Suppression of beta oscillations in the subthalamic nucleus following cortical stimulation in humans

Abstract

It is unclear how subthalamic nucleus activity is modulated by the cerebral cortex. Here we investigate the effect of transcranial magnetic stimulation (TMS) of the cortex on oscillatory subthalamic local field potential activity in the 8-35 Hz (alpha/beta) band, as exaggerated synchronization in this band is implicated in the pathophysiology of parkinsonism. We studied nine patients with Parkinson's disease (PD) to test whether cortical stimulation can modulate synchronized oscillations in the human subthalamic nucleus. With patients at rest, single-pulse TMS was delivered every 5 s over each primary motor area and supplementary motor area at intensities of 85-115% resting motor threshold. Subthalamic local field potentials were recorded from deep brain stimulation electrodes implanted into this nucleus for the treatment of PD. Motor cortical stimulation suppressed beta activity in the subthalamic nucleus from approximately 0.2 to 0.6 s after TMS (repeated measures anova; main effect of time, P < 0.01; main effect of side, P = 0.03), regardless of intensity. TMS over the supplementary motor area also reduced subthalamic beta activity at 95% (P = 0.05) and 115% resting motor threshold (P = 0.01). The oscillatory activity decreased to 80 +/- 26% of baseline (averaged across sites and stimulation intensities). Suppression with subthreshold stimuli confirmed that these changes were centrally driven and not due to peripheral afference. The results may have implications for mechanisms underlying the reported therapeutic benefits of cortical stimulation.

Fig. 2

Changes in power in the right subthalamic nucleus over the frequency range 1–99 Hz induced by: (A) ipsilateral primary motor area (M1) transcranial magnetic stimulation (TMS) at 95% resting motor threshold (RMT) in case 8; (B) ipsilateral M1 TMS at 115% RMT in case 2; (C) ipsilateral M1 TMS at 85% RMT in case 3; and (D) contralateral M1 TMS at 115% RMT in case 8. Data have been thresholded according to 95% confidence limits of the baseline. Stimulus is at 0. White circles indicate beta event-related desynchronization.

Fig. 1

Event-related potentials in Parkinson’s disease patients. (A) Grand average of subthalamic nucleus (STN) activity following transcranial magnetic stimulation (TMS) delivered to the ipsilateral (black, n=14 STNs) and contralateral (grey, n=13 STNs) primary motor area at 115% resting motor threshold (RMT). (B) Grand average of STN activity following TMS over the supplementary motor area (SMA) at 115% RMT (black, n=10 STNs) and at 95% RMT (grey, n=6). (C) Grand average of STN activity following TMS over deep brain stimulation (DBS) leads (n=8 STNs) at 115% RMT. Open arrows denote stimulus artefact. Closed arrows denote evoked potentials. Data from left and right STNs have been combined.

Fig. 3

Group mean changes in subthalamic nucleus (STN) power induced by transcranial magnetic stimulation (TMS) over the primary motor area (M1). Mean changes in power in the STN over the frequency range 1–99 Hz induced by: (A) ipsilateral M1 TMS at 85%/95% resting motor threshold (RMT) (n=14 STNs); (B) ipsilateral M1 TMS at 115% RMT (n=14 STNs); and (C) contralateral M1 TMS at 115% RMT (n=13 STNs). White circles indicate beta event-related desynchronization. (D–F) Grand means of local field potential data averaged to TMS from the same dataset as shown in A–C, with an identical time scale. (G and H) Grand averages of individualized beta activity from same dataset as in A and B. (H inset) Averaged beta response to ipsilateral suprathreshold stimulation from patients recorded in London (black) and Rome (blue). See also supplementary Fig. S2. Red lines indicate values significantly smaller than the lower 95% confidence limit. All changes are expressed as percentages of the baseline mean, taken 1.59–0 s prior to stimulation. Data from both STN sides have been combined in all matrix and line plots. Stimulus at 0 (open arrows). (I) Mean beta power (±SD) in STNs before (1.59–1.23 s) and after (0.19–0.55 s) suprathreshold (filled columns) and subthreshold (open columns) TMS at different cortical sites. Asterisks denote significant differences (P<0.05).

Fig. 4

Changes in subthalamic nucleus (STN) power induced by transcranial magnetic stimulation over the supplementary motor area (SMA) in Parkinson’s disease patients. Group mean colour matrix in response to stimulation at: (A) 95% resting motor threshold (RMT) (n=6 STNs) delivered over the SMA; (B) 115% RMT (n=10 STNs) over the SMA. White circles indicate beta event-related desynchronization. (C and D) Grand average of individualized beta activity from the same dataset as shown in A and B. Red lines indicate values significantly smaller than the lower 95% confidence limit. All changes are expressed as percentages of the baseline mean, taken 1.59–0 s prior to stimulation with data from both STN sides combined. Stimulus at 0 in all plots.

Fig. 5

Sham transcranial magnetic stimulation (TMS). (A) Mean changes in subthalamic nucleus (STN) power over the frequency range 1–99 Hz induced by suprathreshold TMS (115% resting motor threshold) delivered over the deep brain stimulation (DBS) leads. Data have been thresholded according to 95% confidence limits. (B) Grand average of individualized beta activity following suprathreshold TMS delivered over the DBS leads. No values fell below the lower 95% confidence limit. All changes are expressed as percentages of the baseline mean, taken 1.59–0 s prior to stimulation. Data from both STN sides have been combined (n=8 STNs). (C) Mean changes in STN power over the frequency range 1–99 Hz induced by suprathreshold TMS delivered over the occipito-parietal (OP) cortex (n=4 STNs). Data have been thresholded according to 95% confidence limits.