Deep continuous theta burst stimulation of the operculo-insular cortex selectively affects Aδ-fibre heat pain

Abstract

Key points: Deep continuous theta burst stimulation (cTBS) of the right operculo-insular cortex delivered with a double cone coil selectively impairs the ability to perceive thermonociceptive input conveyed by Aδ-fibre thermonociceptors without concomitantly affecting the ability to perceive innocuous warm, cold or vibrotactile sensations. Unlike deep cTBS, superficial cTBS of the right operculum delivered with a figure-of-eight coil does not affect the ability to perceive thermonociceptive input conveyed by Aδ-fibre thermonociceptors. The effect of deep operculo-insular cTBS on the perception of Aδ-fibre input was present at both the contralateral and the ipsilateral hand. The magnitude of the increase in Aδ-heat detection threshold induced by the deep cTBS was significantly correlated with the intensity of the cTBS pulses. Deep cTBS delivered over the operculo-insular cortex is associated with a risk of transcranial magnetic stimulation-induced seizure.

Abstract: Previous studies have suggested a pivotal role of the insular cortex in nociception and pain perception. Using a double-cone coil designed for deep transcranial magnetic stimulation, our objective was to assess (1) whether continuous theta burst stimulation (cTBS) of the operculo-insular cortex affects differentially the perception of different types of thermal and mechanical somatosensory inputs, (2) whether the induced after-effects are lateralized relative to the stimulated hemisphere, and (3) whether the after-effects are due to neuromodulation of the insula or neuromodulation of the more superficial opercular cortex. Seventeen participants took part in two experiments. In Experiment 1, thresholds and perceived intensity of Aδ- and C-fibre heat pain elicited by laser stimulation, non-painful cool sensations elicited by contact cold stimulation and mechanical vibrotactile sensations were assessed at the left hand before, immediately after and 20 min after deep cTBS delivered over the right operculo-insular cortex. In Experiment 2, Aδ-fibre heat pain and vibrotactile sensations elicited by stimulating the contralateral and ipsilateral hands were evaluated before and after deep cTBS or superficial cTBS delivered using a flat figure-of-eight coil. Only the threshold to detect Aδ-fibre heat pain was significantly increased 20 min after deep cTBS. This effect was present at both hands. No effect was observed after superficial cTBS. Neuromodulation of the operculo-insular cortex using deep cTBS induces a bilateral reduction of the ability to perceive Aδ-fibre heat pain, without concomitantly affecting the ability to perceive innocuous warm, cold or vibrotactile sensations.

Keywords: Continuous Theta Burst Stimulation; Insula; Nociception; Operculum; Psychophysics.

© 2018 The Authors. The Journal of Physiology © 2018 The Physiological Society.

Figures

Figure 1. Experimental design

Both experiments followed the same procedure. Detection thresholds were monitored before (T0), immediately after (T1) and 20 min after cTBS (T3). The perceived intensity elicited by suprathreshold stimuli was monitored before (T0) and 10 min after cTBS (T2). Note that the threshold measurements at T0 and T1 and the perception evaluation at T0 and T2 were separated by approximately the same amount of time. In Experiment 1, thresholds and perception were assessed for four modalities: Aδ‐heat, C‐heat, Aδ‐cool and Aβ‐vibrotactile stimuli delivered on the contralateral hand (left) relative to the right insular cortex onto which deep cTBS was applied. In Experiment 2, sensory changes were assessed for two modalities: Aδ‐heat and Aβ‐vibrotactile stimuli delivered on the contralateral and ipsilateral hand at the same time points before and after deep cTBS of the operculo‐insular cortex or superficial cTBS of the operculum. The experimental procedures were completed within 30 min following cTBS.

Figure 2. Localization of the target sites in the dorso‐posterior insular cortex and the localization of the positions of the TMS coil

A, position of the TMS double cone coil with the handle pointing backwards during the cTBS protocol; the participants were lying in left lateral decubitus position. B, localization of the target sites in the dorso‐posterior insular cortex (black circles; MNI coordinates x: [34 to 42], y: [−17 to −5], z: [4 to 14]) defined on individual 3D structural MRI image for the MRI‐guided neuronavigation system. C, projections on the cortical surface of the position of the centre of the TMS coil (white circles) at which cTBS was applied for all participants; the corresponding target markers in the dorsal posterior insula are indicated by the black circles; the location of the insula is indicated in dark grey. The apparent discrepancy between the target markers and the position of the coil is due to the 2D visualisation of the different orientations of the TMS coil according to the individual curvature of the head (MNI Colin 27 brain reconstruction adapted from JuBrain; Mohlberg et al. 2012).

Figure 3. Example of the threshold intermingled staircase procedure for Experiment 1

In this representative participant, the first stimulus delivered was Aβ‐vibrotactile (1; lower panel) followed by a C‐heat stimulus (2) followed by a Aδ‐cool stimulus (3) followed by an Aδ‐heat stimulus (4; upper panel). The order is indicated by the arrows. This sequence was repeated until four reversals (open circles) were obtained for each modality. The order of the modalities was counterbalanced across participants. The different thresholds were computed within each modality by averaging stimulation intensities of the first four staircase reversals. In Experiment 2, the procedure was the same with alternative delivery of Aδ‐heat and Aβ‐vibrotactile stimuli. This procedure was conducted on one hand at a time. The order of the first tested hand was counterbalanced across participants.

Figure 4. The effect of deep cTBS over the right operculo‐insular cortex on detection thresholds in Experiment 1

Bar graphs represent the individual changes (increase in red, decrease in blue) in detection thresholds (n = 10) immediately after cTBS (T1–T0) and 20 min after cTBS (T3–T0) for the four somatosensory modalities: Aδ‐heat, C‐heat, Aδ‐cool and Aβ‐vibrotactile delivered on the left contralateral hand. The lower graphs show the individual absolute thresholds at T0, T1 and T3; group‐level average is displayed in green (mean ± SD).

Figure 5. Effects of deep cTBS over the right operculo‐insular cortex on perceived intensity in Experiment 1

Bar graphs represent the individual changes in perception (numerical rating scale, NRS; increase in red, decrease in blue; n = 10 except for Aβ‐vibrotactile where n = 9 because of missing data) elicited by the four different suprathreshold stimuli delivered on the contralateral hand: Aδ‐heat (60°C), C‐heat (44°C), Aδ‐cool (10°C) and Aβ‐vibrotactile (95 μm) 10 min after cTBS (T2–T0). The lower graphs show the individual intensity of perception at T0 and T2; group‐level average is displayed in green (mean ± SD).

Figure 6. The effect of cTBS on detection thresholds in Experiment 2

Effects of deep and superficial cTBS of the operculo‐insular cortex on Aδ‐heat and Aβ‐vibrotactile thresholds. Bar graphs represent the individual changes (increase in red, decrease in blue) in threshold (n = 7) immediately after cTBS (T1–T0) and 20 min after cTBS (T3–T0). The lower graphs show the individual thresholds; group‐level average is displayed in green (mean ± SD).

Figure 7. Effect of deep and superficial cTBS on intensity of perception in Experiment 2

Bar graphs indicate individual changes in perception (numerical rating scale, NRS; increase in red, decrease in blue; n = 7) elicited by Aδ‐heat (60°C) and Aβ‐vibrotactile (95 μm) stimuli. The lower graphs show the individual perception; group‐level average is displayed in green (mean ± SD).

Figure 8. Relative frequency distribution of reaction times to suprathreshold Aδ‐heat and Aβ‐vibrotactile stimuli in Experiment 2

A, effect of deep and superficial cTBS on reaction times to suprathreshold Aδ‐heat stimuli (60°C). The relative frequency distribution of RTs are displayed before (T0) and 10 min after (T2) cTBS. B, effect of deep and superficial cTBS on RTs to suprathreshold Aβ‐vibrotactile stimuli (95 μm) at T0 and T2.

Figure 9. Quality of perception of suprathreshold stimuli in Experiment 2

A, quality of perception of suprathreshold Aδ‐heat stimuli. B, quality of perception of suprathreshold Aβ‐vibrotactile stimuli (95 μm). Pie charts represent the proportion of the use of each descriptor before (T0) and 10 min after (T2) cTBS. In both cases, all stimuli were perceived; in all conditions Aδ‐heat stimuli were mainly perceived as painful (burning or pricking) before and after deep or superficial cTBS.

Figure 10. Relationship between the effect of deep cTBS on thresholds and intensity of cTBS pulses

A, linear regression for all the participants (Experiment 1 and 2) between the intensity of cTBS pulses used during deep cTBS of the operculo‐insular cortex and the relative difference in Aδ‐heat threshold 20 min after cTBS (T3–T0). The increase of Aδ‐heat threshold was significantly positively correlated with the intensity of cTBS pulses (r = 0.733; n = 16; P = 0.001; one observation, indicated by *, was excluded due to a standardized residual greater than 3 standard deviations; when all participants were included: r = 0.613; n = 17; P = 0.009). B, linear regression for all the participants (Experiment 1 and 2) between the intensity of cTBS pulses used during deep cTBS of the operculo‐insular cortex and the relative difference in Aβ‐vibrotactile threshold 20 min after cTBS (T3–T0). There was no significant correlation (r = 0.110; n = 17; P = 0.676). C, linear regression for all the participants (Experiment 1 and 2) between the intensity of deep cTBS pulses, corrected for the coil–cortical target distance, used during deep cTBS of the operculo‐insular cortex and the relative difference in Aδ‐heat threshold 20 min after cTBS (T3–T0). The increase of Aδ‐heat threshold was significantly positively correlated with the intensity of cTBS pulses (r = 0.592; n = 16; P = 0.016).