Pallidal stimulation for Holmes tremor: clinical outcomes and single-unit recordings in 4 cases

Abstract

OBJECT Holmes tremor (HT) is characterized by irregular, low-frequency (< 4.5 Hz) tremor occurring at rest, with posture, and with certain actions, often affecting proximal muscles. Previous reports have tended to highlight the use of thalamic deep brain stimulation (DBS) in cases of medication-refractory HT. In this study, the authors report the clinical outcome and analysis of single-unit recordings in patients with medication-refractory HT treated with globus pallidus internus (GPi) DBS. METHODS The authors retrospectively reviewed the medical charts of 4 patients treated with pallidal DBS for medication-refractory HT at the University of California, San Francisco, and San Francisco Veterans Affairs Medical Center. Clinical outcomes were measured at baseline and after surgery using an abbreviated motor-severity Fahn-Tolosa-Marin (FTM) tremor rating scale. Intraoperative microelectrode recordings were performed with patients in the awake state. The neurophysiological characteristics identified in HT were then also compared with characteristics previously described in Parkinson's disease (PD) studied at the authors' institution. RESULTS The mean percentage improvement in tremor motor severity was 78.87% (range 59.9%–94.4%) as measured using the FTM tremor rating scale, with an average length of follow-up of 33.75 months (range 18–52 months). Twenty-eight GPi neurons were recorded intraoperatively in the resting state and 13 of these were also recorded during contralateral voluntary arm movement. The mean firing rate at rest in HT was 56.2 ± 28.5 Hz, and 63.5 ± 19.4 Hz with action, much lower than the GPi recordings in PD. GPi unit oscillations of 2–8 Hz were prominent in both patients with HT and those with PD, but in HT, unlike PD, these oscillations were not suppressed by voluntary movement. CONCLUSIONS The efficacy of GPi DBS exceeded that reported in prior studies of ventrolateral thalamus DBS and suggest GPi may be a better target for treating HT. These clinical and neurophysiological findings help illuminate evolving models of HT and highlight the importance of cerebellar–basal ganglia interactions.