Although vagal nerve stimulation (VNS) is a common alternative add-on treatment for epilepsy, the effects of the possible combination of VNS programmable options other than those established in antiepileptic protocols (eg, current intensity, on-off operating time) have not been fully investigated. This study reports the effect of low-intensity, low cycling VNS on a subject with persistent cerebellar tremor due to multiple sclerosis (MS). Case report. The patient had sudden left hemiparesis, diplopia, and ataxia at age 17, followed by ataxia and postural and intentional cerebellar tremor, and was diagnosed with MS according to the Poser criteria. In subsequent years he was treated with interferon-1b and 1a; in addition, 10 cycles of immunoglobulins (IVIg) were administered. The Expanded Disability Status Scale (EDSS) 1 was 8/10 at the beginning of treatment with IVIg and 7/10 1 year later. However, there was further worsening of cerebellar tremor, despite treatment with primidone, gabapentin, and carbamazepine, singly or in combination. Since variations in cerebellar tremor are relevant in the present case, in addition to EDSS, this disability was scored according to the Klockgether Rating Scale (KRS), 2 which is used in studies of primary cerebellar diseases. This scale is based on a five-point rating system evaluating dysdiadochokinesia, intention tremor, dysarthria, upper and lower limb ataxia, gait, and stance. In ad hoc evaluation of this case report, we added assessment of spontaneous head tremor. Mild disability scored 1 point, maximum disability 5 points (total impairment: 40 points). According to this scale, our subject scored 35/40 on baseline evaluations (table).

The deterioration of cerebellar tremor despite treatment prompted use of VNS. Accordingly, VNS was implanted at the end of 2002 following patient and parental consent and approval by the Ethics Committee. VNS operating parameters conformed to high stimulation criteria, the default setting for the treatment of epilepsy. Telemetric control of VNS and motor assessment were monitored separately, as the rehabilitation team was blinded for VNS procedures. For the following eight months, the patient was assessed using both the EDSS and KRS. One year after VNS implant and due to the modest results obtained (KRS: 32/40), operating settings were changed under the hypothesis that low cycles are involved in the pathophysiology of experimental cerebellar tremor. 3