Nociception is a signal.
Pain is what the nervous system makes of it.

1. Pain is an experience; nociception is a neural process. Each can occur without the other.
2. Aδ is fast, sharp and well localised: first pain. C is slow, dull and diffuse: second pain.
3. The dorsal root ganglion holds the first-order cell body; the dorsal horn holds the second. The second-order axon crosses within one or two segments.
4. Modulation happens at the first synapse. Large-fibre input and descending fibres from the periaqueductal grey, rostral ventromedial medulla and locus coeruleus inhibit transmission, presynaptically and postsynaptically.
5. Injury turns the signal up. Peripherally, mediators lower the nociceptor’s threshold; centrally, the NMDA receptor raises the dorsal horn’s gain.
6. Pain is a whole-body event. Segmental reflexes, the sympathetic system and the neuroendocrine stress response are all driven by the same input.

A nociceptor is a peripherally located neuron preferentially sensitive to a damaging or potentially damaging stimulus. It is pseudounipolar: its cell body lies in the dorsal root ganglion, or the trigeminal ganglion for the face, with one process running to the tissue and one into the central nervous system. Its receptive endings are free, unencapsulated nerve endings in skin, muscle, joints, connective tissue, vessel walls and the thoracic and abdominal viscera.

Heat thresholds are quoted slightly differently depending on what is described: TRPV1 opens above about 43 °C, and thermal nociceptors are activated by skin temperatures above about 42 °C.

The two pains are separated by conduction alone. Over one metre of limb, an Aδ volley arrives in 33 to 83 ms; a C volley takes 0.5 to 2.0 s. That gap is the pause between the sharp, precisely placed pain of a stubbed toe and the spreading ache that follows it. Published Aδ velocities vary between textbooks, from about 5 to 35 m/s; the ranges here are the fibre classification used throughout this module.

The central process enters through the dorsal root and divides in Lissauer’s tract, running 2 to 3 segments up and down the cord and giving collaterals into the dorsal horn along the way, which spreads one input across several segments. The grey matter is divided into ten Rexed laminae: lamina I is the most superficial, the dorsal horn extends to lamina VI, the ventral horn is VII to IX, and X surrounds the central canal.

Two classes of second-order neuron receive this input. Nociceptive-specific (high-threshold) neurons respond only to noxious stimuli. Wide dynamic range (convergent) neurons, mostly in deeper laminae, respond to noxious and non-noxious input alike. A wide dynamic range neuron normally does not signal pain in response to touch; once it is sensitised it may discharge at a high rate to a tactile stimulus, and if that exceeds a threshold the touch is felt as pain. Allodynia is therefore a dorsal horn phenomenon.

The second-order axon crosses in the ventral (anterior) white commissure within one or two segments of its origin and ascends in the anterolateral quadrant of the opposite side; 85 to 90% of spinothalamic cells project contralaterally. That crossing level is why damage to one side of the cord removes pain and temperature sensation from the opposite side, beginning one or two segments below the lesion.

Perception

Functional imaging shows a network, the pain matrix, active during acute pain, rather than a single centre. Broadly, the somatosensory cortices serve the sensory-discriminative component (where, how intense, how long) and the limbic and paralimbic regions the motivational-affective one (how unpleasant).

Two observations show that the components can be separated: damage to the prefrontal cortex reduces the ability to judge how severe a pain is, and cingulotomy reduces its emotional component. The anterior cingulate and prefrontal cortex also project back to the periaqueductal grey, which is where the descending system of section 08 begins.

Melzack and Wall proposed in 1965 that transmission from the primary afferent to the projection neuron is gated by inhibitory interneurons in the substantia gelatinosa, and that the position of the gate is set by the balance of large- and small-fibre activity. Two predictions can be read off the circuit.

What has lasted of the theory is its central claim: the first synapse is modulated, not a fixed relay. Later work has described how.

How inhibition works at the synapse

GABA and glycine provide tonic inhibition of nociceptive input, and losing it produces features of neuropathic pain such as allodynia. α-adrenoceptors in the cord are analgesic when activated, whether by noradrenaline from descending fibres or by clonidine given spinally, and their effect is synergistic with that of opioids.

Electrical stimulation of the PAG produces more profound analgesia than stimulation of the nucleus raphe magnus or the locus coeruleus, and naloxone reverses it: endogenous opioids are part of the circuit. Recordings in the rostral ventromedial medulla find two populations, off cells, which pause just before a withdrawal response and inhibit dorsal horn transmission, and on cells, which fire just before it and facilitate transmission. Descending control is modulation, not only inhibition.

The endogenous opioids

Endomorphin-1 and endomorphin-2 are short peptides highly selective for the µ receptor. Nociceptin acts at the fourth member of the family and was named because, unlike the others, it lowered the pain threshold under some conditions; in the rostral ventromedial medulla it inhibits both the facilitating and the inhibiting neurons, so its net effect depends on the state the circuit is already in.

Opioids also act outside the central nervous system. Opioid receptors are made in the dorsal root ganglion and transported to the peripheral terminals, and in inflamed tissue immune cells release opioid peptides onto them: a peripheral gate at the site of injury.

Modulation by stress, expectation and mood

Stress-induced analgesia, the wounded soldier who feels little until the fighting stops, depends on this system: β-endorphin shares its precursor with ACTH, and mice unable to make β-endorphin lose naloxone-reversible stress-induced analgesia while keeping their response to morphine. Placebo analgesia acts on the same brain regions as an opioid, presumably by releasing endogenous opioids. Attention, expectation, anxiety and mood reach the circuit through the cortical projections to the PAG, which is part of why the same injury hurts differently in different people.

Inflammatory mediators, acting through G-protein-coupled receptors, second messengers and kinases that phosphorylate the terminal’s ion channels, change the nociceptor in three measurable ways: its threshold for activation falls, it discharges faster when activated, and it discharges spontaneously. Silent nociceptors wake, so more afferents fire as well as each one firing more. Nerve growth factor, made by fibroblasts, Schwann cells and keratinocytes and acting through trkA receptors, alters gene transcription and increases the synthesis of neuropeptides and ion channels.

The result is primary hyperalgesia, confined to the injured tissue: low-intensity mechanical stimuli now hurt, and so does heat. Hold on to the heat, because the zone around the injury behaves differently (section 10).

Neurogenic inflammation: a sensory nerve acting outwards

An impulse started in one branch of a nociceptor travels centrally and also antidromically into the fibre’s other branches. There it releases substance P, neurokinin A and CGRP, which cause vasodilatation, plasma extravasation and mast cell degranulation. Histamine from mast cells and serotonin from platelets then activate and sensitise neighbouring nociceptors, so the sensitised area spreads beyond the first terminal.

A barrage of nociceptive input, of the kind surgery produces, changes how dorsal horn neurons respond. Central sensitisation is the result: persistent changes in the central nervous system after injury that result in pain hypersensitivity, with exaggerated and prolonged responses to normal afferent input. The relationship between stimulus and response stops being fixed.

In the zone of secondary hyperalgesia there is increased sensitivity to innocuous mechanical stimuli but no change in the threshold to heat. That follows from where the change is. The skin in that zone was never injured, so its nociceptors are normal; what has changed is the dorsal horn neurons that receive its mechanical input.

Wind-up and the NMDA receptor

Wind-up lasts over the course of a train of inputs. Other forms of central sensitisation outlast the stimulus that started them, and the lasting strengthening of synaptic transmission after activity, long-term potentiation, shares its physiology with the process of memory and may matter more than wind-up for the development of chronic pain. Central sensitisation also depends on continuing peripheral input, for its maintenance as well as its induction, which is the argument for treating the periphery even when pain is established.

Most visceral nociceptors are latent. As few as 3% respond to intraluminal pressure in the physiological range, and inflammation raises that about 4-fold, so a relatively minor stimulus to an inflamed organ causes severe pain. Visceral pain also shows spatial summation: stimulating a larger area lowers the threshold.

Referred pain

Visceral and somatic afferents from the same segment converge on the same second-order neurons in the dorsal horn, which project on to the thalamus and cortex. The brain attributes the activity to the skin it usually comes from. Pain is therefore referred to a structure that developed from the same embryonic segment as the organ.

Labour

Labour pain is a worked example of segmental innervation. Adequate analgesia in the first stage needs block of T10 to L1; in the second stage it must extend to S2 to S4. The blocks that work locate the routes: a paracervical block, beside the cervix, relieves first-stage pain, and a block of the pudendal nerve, from S2 to S4, relieves pain in the second stage. How much labour hurts varies widely, with parity, the size and shape of the pelvis, fetal size and presentation, anxiety, support in labour, and whether contractions are induced or augmented.

Neuropathic pain is relatively resistant to opioids. One proposed reason is cholecystokinin, which antagonises µ agonists at the receptor without being analgesic itself, and which rises in neuropathic pain and falls in inflammatory pain; loss of presynaptic opioid receptors may add to it. In neuropathic pain, morphine may also increase NMDA receptor activity while losing efficacy at the µ receptor.

Complex regional pain syndrome follows trauma, with or without a nerve lesion (type I after an initiating noxious event, type II after nerve injury), and combines spontaneous burning pain, hyperalgesia and allodynia with vasomotor, sudomotor and trophic change. Phantom sensations, usually painful, occur in 50 to 80% of amputees.

From acute to chronic

The line between acute and chronic pain is less clean than a 3-month definition suggests. A noxious stimulus induces new gene expression in the dorsal horn within 1 hour, enough to change behaviour in the same period, and the intensity of acute postoperative pain predicts chronic postoperative pain. Chronic persistent postsurgical pain follows 10 to 65% of operations, depending on the procedure, and is severe in 2 to 10%.

The predictors map onto this lesson: severe acute pain, the area of postoperative hyperalgesia, and pain before surgery, which may already have sensitised the central nervous system. Activated microglia in the dorsal horn take part in the transition, and so do psychological factors. Chronic pain is best understood through the biopsychosocial model, in which biological change, cognition and memory, and social reinforcement interact, which is why treating only one of them rarely works.

The same response that once had a protective purpose can harm a patient after modern surgery: more myocardial work at a time of higher demand, thrombosis, pulmonary complications and delayed recovery. Reducing the nociceptive input reduces the stress response, the sympathetic outflow and the inhibitory spinal reflexes together, which is why effective analgesia is part of recovery rather than only comfort.

Read down the table and multimodal analgesia is an argument rather than a habit: drugs that act by different mechanisms on different receptors along the pathway combine for a greater, synergistic effect, whereas two drugs acting at the same step mostly combine their side effects. Covering the whole perioperative period, rather than a moment before incision, is what turns that into preventive analgesia (section 12).