The Sympathetic Nervous System and Pain
The sympathetic nervous system (SNS) and pain interact on many levels of the neuraxis. In healthy subjects, activation of the SNS in the brain usually suppresses pain mainly by descending inhibition of nociceptive transmission in the spinal cord. Furthermore, some experimental data even suggest that the SNS might control peripheral inflammation and nociceptive activation. However, even subtle changes in pathophysiology can dramatically change the effect of SNS on pain, and vice versa. In the periphery, inflammation or nociceptive activation is enhanced, spinal descending inhibition is reversed to spinal facilitation, and finally the awareness of all these changes will induce anxiety, which furthermore amplifies pain perception, affects pain behavior, and depresses mood. Unraveling the detailed molecular mechanisms of how this interaction of SNS and pain is established in health and disease will help us to treat pain more successfully in the future.
KeywordsSympathetic nervous system Nociception Descending inhibition Neurogenic inflammation Complex Regional Pain Syndrome
In humans, the afferent nociceptive and the efferent autonomic system interact in many ways. On the one hand pain as a stressor activates the sympathetic nervous system (SNS); on the other hand, the activation of the autonomic nervous system (ANS) can suppress or in pathological states augment pain. This interaction might take place in the periphery or the central nervous system (CNS) (Cortelli and Pierangeli 2003; Benarroch 2006).
The Healthy State
The situation is quite different when the second transmitter of the SNS—acetylcholine (ACh)—is increased in the skin. Injection of pharmacological doses of ACh dose dependently induces a burning sensation (Vogelsang et al. 1995) and iontophoresis of ACh, which induces axon-reflex sweating, simultaneously activates a neurogenic flare response, which is indicative of the activation of chemosensitive C-fibers (Schlereth et al. 2005). However, all these studies used pharmacological ACh doses. Physiological concentrations would be rapidly degraded by choline esterases in human tissue (Schlereth et al. 2005). Therefore, it is still unclear whether ACh release in human skin normally results in C-fiber activation or even helps to suppress nociceptor responses, depending on which type of ACh receptors, nicotinergic or muscarinergic, are activated (Bernardini et al. 2001).
Experimental Pain and the Sympathetic Nervous System
The effect of SNS activation on experimental pain in humans has been mainly studied employing the capsaicin model. If capsaicin, an ingredient of the red chili pepper, is injected into the skin, it binds to the TRPV1 channel which is located on many peripheral C-fibers. This binding induces the immediate generation of an action potential barrage in primary afferent fibers, thereby inducing strong burning pain and the release of neuropeptides from the C-fibers in the vicinity of the injection site (Simone et al. 1989). It has been shown that controlled physiological activation of the SNS by cold thermoregulation has no impact on capsaicin-induced pain, neither in the skin (Baron et al. 1999) nor in muscles (Elam et al. 1999; Wasner et al. 2002). Another approach to study the impact of the SNS on experimental pain was to employ the heat pain hyperalgesia model (Pedersen et al. 1997). As with capsaicin, sympathetic blocks were also shown not to affect pain in these subjects.
On the other hand hyperalgesia to heat, which frequently occurs after capsaicin injection, was enhanced if norepinephrine was co-applied by iontophoresis or injection in pharmacological doses (Drummond 1998; Fuchs et al. 2001). This effect was reversed by combined blocking of alpha1- and alpha2- receptors (Liu et al. 1996; Kinnman et al. 1997). However, enhancement of capsaicin induced heat hyperalgesia was also found for other vasoconstrictive agents like vasopressin, angiotensin or the occlusion of blood flow (Drummond 1998), and it was antagonized by vasodilators like nitroprusside (Drummond 1999). These findings indicate that at least a part of catecholamine-amplification of heat sensitization must be mediated by local vasoconstriction rather than a specific catecholaminergic effect.
The Sympathetic Nervous System and Inflammation
There are some clinical symptoms which suggest a link between the SNS and immune reactions in human skin. In CRPS patients (see below), who are characterized by sympathetic dysfunction, Langerhans cells were found extensively in skin biopsies (Calder et al. 1998). Furthermore in these patients sympatholytic procedures ameliorate pain but also inflammation and edema. Another example is hemispherical stroke, which amplifies sympathetic outflow to the paretic side by disinhibition of the SNS. In stroke patients, T-cell responses of the delayed-type hypersensitivity reaction were significantly attenuated (Tarkowski et al. 1995). All these findings triggered in-depth studies on the mechanisms of sympathetic–immune interaction.
The SNS influences dendritic cells, which play a crucial role in the innate immune response. Dendritic cells express a variety of adrenergic receptors with alpha1-Rs playing a stimulatory and beta2-Rs an inhibitory effect on dendritic cell migration. It has been shown that beta2-Rs in skin and bone marrow-derived dendritic cells when stimulated respond to norepinephrine (NE) by decreased interleukin-12 (IL-12) and increased anti-inflammatory IL-10 production which in turn downregulates inflammatory cytokine production (Maestroni 2006). The situation can dramatically change in inflammation, where immune cells downregulate their expression of beta2-Rs and upregulate their expression of alpha1-Rs (Heijnen et al. 1996). Alpha1-adrenoceptors stimulate the production and release of proinflammatory cytokines. If alpha1-Rs were to become expressed by the resident or recruited immune cells, then sympathetic activation would be predicted to cause pain and other inflammatory signs via cytokine release (see Fig. 2). Another phenomenon suggesting the involvement of the SNS in allergic responses is the fact that contact sensitizers under certain circumstances inhibit the local norepinephrine turnover in the skin. Thereby, the immune reaction could be amplified (Maestroni 2004). A more indirect contribution of the SNS to inflammation symptoms is the fact that lymph nodes and lymph vessels are innervated by catecholaminergic sympathetic nerves. If the SNS is overactive due to, for example, chronic pain, lymphatic vessel constriction could amplify inflammation related edema (Howarth et al. 1999).
On the other hand every inflammatory condition has a neurogenic component. This means that sensitization and activation of C-fibers by immune mechanisms contributes to the clinical presentation of inflammation. This is another possibility for interaction between the SNS and primary afferents. After activation of nociceptors, which is facilitated by the peripheral immune sensitization, C-fibers release neuropeptides. Since these neuropeptides induce vasodilation, increase skin temperature and sometimes even cause edema, this reaction is called neurogenic inflammation. Particularly, vasodilation and an increase of skin temperature might indirectly increase local sweat rate by increasing sweat gland capacities (Nadel et al. 1971). In addition calcitonin-gene related peptide (CGRP), the most abundant neuropeptide in human skin, has some direct neuronal effects. In a recent study we found that cholinergic sweating was significantly increased by local CGRP in physiological concentrations. Other neuropeptides like vasoactive intestinal peptide or substance P had no such effects (Schlereth et al. 2006). Previous patch clamp studies revealed that the whole molecule CGRP has an inhibitory effect on nicotinic ACh-receptors. However, in human skin CGRP is rapidly degraded by peptidases and smaller fragments of CGRP occur (CGRP1–6, CGRP1–5, or CGRP1–4). These fragments shift the response function of nicotinic ACh-receptors to the left and render them more sensitive (Di Angelantonio et al. 2003). Thus depending on the metabolism of CGRP there is a facilitating or inhibiting effect of activation of nociceptors on function of the SNS.
An Example of Clinical Pain and Sympathetic Disturbances: Complex Regional Pain Syndromes (CRPS)
The SNS is assumed to be involved in the pathophysiology of CRPS because of three phenomena: The patients show the clinical picture of autonomic dysfunction with skin blood flow, temperature, sweating and trophic abnormalities in the affected extremity (Stanton-Hicks et al. 1995); blocking the efferent sympathetic pathways to the painful body region alleviates the pain in CRPS patients (Baron et al. 2002), a symptom called sympathetically maintained pain (SMP). Third, there are repeated reports about painfulness of norepinephrine injections in the skin of CRPS patients; in contrast to healthy people (Ali et al. 2000). The most likely explanation for the latter phenomenon is that at least in some CRPS cases functionally relevant adrenoreceptors may be present on primary nociceptive afferents (Sato and Perl 1991).
Clinical observations in CRPS showed that cutaneous vasomotor disturbances may vary considerably with the duration of CRPS and sometimes even from day-to-day (Sherman et al. 1994; Baron and Maier 1996). In acute CRPS, skin temperature is mostly increased while in chronic CRPS it is almost uniformly decreased on the affected side (Birklein et al. 1998). The increase of skin temperature in acute stages depends on local inflammation (Weber et al. 2001), but also on inhibition of sympathetic vasoconstrictor activity (Janig and Baron 2002). Concentrations of norepinephrine, its intracellular metabolite DHPG (dihydroxyphenylglycol), and neuropeptide Y were reduced in venous blood from the painful side in CRPS patients (Drummond et al. 1991, 1994). Concentrations of adrenaline, which is released from the adrenal medulla, did not differ significantly between the symptomatic and nonsymptomatic limbs (Drummond et al. 1991; Harden et al. 1994). Thus, the release of sympathetic neurotransmitters appears to be compromised in the symptomatic limb. The coincidence with increased sweating despite cold temperature (particularly in chronic CRPS) suggests thermoregulatory disturbances. These thermoregulatory disturbances have been further confirmed and investigated by controlled alterations (warming and cooling) of sympathetic activity in CRPS (Wasner et al. 2001). In CRPS patients, three vascular regulation patterns were found: In the acute stage, the affected limb was warmer during the entire thermoregulatory cycle, in the intermediate type it was either warmer or colder, and in chronic CRPS it was always cooler.
Unilateral sweating abnormalities in CRPS are also very common (Birklein et al. 2000). Two tests were used to localize sudomotor dysfunction along the neuraxis: The thermoregulatory sweat test (TST) activates central thermoreceptors in the ventral hypothalamus (Janig 1990), and sudomotor axon reflex (QSART) tests only the peripheral sudomotor pathways (Low et al. 1983). In acute CRPS both TST and QSART are increased, whereas in chronic CRPS only the TST is increased. The increased peripheral sweating in acute CRPS is explained by local inflammation—as discussed in the previous chapter. Central sweating, however, is increased in chronic CRPS even when peripheral inflammation is no longer prevalent. Such a hyperhidrosis can also be found acutely in stroke and probably reflects some kind of cortical disinhibition of central sudomotor neurons (Riedl et al. 2001).
In order to explain the shift from initially warm inflammatory to subsequent cold CRPS there are once again three possibilities. The first possibility is that cold limbs might be the result of increased sympathetic vasoconstrictor activity in chronic CRPS. However, this is most likely not the case, because bilateral microneurographic recordings in chronic CRPS patients with marked cutaneous vasoconstriction did not show increased discharges in sympathetic neurons (Casale and Elam 1992).
The second explanation is the development of peripheral vasoconstriction (Groeneweg et al. 2006) as a consequence of decreased sympathetic drive and inflammation in acute stages of CRPS. In chronic stages of CRPS adrenergic supersensitivity might overcompensate for an underlying sympathetic deficit in acute stages. Sensitivity to a neurotransmitter increases after an organ has been deprived of that neurotransmitter (Cannon and Rosenbleuth 1949). Multiple mechanisms contribute to this supersensitivity, including a decrease in neurotransmitter reuptake or enzymatic degradation, an increase in post-junctional receptor density, an increase in the binding affinity of agonists to receptors, reduced efficiency of the sodium–potassium pump, and an upregulation of second messenger signalling systems (Fleming and Westfall 1988). The mechanism varies across different types of tissue and also depends, in part, on the cause of neurotransmitter depletion. For example, mechanisms of supersensitivity differ after decentralization, injury to a peripheral nerve, and drug-induced neurotransmitter depletion.
In a pilot study comprising five chronic CRPS I patients and seven controls, quantitative autoradiography revealed an increase of adrenergic receptors in symptomatic tissue in CRPS patients (Drummond et al. 1996). On a functional level the responsiveness of superficial dorsal hand veins in CRPS were found to be more responsive to norepinephrine than those of controls (Arnold et al. 1993). These findings are consistent with the development of adrenergic supersensitivity in response to decreased sympathetic outflow in CRPS.
The third possibility to explain the shift from warm to cold CRPS comes again from the investigation of skin temperature in stroke patients (Riedl et al. 2001). Within 1-day skin temperature decreased on the paretic limb too fast to be mediated by supersensitivity mechanisms. This acute change must be the result of the loss of cortical control on the paretic side. Since CRPS patients also show features, which can also occur after brain damage including a shift of the visual body midline (Sumitani et al. 2007), changes of body scheme and body perception (Moseley 2005) and neglect-like phenomena at the painful side (Frettloh et al. 2006), the loss of cortical control evolving with duration of CRPS might be another reason why autonomic symptoms change in CRPS.
The most compelling evidence of adrenergic involvement in CRPS pain is that injection of adrenergic agonists into the symptomatic limb sometimes provokes or enhances pain, even after the limb has been sympathectomized (Wallin et al. 1976; Torebjork et al. 1995; Ali et al. 2000). These findings are consistent with the clinical observation that CRPS pain increases in cold weather or in response to psychological stress, when catecholamine secretion would be expected to increase (Drummond et al. 2001). This raises the possibility that an adrenergic mechanism aggravates pain in CRPS. Depletion of sympathetic neurotransmitters in the symptomatic limb might not only provoke adrenergic supersensitivity in blood vessels but also the expression of α-adrenergic receptors on nociceptive afferents (Perl 1999). Alternatively, adrenergic supersensitivity that compromises limb blood flow might aggravate tissue inflammation. Metabolic disturbances characteristic of chronic inflammation develop in the symptomatic limb of patients with CRPS. An inadequate blood supply could aggravate the metabolic disorder or delay the washout of nociceptive mediators (e.g., protons) produced during inflammation. Finally, adrenergic supersensitivity might interfere with the normal production and release of nociceptive mediators such as nerve growth factor. Tuttle et al. (Tuttle et al. 1993) demonstrated that α-adrenergic activation increases the secretion of nerve growth factor from cultured vascular smooth muscle. An oversupply of nerve growth factor is likely to sensitize nociceptive afferents directly (Shu and Mendell 1999).
There are many possibilities of how nociception and pain on the one hand, and the sympathetic nervous system and its transmitters on the other hand interact. Physiologically pain activates the SNS, and if this pain becomes chronic neuroplasticity will be responsible for a change of this physiological interaction. The sympathetic disturbances might occur and these disturbances contribute to amplification and prolongation of pain. Since all these changes are usually recognized by the patients causing a deep uncertainty about what is going on in the own body, subsequent development of anxiety will further perturb pain and pain behavior. Unfortunately, we do not know most of the underlying molecular mechanisms for sympathetically maintained pain, since many of them are unique to human consciousness and cannot simply be modeled in animals. Nevertheless, unraveling the molecular targets is a prerequisite for any selective treatment of sympathetically maintained pain.
Supported by the German Research Foundation, DFG Bi 579/1 and Bi 579/4 to F.B.
- Benarroch, E. E. (2006). Pain-autonomic interactions. Neurological Science, 27(Suppl. 2), S130–S133.Google Scholar
- Calder, J. S., Holten, I., & McAllister, R. M. (1998). Evidence for immune system involvement in reflex sympathetic dystrophy. Journal of Hand Surgery [British], 23, 147–150.Google Scholar
- Cannon, W. B., & Rosenbleuth, A. (1949). The supersensitivity of denervated structures. A law of denervation. New York: MacMillan.Google Scholar
- Cortelli, P., & Pierangeli, G. (2003). Chronic pain-autonomic interactions. Neurological Science, 24(Suppl. 2), S68–S70, S68–S70.Google Scholar
- Drummond, P. D., Skipworth, S., & Finch, P. M. (1996). Alpha 1-adrenoceptors in normal and hyperalgesic human skin. Clinical Science (London), 91, 73–77.Google Scholar
- Fleming, W., & Westfall, D. P. (1988). Adaptive supersensitivity. In U. Trendelenberg & N. Weiner (Eds.), Catecholamines. I. Handbook Exp. Pharmacol. (pp. 509–559). Berlin: Springer-VerlagGoogle Scholar
- Heijnen, C. J., Rouppe van der Voort, C., Wulffraat, N., et al. (1996). Functional alpha 1-adrenergic receptors on leukocytes of patients with polyarticular juvenile rheumatoid arthritis. Journal of Neuroimmunology, 71, 223–226.Google Scholar
- Janig, W. (1990). Functions of the sympathetic innervation of the skin. In A. D. Loewy (Ed.), Central regulation of autonomic functions (pp. 334–348). New York: Oxford University Press.Google Scholar
- Manning, B. H. (2004). Preclinical studies of pain modulation: Lessons learned from animals. In D. D. Price & M. C. Bushnell (Eds.), Progress in pain research and management (pp. 43–71). Seattle: IASP Press.Google Scholar
- Schlereth, T., Brosda, N., & Birklein, F. (2005). Spreading of sudomotor axon reflexes in human skin. Neurology, 26, 1417–1421.Google Scholar
- Tarkowski, E., Blomstrand, C., & Tarkowski, A. (1995). Stroke induced lateralization of delayed-type hypersensitivity in the early and chronic phase of the disease: A prospective study. Journal of Clinical & Laboratory Immunology, 46, 73–83.Google Scholar
- Wallin, G., Torebjörk, E., & Hallin, R. (1976). Preliminary observations on the pathophysiology of hyperalgesia in the causalgic pain syndrome. In I. Zotterman (Ed.), Sensory functions of the skin in primates: With special to man. Wenne-Gren Center International Symposium reference Series (pp. 489–502). Oxford: Pergamon Press.Google Scholar