The role of the autonomic nervous system in acute surgical pain processing – what do we know?

For all anaesthetists, the link between peri-operative pain and sympathetic nervous system activity seems obvious, and has actually become second nature to their profession. Changes in heart rate and blood pressure in response to painful stimuli, for example, are routinely used as surrogate markers to assess analgesia in anaesthetised patients. If asked to describe the nature of the relationship between pain and sympathetic nervous system activity, most anaesthetists would therefore probably regard an increased sympathetic outflow as a mere epiphenomenon or symptom of pain. This school of thought is based not only on textbook knowledge but also predominantly on our day-to-day clinical experience. It implies a unidimensional, one-way, reflex pathway, with the obligatory noxious stimulus at the start and the observed sympathetic nervous system activity as the downstream endpoint. That this notion might be an oversimplification of pathway interactions, even in acute surgical pain, is one of the exciting findings of the case series of McDonnell et al. in this issue [1]. The authors showed a remarkable reduction in postoperative analgesic requirements after blockade of the sympathetic stellate ganglion with the local anaesthetic and sodium channel blocker, lidocaine. The stellate ganglion contains the somata of postganglionic sympathetic fibres and afferent fibres. Both of these might be inhibited by lidocaine. Therefore, either efferent or afferent pathways could be involved, thus leaving the question open as to whether modulation of pain sensation is achieved in the periphery or centrally. However, in any case, their work suggests an autonomic nervous system involvement in nociceptive processing after upper limb surgery. Based on these results, one can think of two main principles of how the autonomic nervous system might contribute to modulation of pain pathways: (i) through excitatory sympathetic or (ii) through inhibitory parasympathetic mechanisms. Employment of excitatory sympathetic nervous system mechanisms, for example, could mean that afferent sympathetic fibres are directly capable of either conducting noxious signals or enhancing Aδ- or C-fibres activity. In contrast, parasympathetic inhibition would reduce these signals or nerve fibre activity. The participation of the sympathetic nervous system in pain generation is well recognised and best described in chronic, usually neuropathic, pain states such as complex regional pain syndrome, but also for instance in cancer, vascular or visceral pain [2, 3]. Subsequently, for those phenomena the term ‘sympathetically maintained pain’ has been coined [4]. However, whether or not the sympathetic nervous system is similarly involved in acute pain is still uncertain. Data from animal and human volunteer studies strongly indicate that under physiological conditions, sympathetic outflow has no effect on nociceptor function and activity [5, 6]. However, this might change under circumstances in which primary afferents have been sensitised by trauma or inflammation. In animal models, sympathomimetics have been found to increase nociception-induced behaviour after nerve or heat injury, as well as in response to chemically induced inflammatory processes, whilst sympatholytics reduce nociception [5–7]. These findings are in agreement with experiments where the influence of sympathetic stimulation on the electrical activity of Aδ- and C-fibres was tested before and after sensitisation [8, 9]. The results thus obtained showed that inflammation and hence inflammatory mediators changed the excitation characteristics of these fibres. They subsequently became receptive to sympathetic modulation and therefore sensitised to sympathetic stimulation [8, 9]. In contrast to the animal data, surprisingly little is known about involvement of the sympathetic nervous system in acute somatic pain pathways in humans. To date, only one study has investigated the effect of sympathetic blockade on heat injury-induced inflammatory pain in healthy volunteers [10]. In this study, Pedersen and colleagues could find no effect of lumbar sympathetic blockade on acute pain, pain thresholds or hyperalgesia [10]. Therefore, although evidence from humans is lacking, animal research indicates the possibility that the sympathetic nervous system is capable of modulating afferent pain pathways, probably in an excitatory pain-enhancing way. However, prior sensitisation of the nervous system through disease processes seems to be required. If this is the evidence for the sympathetic nervous system, what then can be said about the role of the parasympathetic nervous system, and especially its predominant component, the vagus nerve? Anatomically, the vagus nerve consists of afferent as well as efferent branches, supplying mostly visceral organs. Not surprisingly, its function has therefore been best described in relation to sensory information processing from these structures [11]. Because of this role as a major pathway of information transfer from inner organs to the central nervous system, an involvement of the vagus in visceral pain has also been suggested. However, results in this regard are conflicting and hence a vagal contribution to visceral pain is still contentious [12]. In contrast, more and more data are now mounting to indicate involvement of the vagus nerve in acute somatic nociception. For example, vagal afferents in the lumbar spinal cord have been found either to facilitate or to reduce peripheral nociceptive processing after noxious thermal stimuli in rats [13]. This biphasic effect was a function of the intensity of the stimuli applied to the afferents, with increasing intensity shifting the balance from facilitation to inhibition [13]. Anatomically, primary afferent vagal fibres terminate in the nucleus tractus solitarius in the brainstem. From here, descending inhibitory pathways originate to descend back to the spinal cord. Hence, both the vagus and the nucleus tractus solitarius are parts of a spinal-bulbo-spinal anti-nociceptive circuit [14]. Evidence that the inhibitory anti-nociceptive parasympathetic pathways are clinically more important than excitatory pro-nociceptive parasympathetic mechanisms was obtained by studies involving epileptic patients with implanted vagal nerve stimulators. Pain during tonic pressure stimuli, mechanical impact and noxious heat was assessed before and after stimulator implantation. Results showed that stimulation of the vagus nerve is capable of reducing acute pressure pain but not thermal or mechanically induced pain [15, 16]. Based on their results, the authors hypothesised that the electrical stimulation caused an activation of the afferent part of the vagus. This signal would then be transmitted to the nucleus tractus solitarius and from there to a number of higher autonomic nuclei, as well as the nucleus raphe magnus and locus coereleus, to cause subsequent activation of descending inhibitory pathways [15]. In addition to this centrally induced analgesic effect of vagal nerve stimulation, in a second series of experiments the authors also found evidence for a peripheral mode of action [16]. This is because nerve stimulation in those experiments significantly reduced pain-induced local vasodilation. Whether this finding could be explained via an anti-inflammatory effect of vagus nerve stimulation as suggested by preclinical data [17], or a reduced mediator release from C-fibres as would follow from Kolzenburg’s work [18], remains currently elusive. So far, the potential involvement of each division of the autonomic nervous system in acute pain has been discussed in isolation. However, daily experience, especially as far as the cardiovascular system is concerned, tells us that modulation of one part of the autonomic nervous system subsequently also affects the other [19]. Therefore, it might be reasonable to assume similar interactions for the sensory function of the autonomic nervous system as well. The possible communications between the sympathetic and parasympathetic nervous systems have been comprehensively reviewed by Benarroch [20]. They occur as direct local interactions in the periphery and also as part of complex integrated signalling circuits in the entire nervous system. Primary afferents, for example, can synapse with preganglionic sympathetic fibres in the spinal cord to form reflex pathways. However, of particular interest in pain processing are spinal-bulbo-spinal connections, since these are integrating information from both the periphery and higher brain centres. The participating brainstem nuclei receive inputs from both visceral and nociceptive afferents, and are thought to be relay stations that serve the purpose of modulating incoming nociceptive signals. As such, they can either exert excitatory or inhibitory actions on the afferent impulses [21–23]. Since this helps to augment or attenuate a painful signal, respectively, the autonomic nervous system is therefore part of the endogenous descending pain modulating system. Furthermore, some of the brainstem nuclei – apart from receiving sensory noxious information – are also involved in the control of cardiac and vascular function. Through this dual role, they are capable of translating nociceptive input into autonomic outflow. Thus, they are responsible for inducing changes in heart rate and blood pressure in response to pain – changes the anaesthetist in the operating theatre relies upon to estimate stimulus intensity and hence patient comfort during surgery. Brainstem nuclei classically associated with autonomic functions and the descending control of pain are the parabrachial nucleus in the pons, the peri-aqueductal grey in the midbrain, the nucleus tractus solitarius, the ventrolateral reticular formation and the raphe nuclei in the medulla. These centres are interlinked not only with each other but also with regions in the brain that participate in autonomic regulation, such as the insular cortex, the anterior cingulated cortex, the amygdala and the hypothalamus. They therefore form a network within the central nervous system that is concerned with autonomic processing and hence called the ‘central autonomic network’ [24]. Interestingly, both the insular cortex and the anterior cingulated cortex are also part of the ‘pain matrix’ [25]. The pain matrix is defined from MRI studies in man, and is formed by those brain areas that are consistently activated in response to acute noxious stimuli [26]. In summary, although clinical evidence about an involvement of the autonomic nervous system in the processing of acute pain is scarce, there are some data to suggest a contribution of both sympathetic and parasympathetic pathways. However, the nature, extent and therapeutic potential of this contribution remain unclear and warrant further research. Finally, from the evidence reviewed herein it appears that sympathetic and parasympathetic influences on nociception differ in one important point. A significant role in supra-spinal pain modulation under physiological conditions seems to be likely for the parasympathetic nervous system, but not the sympathetic system. Sympathetic mechanisms seem to require sensitisation or ‘priming’ of the sensory system through already established pathological processes before they can become effective. Since trauma and inflammation, the basic ingredients of every surgical insult, have been identified as important primers, it appears plausible that the sympathetic nervous system participates in peri-operative pain. McDonnell et al.’s [1] results add substance to this idea and we hope will encourage further studies – under properly controlled conditions [27] – to determine the role of the autonomic nervous system in acute pain processing and the scope for analgesic treatment. No external funding and no competing interests declared.