects, and lesions of one fiber group are not sufficient to prevent seizure suppression. The authors are unaware of any technique to lesion specifically either capsaicin-insensitive C fibers, or A and B fibers, that would not disrupt capsaicin-sensitive C-fiber responses as well. Therefore, testing this alternative hypothesis may not be possible. However, the clinical evidence points to the third possibility: VNS-induced seizure suppression results from activation of vagal A and B fibers. Lack of clinical autonomic side effects and the presence of side effects mediated by A and B fibers (e.g., hoarseness) suggest that C fibers are not recruited in sufficient numbers to account for the therapeutic mechanisms of VNS.
We do not believe that our results contradict the early animal studies, which suggested that C fibers are necessary for VNS-induced seizure suppression (4); we differ only in our interpretation of the results. The data supporting C-fiber involvement were obtained mainly from anesthetized rats, and may not be relevant to VNS in awake rats. Compound action potentials recorded from the vagus nerve of their unanesthetized rats indicated that the half-maximal stimulation threshold of vagal C fibers at pulse durations used in the present study (0.5 ms) is >3 mA, 10 times higher than in anesthetized rats because of fluid shunting. We used a stimulation amplitude of only 1 mA in the same unanesthetized preparation, well below the C-fiber threshold, yet we also obtained significant seizure suppression.