Theta burst stimulation applied over primary motor and somatosensory cortices produces analgesia unrelated to the changes in nociceptive event-related potentials

Abstract

Continuous theta burst stimulation (cTBS) applied over the primary motor cortex (M1) can alleviate pain although the neural basis of this effect remains largely unknown. Besides, the primary somatosensory cortex (S1) is thought to play a pivotal role in the sensori-discriminative aspects of pain perception but the analgesic effect of cTBS applied over S1 remains controversial. To investigate cTBS-induced analgesia we characterized, in two separate experiments, the effect of cTBS applied either over M1 or S1 on the event-related brain potentials (ERPs) and perception elicited by nociceptive (CO2 laser stimulation) and non-nociceptive (transcutaneous electrical stimulation) somatosensory stimuli. All stimuli were delivered to the ipsilateral and contralateral hand. We found that both cTBS applied over M1 and cTBS applied over S1 significantly reduced the percept elicited by nociceptive stimuli delivered to the contralateral hand as compared to similar stimulation of the ipsilateral hand. In contrast, cTBS did not modulate the perception of non-nociceptive stimuli. Surprisingly, this side-dependent analgesic effect of cTBS was not reflected in the amplitude modulation of nociceptive ERPs. Indeed, both nociceptive (N160, N240 and P360 waves) and late-latency non-nociceptive (N140 and P200 waves) ERPs elicited by stimulation of the contralateral and ipsilateral hands were similarly reduced after cTBS, suggesting an unspecific effect, possibly due to habituation or reduced alertness. In conclusion, cTBS applied over M1 and S1 reduces similarly the perception of nociceptive inputs originating from the contralateral hand, but this analgesic effect is not reflected in the magnitude of nociceptive ERPs.

Figure 1. Experimental design.

EEG responses to nociceptive and non-nociceptive somatosensory stimuli recorded before and after applying continuous theta burst stimulation (cTBS) over the left or right M1 or S1 cortex. The second EEG recording was always completed within 20 minutes after the end of cTBS. Nociceptive ERPs, late-latency non-nociceptive ERPs and early-latency non-nociceptive ERPs were recorded following stimulation of the left and right hand, in six separate blocks. The order of the blocks was counterbalanced across subjects, but identical in the two recording sessions.

Figure 2. S1 coil target location in in nine representative subjects of experiment S1.

Left panel. The S1 target was identified using a custom MRI-guided neuronavigation system. The position of the coil was adjusted to target the post-central gyrus at a location mirroring the M1 hotspot relative to the central sulcus, i.e. the location expected to correspond to the representation of the hand within S1. The M1 (blue) and S1 (red) targets are shown on the cortical surface reconstructed from the individual MRI data of nine representative subjects. Right panel. Using our MRI-guided approach to target S1, we found that the actual location of the coil on the scalp surface was both more posterior and more lateral relative to the M1 coil position (x-axis: medial-lateral distance relative to the M1 coil position; y-axis: anterior–posterior distance relative to the M1 coil position).

Figure 3. Intensity of perception.

Left panel. Following cTBS applied over M1 and S1, the intensity of the percept elicited by nociceptive stimuli delivered to the hand contralateral to the stimulated hemisphere was significantly reduced. Right panel. In contrast, cTBS did not modulate the percept elicited by non-nociceptive somatosensory stimuli.

Figure 4. Nociceptive somatosensory ERPs (group-level average waveforms).

There was no significant effect of cTBS on the magnitude of the N160 wave. In contrast, the magnitude of the N240 and P360 waves was significantly reduced following cTBS, regardless of whether the nociceptive stimuli were delivered ipsilateral vs. contralateral to the hemisphere onto which cTBS was applied.

Figure 5. Time-frequency analysis of the EEG responses to nociceptive somatosensory stimuli.

The colour maps represent the group-level average EEG signal amplitude expressed as percentage of change relative to baseline (ER%). x-axis: time (s); y-axis: frequency (Hz). Three time-frequency regions of interest were defined: ROI-ERP circumscribing the phase-locked nociceptive ERP, ROI-ERS circumscribing an early, non phase-locked enhancement of signal power between 10–20 Hz and ROI-ERD circumscribing a long-lasting desynchronization of alpha-band power. Mirroring the effect of cTBS on the magnitude of the N240 and P360 waves, the magnitude of ROI-ERP and ROI-ERS was significantly reduced following cTBS, regardless of whether the nociceptive stimuli were delivered ipsilateral vs. contralateral to the stimulated hemisphere. The magnitude of ROI-ERD was unaffected by cTBS.

Figure 6. Late-latency non-nociceptive somatosensory ERPs (group-level average waveforms).

There was no significant effect of cTBS on the magnitude of the N140 wave. In contrast, the magnitude of the P200 wave was significantly reduced following cTBS, regardless of whether the nociceptive stimuli were delivered ipsilateral vs. contralateral to the hemisphere onto which cTBS was applied.

Figure 7. Early-latency non-nociceptive somatosensory ERPs (group-level average waveforms).

There was no effect of cTBS on the magnitude of the N20, P27, N30, P45 and N60 waves. In contrast, there was a specific effect of cTBS on the magnitude of the P100 wave. The magnitude of the P100 elicited by stimuli delivered on the hand contralateral to the hemisphere onto which cTBS was applied was decreased following cTBS applied over M1 and increased following cTBS over S1. The opposite pattern was observed for the P100 elicited by stimuli delivered to the hand ipsilateral to the stimulated hemisphere. However, this observation was not confirmed by post-hoc pairwise comparisons.

Figure 8. Time-frequency analysis of the EEG responses to non-nociceptive somatosensory stimuli.

The colour maps represent the group-level average EEG signal amplitude expressed as percentage of change relative to baseline (ER%). x-axis: time (s); y-axis: frequency (Hz). Three time-frequency regions of interest were defined: ROI-ERP circumscribing the phase-locked ERP, ROI-ERS circumscribing an early, non phase-locked enhancement of signal power between 10–20 Hz and ROI-ERD circumscribing a long-lasting desynchronisation of alpha-band power. The magnitudes of ROI-ERP, ROI-ERS and ROI-ERD were not significantly modulated by cTBS.