High-fidelity transmission of high-frequency burst stimuli from peripheral nerve to thalamic nuclei in children with dystonia

Abstract

High-frequency peripheral nerve stimulation has emerged as a noninvasive alternative to thalamic deep brain stimulation for some patients with essential tremor. It is not known whether such techniques might be effective for movement disorders in children, nor is the mechanism and transmission of the peripheral stimuli to central brain structures understood. This study was designed to investigate the fidelity of transmission from peripheral nerves to thalamic nuclei in children with dystonia undergoing deep brain stimulation surgery. The ventralis intermediate (VIM) thalamus nuclei showed a robust evoked response to peripheral high-frequency burst stimulation, with a greatest response magnitude to intra-burst frequencies between 50 and 100 Hz, and reliable but smaller responses up to 170 Hz. The earliest response occurred at 12-15 ms following stimulation onset, suggesting rapid high-fidelity transmission between peripheral nerve and thalamic nuclei. A high-bandwidth, low-latency transmission path from peripheral nerve to VIM thalamus is consistent with the importance of rapid and accurate sensory information for the control of coordination and movement via the cerebello-thalamo-cortical pathway. Our results suggest the possibility of non-invasive modulation of thalamic activity in children with dystonia, and therefore the possibility that a subset of children could have beneficial clinical response without the need for invasive deep brain stimulation.

Conflict of interest statement

The authors declare no conflict of interest. T.D.S serves as a paid consultant and owns stock in CALA HEALTH Inc., a company that develops and manufactures noninvasive electrical stimulation devices for treatment of movement disorders in adults.

Figures

Figure 1

3D rendering of group based DBS electrodes in the bilateral ventral intermediate nuclei. Axial view from normalized scans into MNI space. The VIM boundaries were defined by the DISTAL atlas. Each pair of DBS electrodes corresponds to a patient and are represented by a different color for each patient. A total of 20 DBS electrodes in 10 dystonic patients are represented through DSI Studio, V3 (http://dsi-studio.labsolver.org).

Figure 2

Average thalamic EPs in responses to contralateral single-pulse median nerve stimulation. Note that VIM and VPL EPs have an onset at approximately 11 ms and first peak at 15 ms.

Figure 3

Robust evoked responses are observed in the microelectrodes proximal to the target thalamic nuclei, with a change in the amplitudes as a function of burst frequency. Representative evoked VIM responses are presented for each of the 10 micro-contacts.

Figure 4

EP in response to a single 5-Hz pulse (top row) and 50–170-Hz burst stimulation for a single VIM contact in a single subject. (A) Average traces for the 100-ms stimulation period (negative time values) and the 100-ms EP response used for analysis (positive time values). Stimulation artifacts appear as thick vertical lines. (B) Power spectral density (PSD) calculated from the 100-ms period after the last stimulus pulse. Dotted line indicates stimulation frequency.

Figure 5

EP amplitudes from left (blue) and right (red) VIM as a function of stimulation frequency during stimulation of the contralateral median nerve, for each subject. Horizontal axes indicate stimulation frequency. Vertical axes represent the normalized EP values.

Figure 6

Average EP amplitude as a function of pulse frequency for all subjects. EP amplitude differed significantly between 50 vs. 140 Hz (p < 0.01), 50 vs. 170 Hz (p < 0.001), and 100 vs. 170 Hz (p < 0.001). Error bars indicate standard error.