Cerebral processing of sharp mechanical pain measured with arterial spin labeling

Abstract

Introduction: Arterial spin labeling (ASL) is a functional neuroimaging technique that has been frequently used to investigate acute pain states. A major advantage of ASL as opposed to blood-oxygen-level-dependent functional neuroimaging is its applicability for low-frequency designs. As such, ASL represents an interesting option for studies in which repeating an experimental event would reduce its ecological validity. Whereas most ASL pain studies so far have used thermal stimuli, to our knowledge, no ASL study so far has investigated pain responses to sharp mechanical pain.

Methods: As a proof of concept, we investigated whether ASL has the sensitivity to detect brain activation within core areas of the nociceptive network in healthy controls following a single stimulation block based on 96 s of mechanical painful stimulation using a blunt blade.

Results: We found significant increases in perfusion across many regions of the nociceptive network such as primary and secondary somatosensory cortices, premotor cortex, posterior insula, inferior parietal cortex, parietal operculum, temporal gyrus, temporo-occipital lobe, putamen, and the cerebellum. Contrary to our hypothesis, we did not find any significant increase within ACC, thalamus, or PFC. Moreover, we were able to detect a significant positive correlation between pain intensity ratings and pain-induced perfusion increase in the posterior insula.

Conclusion: We demonstrate that ASL is suited to investigate acute pain in a single event paradigm, although to detect activation within some regions of the nociceptive network, the sensitivity of our paradigm seemed to be limited. Regarding the posterior insula, our paradigm was sensitive enough to detect a correlation between pain intensity ratings and pain-induced perfusion increase. Previous experimental pain studies have proposed that intensity coding in this region may be restricted to thermal stimulation. Our result demonstrates that the posterior insula encodes intensity information for mechanical stimuli as well.

Keywords: acute pain; cerebral blood flow; functional neuroimaging; perfusion.

FIGURE 1

Schematic representation of the paradigm. Mechanical pain was applied with a blunt blade to the left forearm within an area of approximately 7 cm² (shaded area in light red) for 96 s. Pain ratings were acquired 5 min after pain offset

FIGURE 2

Brain mask used in the ROI analysis. The yellow‐colored area represents the posterior insula as defined in the Juelich histological atlas, thresholded at a probability value of 0.2 and binarized. The area shaded in green represents the statistically significant voxels of the pain activation that intersected with the brain mask. The ROI analysis was performed with the average perfusion values of the voxel within the green area

FIGURE 3

(a) 3D view of significant clusters after cluster correction (cluster‐building threshold z > 3.5, p = .05). (b) Transversal view of the clusters at Z = 64/41/23/2/−16/−26, respectively

FIGURE 4

Normalized perfusion within the posterior insula (mask see Figure 3). Red dashed area: Time window in which noxious stimuli were applied. Red stippled line: Baseline perfusion in the absence of pain. Blue line: Perfusion averaged across 19 subjects with error bars (± standard error of the mean). Solid blue line: Average perfusion smoothed for better visualization

FIGURE 5

Correlation (Spearman's rho) between normalized perfusion within our posterior insula mask and (a) pain intensity: medium correlation (r = .444, p = 0.028) and (b) pain unpleasantness: large correlation (r = .506, p = .014). For the purpose of visualization, regression lines are plotted

FIGURE 6

Boxplot diagrams of (a) pain intensity ratings, (b) pain unpleasantness ratings, and (c) log‐transformed mechanical pain thresholds. Middle line = median; x = mean value; lower and upper bound = end of first and third quartile; whisker = minimum and maximum value; blue dot = outlier (defined as 1.5 times the interquartile range)