Brainstem involvement in the initial response to pain

Abstract

The autonomic responses to acute pain exposure usually habituate rapidly while the subjective ratings of pain remain high for more extended periods of time. Thus, systems involved in the autonomic response to painful stimulation, for example the hypothalamus and the brainstem, would be expected to attenuate the response to pain during prolonged stimulation. This suggestion is in line with the hypothesis that the brainstem is specifically involved in the initial response to pain. To probe this hypothesis, we performed a positron emission tomography (PET) study where we scanned subjects during the first and second minute of a prolonged tonic painful cold stimulation (cold pressor test) and nonpainful cold stimulation. Galvanic skin response (GSR) was recorded during the PET scanning as an index of autonomic sympathetic response. In the main effect of pain, we observed increased activity in the thalamus bilaterally, in the contralateral insula and in the contralateral anterior cingulate cortex but no significant increases in activity in the primary or secondary somatosensory cortex. The autonomic response (GSR) decreased with stimulus duration. Concomitant with the autonomic response, increased activity was observed in brainstem and hypothalamus areas during the initial vs. the late stimulation. This effect was significantly stronger for the painful than for the cold stimulation. Activity in the brainstem showed pain-specific covariation with areas involved in pain processing, indicating an interaction between the brainstem and cortical pain networks. The findings indicate that areas in the brainstem are involved in the initial response to noxious stimulation, which is also characterized by an increased sympathetic response.

Introduction

Supraspinal processing of pain has been divided between processes in brainstem networks and in the cortex (Price, 2000). Although these systems are highly interactive, it has been postulated that they can be more or less involved in different phases of pain processing. While the brainstem seems to be more important in the immediate pain response, higher order regions seem to be more involved in the later response (Price, 2000).

However, the brainstem mediates a response to a complex situation, which may not just consist of pain. Instead, it is a way for the organism to rapidly cope in a life-threatening situation with a set of prepackaged behaviors to increase the chance for survival Bolles, 1970, Fanselow, 1994, Timberlake, 1993. These states have been described for experimental animals both in encounters with predators Blanchard et al., 1990, Fanselow and Sigmundi, 1986 but also elicited by noxious stimulation (Fanselow, 1982), for example, the dorsolateral periaqueductal grey (PAG) may independently of cortical networks increase autonomic responses, induce flight or fight behaviors, and nonopioid analgesia in direct response to noxious stimulation. Thus, it may be proposed that the dorsolateral PAG is involved in the initial response to noxious input that also includes an autonomic response (Fanselow, 1994).

It is unlikely that the dorsolateral PAG is the only brainstem region involved in this response. Several other brainstem nuclei, such as the parabrachial nucleus, the locus coeruleus, and the hypothalamus, also receive direct nociceptive information from the spinobulbar tract (Craig and Dostrovsky, 1999). These regions interact with a more elaborate brainstem network that may both modulate cortical and spinal processing Craig and Dostrovsky, 1999, Fields and Basbaum, 1999, Parvizi and Damasio, 2001.

The studies presented above are in line with the known relationship between acute pain and the sympathetic response on a behavioral level in human subjects Gracely, 1999, Jäning, 1995. The idea that the brainstem is more involved in the initial pain response may explain habituation of the autonomic responses during tonic pain (Lovallo, 1975) although the subjective pain ratings increase or remain equal Rainville et al., 1992, Tassorelli et al., 1995. Brain systems involved in the autonomic response during pain should therefore attenuate their activity during prolonged painful stimulation, while other regions participating in the processing of pain intensity and unpleasantness should increase in activity or at least remain equally activated during the prolonged pain. Thus, stimulus duration should have an opposite effect on brainstem vs. cortical activity during the processing of nociceptive signals. To test this hypothesis, we performed a PET study where we scanned different time components in prolonged painful cold stimulation and nonpainful cold stimulation. A rapid communication from the same data set on the S1 functional connectivity has previously been published (Petrovic et al., 2002b).