Increased cerebral functional connectivity underlying the antinociceptive effects of hypnosis

Abstract

The neural mechanisms underlying the antinociceptive effects of hypnosis are not well understood. Using positron emission tomography (PET), we recently showed that the activity in the anterior cingulate cortex (midcingulate area 24a′) covaries with the hypnosis-induced reduction of affective and sensory responses to noxious thermal stimulation [Faymonville et al., Anesthesiology 92 (2000) 1257–1267]. In the present study, we assessed changes in cerebral functional connectivity related to the hypnotic state, compared to simple distraction and the resting state. Nineteen highly hypnotizable right-handed volunteers were studied using H₂¹⁵O-PET. The experimental conditions were hot noxious or warm non-noxious stimulation of the right hand during resting state, mental imagery and hypnotic state. Using a psychophysiological interaction analysis, we identified brain areas that would respond to noxious stimulations under the modulatory action of the midcingulate cortex in, and only in, the hypnotic state. Hypnosis, compared to the resting state, reduced pain perception by 50%. Pain perception during rest and mental imagery was not significantly different. Analysis of PET data showed that the hypnotic state, compared to normal alertness (i.e., rest and mental imagery), significantly enhanced the functional modulation between midcingulate cortex and a large neural network encompassing bilateral insula, pregenual anterior cingulate cortex, pre-supplementary motor area, right prefrontal cortex and striatum, thalamus and brainstem. These findings point to a critical role for the midcingulate cortex in the modulation of a large cortical and subcortical network underlying its influence on sensory, affective, cognitive and behavioral aspects of nociception, in the specific context of hypnosis.

Introduction

Hypnosis combined with slight conscious sedation (i.e., hypnosedation) and local anesthesia is now considered a valuable alternative to general anesthesia in specific indications [18], [20], [21], [30], [31], [37]. Since 1992 we have used hypnosis routinely in more than 3300 surgical procedures. The underlying neuromodulatory effects of hypnosis remain, however, not fully understood. Studies of the antinociceptive effects of hypnosis have labored under a double burden: both hypnotic experience and pain experience are highly subjective phenomena. Factors that evoke pain reduction range from extrinsic psychosocial (e.g., interactions between clinician and patient) to intrinsic psychophysiological (e.g., modulation of pain signal transmission [47]). Recent positron emission tomography (PET) studies have demonstrated that the decreased perception of pain during hypnosis is related to changes in the activity (i.e., regional cerebral blood flow—rCBF) measured in the midcingulate cortex (area 24′ [19], [48]). We here test the hypothesis that hypnosis-induced analgesia can be explained by an enhanced modulation between the midcingulate cortex and the large neural networks involved in sensory, affective and cognitive aspects of noxious processing. Using a psychophysiological interaction analysis [23], we assessed hypnosis-specific increases in functional connectivity between the midcingulate cortex, identified in our previous study [19], and the rest of the brain.

Complementary to the concept of functional segregation as a principle of organization of the human brain (i.e., localizing a function to a cerebral area), recent neuroimaging techniques have focused on functional integration (i.e., assessing the interactions between functionally segregated areas mediated by changes in functional connectivity). Functional connectivity is defined as the temporal correlation of a neurophysiological index (i.e., rCBF) measured in different remote brain areas. Anatomical connectivity (e.g., neuroanatomic tracer studies obtained in animals) is a necessary underpinning for the assessment of functional connectivity. A psychophysiological interaction means that the contribution of one area to another (i.e., regression slope) changes significantly with the experimental context [23]. The psychophysiological interaction analysis used in the present study, aims at explaining the activity in one cortical area in terms of an interaction between the influence of a chosen area (i.e., midcingulate cortex) and some experimental condition (i.e., being in a hypnotic state or not). Pain is a multidimensional experience including sensory-discriminative, affective-emotional, cognitive and behavioral components. Its cerebral correlate is best described in terms of neural circuits or networks, referred to as the ‘neuromatrix’ for pain processing, and not as a localized ‘pain center’ [25]. The aim of the present study is to explore the modulatory role of the midcingulate cortex on the activity of this ‘neuromatrix’ in the specific context of hypnosis.