Pain Physiology

1. What is pain?

The "standard" definition of pain is that of the International Association for the Study of Pain:-

"An unpleasant sensory or emotional experience associated with actual or potential tissue damage, or described in terms of such damage. Pain is always subjective. Each individual learns the application of the word through experiences related to injury in early life. It is unquestionably a sensation in a part of the body, but it is also unpleasant, and therefore also an emotional experience. Many people report pain in the absence of tissue damage or any likely pathophysiological cause; usually this happens for psychological reasons. There is no way to distinguish their experience from that due to tissue damage, if we take this subjective report".
(IASP. Pain 1979(6)249-252, ex Shipton, 1993).

This definition is extremely unfortunate. Definitions tend to force people into particular ways of thinking. By concentrating on the subjective nature of pain, this definition allows us to conveniently ignore individuals whose physical findings are all consistent with a diagnosis of "pain", but who cannot relate a subjective feeling of pain. Indeed, it tells us that (appearances to the contrary) such people are not in pain!

Let us consider a few examples:
  • A patient is under anaesthesia, and has a tachycardia and hypertension. We administer analgesia, and both settle. We wake her up and - voila - she reports no sensation or recollection of pain. By the above definition, she was never in pain!

  • Even more tellingly, we now have a fairly good "molecular memory for pain stimulation". The gene c-fos is rapidly expressed in the spinal cord in response to painful (but not to other) peripheral stimuli.

    This is further evidence that we should not rely only on pain being "always subjective". Pain is also something that can now be documented in a fairly objective fashion, albeit only using fancy tools in the laboratory. Clinically it is still, and always will be vitally important to listen to the patient who reports pain. But pain does not stop there. As Cross [1994] has said:

    "The affective-motivational aspects of pain originate in the periphery, and suffering is not merely a matter for the neocortex; it is profoundly more ancient and primitive phylogenetically and is reflected in fibre tracts and neural networks throughout the nervous system".

    Note that c-fos is a proto-oncogene, and is therefore likely to be involved in regulation of major structural and functional changes in the cells where it is induced. We know that volatiles and N2O are not adequate to suppress c-fos expression, and that opiates are not as effective at this task as local or neuraxial block. Some evidence suggests that the latter approaches are more effective in preventing chronic pain states.

    We must also remember that although acute pain may have survival value (causing e.g. removal of the injured limb from a harmful stimulus), chronic pain is of no value whatsoever, and is indeed a major scourge of humanity.

    2. Pain pathways.

    Previously, pain pathways were seen as having three components:-

    1. A first order neurone (cell body in dorsal root ganglion) which transmits pain from a peripheral receptor to ..
    2. A second-order neurone in the dorsal horn of the spinal cord, which axon crosses the midline to ascend in the spinothalamic tract to the thalamus where..
    3. A third-order neurone projects to the postcentral gyrus (via the internal capsule).
    This scenario, while partially correct, is now known to be horribly over-simplified. We will consider the following components of pain pathways:-
    1. Peripheral receptors;
    2. Neural pathways;
    3. Spinal Cord mechanisms & long tracts;
    4. Brainstem, thalamus, cortex & other areas.
    5. Descending pathways.


    a. Peripheral receptors

    We must even here make the important distinction between:

    1. First pain
    2. Second pain
    It is important to realise that there are two distinct responses to a painful stimulus, a "first pain" and a "second pain". The first pain is well-localised and brief, the second is more diffuse and protracted.

    First pain is described as sharp, and "pricking". It localises to a well-defined part of the body surface.The receptors for this first pain are high threshold mechanoreceptors. There appear to be specific "nociceptors" which mediate pain, and ONLY pain.

    Second pain is due to stimulation of receptors that exist in many tissues (but not in, paradoxically, the brain). It is often described as dull (i.e. not sharp) and aching. It is poorly localised. Receptors for this second pain are termed polymodal nociceptors. This pain tends to last beyond the termination of an acute painful stimulus. Sources, pathways, perception of and treatment of the two types of pain are very different. Visceral pain is predominantly of the "second pain" type.

    Visceral pain can however sometimes be referred to a region of the body surface (for example, shoulder tip pain with sub-diaphragmatic irritation). See [Cervero, F. Physiological Rev 1994(74.1) 95-129pp] for a review of the sensory innervation of the viscera.

    There is some evidence that neurotransmitters such as substance P (=sP), vasoactive intestinal polypeptide (VIP) and calcitonin gene-related peptide are important mediators, either as neurotransmitters, or sensitisers of visceral pain receptors. Prostaglandins, histamine, serotonin, bradykinin, ATP, potassium, and H+ ions also appear important in this regard, especially serotonin, which appears to act mainly on 5HT3 receptors.

    In terms of pain perception, thresholds for feeling pain are remarkably constant from individual to individual. i.e. Peripheral receptor stimulation of sufficient intensity will reproducibly cause pain at the same level in most people. The response of the individual, and his tolerance of the pain, will however differ markedly between individuals.

    Of great interest is "Neurogenic Inflammation". Here, stimulation of C fibres causes a local reaction consisting of vasodilatation and increased capillary permeability. This is due to retrograde transport and local release of sP and calcitonin gene-related peptide. As a consequence, K+, H+, acetylcholine, histamine and bradykinin may be released, and these in turn cause prostaglandin and leukotriene production (which may end up sensitizing high-threshold mechanoreceptors)! Neurogenic inflammation may spread to surrounding tissues antidromically!!

    Analgesic drugs that act peripherally include non-steroidal anti-inflammatory agents, corticosteroids, local anaesthetic agents (which may theoretically inhibit neurogenic inflammation if given early enough, an area of great controversy), and even novel drugs such as substance P antagonists (One such antagonist that does NOT appear to work very well is capsaicin, but opioids, serotonin antagonists, baclofen and clonidine may also inhibit sP release). Of note is the recent identification of two different types of cyclo-oxygenase, with the potential for developing more specific non-steroidal anti-inflammatories, with (perhaps) fewer side-effects.


    b. Neural pathways.

    "First pain" responses are conveyed from the periphery to the dorsal horn of the spinal cord in small myelinated fibres (A delta) while "second" pain is conveyed in non-myelinated C fibres. This is important, especially when considering the "gate control theory" detailed below. Also of importance to this theory are afferent stimuli coming in in large myelinated fibres (A beta fibres), from peripheral vibration / pressure / touch receptors.

    Neurogenic pain, originating in damaged or abnormal C fibers, may respond to membrane- stabilizing drugs such as anticonvulsants (e.g. carbamazepine).


    c. Spinal cord pathways.

    These are complex. We will consider:

    1. initial connections : laminae in the spinal grey matter close to where the fibres enter the spinal cord.
    2. Local interconnections.
    3. Ascending pathways and descending (control) pathways are considered much later.

    1. Spinal pathways: initial connections.
    About 70% of pain fibres enter in the dorsal root, but the rest double back and enter the ventral (so called "motor" root). The grey matter of the spinal cord has ten "laminae" or layers. The important ones that we must consider are:

  • Lamina I - marginal zone
  • Lamina II - substantia gelatinosa
  • Lamina V - part of "nucleus proprius of dorsal horn" (with IV,VI)
  • Laminae VII and VIII - intermediate spinal grey matter (with X)

    Unmyelinated C fibres synapse in laminae I to V, while A delta fibres synapse in laminae I, V and X. These different routes are only the start of dramatically different pathways! The two main pathways are:-

    1. The primitive, spino-reticulo-diencephalic tract (otherwise known to its friends as the "palaeospinothalamic tract"). Impulses pass from type C fibres to several different "second order" neurons.
    2. The more modern neospinothalamic tract (which also answers to the name "anterolateral white funiculus"). This takes most of its fibres from Laminae I and V, and appears to mediate "first" pain;
    Important second order neurons are
    1. nerve cells that are not committed to responding only to pain, responding also to gentle stimuli and increasing their response as the stimulus becomes more marked. This is the Wide Dynamic Range cell (WDR cell). WDR cells are found in lamina V (and also laminae IV and VI)
    2. Nociceptive specific neurones found in lamina I, responding only to noxious stimuli;
    3. Complex neurons which receive numerous inputs and have been poorly studied. They are seen mainly in laminae VII and VIII;

    An important property of WDR neurons is "wind up". This occurs with repetitive stimulation at about once per second via C fibres - each added stimulus increases the response of the WDR cell. Wind-up is seen as a good justification for treating pain early and well! Wind up may be related to stimulation of glutamate receptors (especially the N-methyl D Aspartate or "NMDA" receptor, a receptor which is very topical at present)! Many NMDA antagonists exist, but most have significant side effects (e.g. ketamine).

    Clinically, evidence for wind-up has been hard to come by, although some recent studies attest to the value of pre-emptive analgesia. See, for example, Mansfield(1996). Some evidence suggests that mechanisms other than wind-up may be important in the development of visceral pain, although the NMDA receptor appears to still be a major player [J Physiol(Lond) Dec 1995 489(2) 545-55pp].

    The spino-reticulo-diencephalic system is rich in opiate receptors, while the spinothalamic tract has very few. This explains why opiates (eg. morphine) have good analgesic properties for visceral pain without affecting e.g. pinprick sensation.

    2. Spinal pathways: local interconnections. There are several. Of great importance are connections mediating so-called "gating". The basic idea here is that "painful stimuli" coming into the cord on C fibres can be modified by other inputs, which "close the gate on the incoming pain". These inputs come from:

    This arrangement has several practical consequences:
    1. Transcutaneous Electrical Nerve Stimulation (TENS) works
    2. Dorsal column stimulation (DCS) works
    3. Acupuncture works!
    4. Rubbing the skin locally helps decrease pain!
    TENS works by high-frequency, low amplitude stimulation of larger peripheral fibres, and this inhibits transmission of pain through the "gates". DCS is similar, but works by antidromically stimulating the same A beta fibres. Acupuncture works in a different way: it causes low-frequency high amplitude stimulation of small A delta fibres (amongst other fibres), and this also causes inhibition of pain through gating mechanisms. Another difference between acupuncture and TENS is that the effect of acupuncture can be blocked by giving opioid antagonists, which don't inhibit the effect of TENS.

    Mechanisms of inhibition in the dorsal horn may be very complex - incoming nociceptive fibres express opioid receptors, opioids here blocking sP release, and they probably also act post- synaptically. The presence of these opioid receptors at least partially explains the potent effect of small doses of intrathecal opiates. In addition, GABA, somatostatin, neurotensin, CCK and neuropeptide Y may play a role in pain transmission.

    A wide variety of drugs has been used for "antinociception" at the spinal level, including opioids, alpha-2 agonists, and local anaesthetics. Agents such as ketamine and somatostatin may however be neurotoxic. Of interest is the potent analgesia induced by spinal octreotide! [Rawal N, Ann Med Apr 1995 27(2) 263-8pp].


    d. Higher ascending pathways.

    1. The old spino-reticulo-diencephalic pathway does just that - it mainly ends in the reticular system of the brainstem, but also sends fibres to the thalamus (the medial nuclei of the thalamus). Probably important are connections between the reticular system and the hypothalamus (and thalamus) - these may explain autonomic components of the pain response. `Emotional'/affective responses to pain may be explained by projections that go from the medial nuclei of the thalamus to most of the cortex, especially the frontal areas, and notably to the anterior cingulate gyrus! (In the bad old days people sometimes did prefrontal lobotomies for intractable pain)! The basal ganglia may also be involved in pain discrimination, the affective component of pain, and even in pain modulation [Chudler E & Dong W, Pain 1995 60(1) 3-38pp].

    2. The fresh, new "spinothalamic tract" nips up to the ventrobasal part of the lateral thalamus. Connections go from here to the sensory cortex (postcentral gyrus), which explains the precise localisation of somatic pain. It is obvious that this tract is NOT the main pain pathway, because lesions along this pathway do not cancel out the sensation of pain, and, in fact, may cause severe pain (the "thalamic syndrome" - possibly due to damage to inhibitory pathways).

    Positron emission tomography (PET) has been used to study the response to pain. Acute traumatic nociceptive pain appears to activate the hypothalamus and PAG, but a whole array of other areas also become involved (including the prefrontal cortex, insula, anterior cingulate cortex, posterior parietal cortex, primary motor and somatosensory areas, supplemental motor area and even the cerebellum. Further demonstration of the complex nature of pain. [Pain Feb 1996 64(2) 303-14pp].


    e. Descending pathways.

    Descending modulation of pain sensation originates from three main areas:

    1. Fibres pass from PAG to the reticular formation of the medulla (the nucleus raphe magnus or "NRM", and the closely associated nucleus reticularis gigantocellularis pars alpha, and nucleus reticularis paragigantocellularis, all together called the ventromedian medulla or "VMM") where connections are serotoninergic, and from there axons descend in the "dorsolateral funiculus" of the spinal cord, to end up (surprise, surprise) on interneurones right next to the substantia gelatinosa (lamina II) in the cord. The synapses here are enkephalinergic. Stimulation of this system causes inhibition of incoming pain impulses. Thus, although serotonin applied peripherally augments pain, its action centrally is important in descending inhibition of incoming painful impulses! New evidence suggests that GABA is also important in inhibition of pain pathways by the VMM [Neuroscience Jul 1996 73(2) 509-18pp].

    2. To make matters more complex, quite apart from the above inhibitory pathway that works mainly on opiates, there is a separate, noradrenaline-based pathway that inhibits pain. Noradrenaline-based projections appear to come from the NRM, and also from the locus coeruleus in the pons. This is one explanation why antidepressants (which inhibit noradrenaline uptake) may be effective in controlling pain, but remember that antidepressants also inhibit serotonin uptake, so they may also increase activity in the fibres that pass from PAG to medulla!

    3. "On" and "Off" cells:
      There are probably two main classes of neurones in the PAG and VMM:
      • "On" cells increase pain transmission;
      • "Off" cells decrease it.
      Opioids inhibit "on" neurones and increase transmission in "off" neurones, and noradrenaline also inhibits "on" neurons.

    3. The response to pain.

    Responses to visceral pain are very different from those evoked by somatic pain. Visceral pain generally results in tonic muscular spasm (teleologically, to decrease movement of the affected area) while somatic pain usually causes withdrawal of the affected part of the body ("to protect this region from further damage"). As already mentioned, the sensations reported for the two pains are also quite different.

    We are all also aware that pain (be it somatic or visceral) can have profound autonomic effects. Some of the reasons for this have been alluded to: there is a good degree of cross-over between the somatic and visceral systems, notably at the level of the WDR cell and "complex neurone" in the spinal cord, but also extensively at higher centres, with projections to, for example, the hypothalamus. Also of note is the close relationship between sensory afferents and sympathetic outflow (See Cross, 1994). Recent studies suggest that epidural fentanyl alone blocks the neurohormonal response to surgery (See Harukuni, I et al. anesth Analg. 1995(81)1169)!


    4. A brief note on opiates and their receptors

    There are three main types of opioid receptor: delta, kappa and mu. All of these are widely distributed in the brain, and are not only concerned with modulation of pain perception, but also with a variety of other functions. This explains why in trying to control pain, we encounter many unwanted side-effects. For example, mu receptors are widespread in the brainstem parabrachial nuclei (where stimulation of them causes respiratory depression), and dependence may be related to receptors in the locus coeruleus and ventral tegmentum. Some have asserted that 1 and 2 receptors are mainly concerned with pain and respiratory depression, respectively, but this is probably too simplistic.

    Opioids work in two main ways: they either block neurotransmitter release (by inhibiting calcium influx into the presynaptic terminal), or hyperpolarise neurones by opening a potassium channel (and therefore effectively temporarily knock the neurone out of action)!

    Corresponding (more-or-less) to the three receptor classes, there are three main groups of endogenous opioids: the enkephalins, endorphins and dynorphins. Enkephalins act mainly on delta receptors, dynorphins on kappa receptors, while beta endorphin acts on both mu and delta receptors. Morphine is almost exclusively a mu agonist. Mu and delta receptors appear to be expressed on the same cells, and stimulation of one receptor increases the affinity of the other! Beta endorphin is also a "neurohormone" in that it may have distant effects on neurones with mu receptors, after being released in the hypothalamus.

    The effect of opioids on the "on" and "off" cells in the medulla is most interesting: ONLY the "on" cell is affected by opioids - it is inhibited, probably by hyperpolarization due to opening of potassium channels. The "off" cell certainly becomes more active in the presence of opioids, but this is only secondary to the lessened activity of the "on" cell. The "on" cell is also inhibited by the alpha-2 agonist clonidine. In fact, in some cells the mu receptor and the alpha-2 receptor are linked to the same G protein and the same potassium channel! Of even more interest is evidence that by giving morphine we can activate endogenous circuitry that turns on enkephalin release! See [Fields, Ann Neurology, 1994 35 S42-45pp].

    The roles of the various types of opiate receptor have recently been clarified by studies in opiate receptor knockout mice. Mu-receptor knockout mice had shorter latencies on tail-flick and hot plate tests, than did wild-type mice, and morphine did not reduce responses to pain. These studies support the contention that the mu receptor is the main player in morphine-induced analgesia. [Proc Natl Acad Sci U S A Feb 1997 94(4) 1544-9pp]



    5. References & Further Reading

    1. Richardson, BP. Ann NY Acad Sci 1990 (600) 511-519. Serotonin and Nociception.

    2. Cross, SA. Mayo Clin Proc1994 (69) 375-383. Pathophysiology of Pain. This article on pain mechanisms is still by far the best summary we have encountered!

    3. Helme, RD. et al. Medical J. Australia 1990(153)400-406. Neural Pathways in Chronic Pain.

    4. Mansfield MD. et al. BJA. 1996 (76) 358-361. Influence of dose and timing of administration of morphine on postoperative pain and analgesic requirements.

    5. Editorial. BJA. Jan 1996 76(1) 1-4pp.

      An article on the significance of c-fos. Well worth reading.

    6. Shipton, EA. Pain, Acute & Chronic. W.U.P.1993 (ISBN 1-86814-241-8). Reading this book {prescribed at Wits} gives one a great understanding of pain.

    7. Hamill, RJ & Rowlingson, JC. Handbook of Critical Care Pain Management. McGraw-Hill. 1994. (ISBN 0-07-025814-7).