SENSORY TRANSMISSION IN THE SPINAL CORD Morphology of the spinal cord dorsal horn
The spinal cord is arranged in such a way that primary afferents originating from different regions of the body display specific somatotopic organisations upon entry into the cord. Hence in any given segment, there is a definite laterality (ipsilateral/ contralateral) and a three-dimensional organisation (rostrocaudal, mediolateral, dorsoventral) of the afferent terminations.
The spinal cord is classically divided into white and grey matter (Fig. 21.2). The grey matter can be organised into ten different laminae, which run continuously along the entire length of the spinal cord. Within a given section of a spinal cord, each lamina can be seen as a layer of functionally distinct cells. Laminae I to VI comprise the dorsal horn, laminae VII to IX the ventral horn, and lamina X is the substantia grisea centralis which surrounds the central canal.
Lamina I forms the outer layer of the dorsal horn and contains the large marginal cells of Waldeyer and plays an important role in nociception since it is the layer in which
some nociceptive afferents terminate. It contains a number of cell types, including nociceptive-specific neurons, which received A¿- and C-fibre input, neurons which only respond to innocuous thermal stimuli and wide dynamic range (WDR) neurons. Many marginal cells appear to be projection neurons, which contribute to the lateral spino-cervical (SCT), spinoreticular and spinothalamic tracts (STT). The projections also extend to the periaquaductal grey (PAG), parabrachial nucleus and the nucleus submedius.
Lamina II is also known as the substantia gelatinosa (SG) and can be divided into two layers, the outer layer (Ilo) and the inner layer (IIi). This layer is densely packed with small neurons and lacks myelinated axons. Neurons with cell bodies in IIi receive inputs from low-threshold mechanoreceptive primary afferents, while those in IIo respond to inputs from high-threshold and thermoreceptive afferents. The intrinsic cells which comprise the SG are predominantly stalk and islet cells. Stalk cells are found located in lamina IIo, particularly on the border of lamina I, and most of their axons have ramifications in lamina I although some also project to deeper layers. These cells are thought to predominantly relay excitatory transmission. Islet cells, on the other hand, are located in IIi and have been demonstrated to contain the inhibitory neuro-transmitters, y-aminobutyric acid (GABA), glycine and enkephalins in their dendrites. Hence these cells have been proposed to be inhibitory interneurons.
The cell bodies in lamina III are generally larger and less densely packed than those in the substantia gelatinosa. The main cell type of lamina III includes projection cells, which contribute to the SCT and postsynaptic dorsal column (PSDC). The dendrites of SCT cells are confined to lamina III and do not reach laminae I and IIo. However, those of PSDC are not flattened in the mediolateral plane and extend to laminae I and II, thus forming monosynaptic connections with small primary afferent fibres.
Lamina IV is composed of heterogeneous sized cells and is less densely packed than lamina III due to the number of nerve axons passing in this layer. At least three types of neurons have been identified in lamina IV, based on different dendritic projection patterns and these include SCT and PSDC cells. Another cell type has been described which has a dendritic pattern similar to SCT and PSDC, but with local axon terminations. Somas of STT cells are also found in lamina IV.
The cells comprising lamina V are more diverse than those of lamina IV and their dendrites extend vertically toward the superficial layers. Cell bodies in lamina V contribute to three projection pathways, the SCT, PSDC and STT. However, the STT cells appear to be predominant in this lamina. Lamina V plays an important role in nociception since it receives both A£- and C-fibre inputs. Some cells in lamina V also respond to cutaneous low- and high-threshold mechanical stimuli and receive nociceptive inputs from the viscerae. Many of these neurons also project onto mono-neurons and so act as interneurons in the polysynaptic withdrawal reflex to noxious stimuli.
Lamina VI forms the base of the dorsal horn and can be found only in certain levels of the spinal cord, the cervical and lumbar regions. Few data have been reported on the cell composition of lamina VI. Cells of lamina VI are small compared to those of lamina V and some axons appear to contribute to the STT and SCT pathways.
Nociceptive sensory information arriving from primary afferent fibres enters via the dorsal horn and on entering the spinal cord undergoes considerable convergence and modulation. The spinal cord is an important site at which the various incoming nociceptive signalling systems undergo convergence and modulation and is under ongoing control by peripheral inputs, interneurons and descending controls. One consequence of this modulation is that the relationship between stimulus and response to pain is not always straightforward. The response of output cells could be greatly altered via the interaction of various neurotransmitter systems in the spinal cord, all of which are subject to plasticity and alterations during pathological conditions.
The arrival of action potentials in the dorsal horn of the spinal cord, carrying the sensory information either from nociceptors in inflammation or generated both from nociceptors and intrinsically after nerve damage, produces a complex response to pain. Densely packed neurons, containing most of the channels, transmitters and receptors found anywhere in the CNS, are present in the zones where the C-fibres terminate and while excitatory mechanisms are of importance, the role of controlling inhibitory transmitter systems is perhaps paramount.
Since glutamate is the main excitatory neurotransmitter in the CNS it is not unexpected to find that the vast majority of primary afferents synapsing in the dorsal horn of the spinal cord, regardless of whether they are small or large diameter, utilise this transmitter. It has an excitatory effect on a number of receptors found on both postsynaptic spinal neurons, leading to a depolarisation via three distinct receptor subclasses, the a-amino-3-hydroxy 5-methyl-4-isoxazeloproprionic acid (AMPA) receptor, the N-methyl-D-aspartate (NMDA) receptors and the G-protein-linked meta-botropic family of receptors. In addition, presynaptic kainate receptors for glutamate have been described in the spinal cord. Most is known about the first two receptors, the AMPA and NMDA receptors, named after chemical analogues of glutamate with selective actions on these sites (see Chapter 11).
Glutamate is released in response to both acute and more persistent noxious stimuli and it is fast AMPA-receptor activation that is responsible for setting the initial baseline level of activity in responses to both noxious inputs and tactile stimuli. However, if a repetitive and high-frequency stimulation of C-fibres occurs there is then an amplification and prolongation of the response of spinal dorsal horn neurons, so-called wind-up (Fig. 21.3). This enhanced activity results from the activation of the NMDA-receptor. If there are only acute or low-frequency noxious or tactile inputs to the spinal cord the activation of the NMDA-receptor is not possible. The reason is that under normal physiological conditions the ion channel of this receptor is blocked by the normal levels of Mg2+ found in nervous tissues. This unique Mg2+ plug of the channel requires a repeated depolarisation of the membrane to be removed and allows the NMDA receptor-channel to be activated. Here it is likely that the co-release of the peptides such as substance P and CGRP that are found in C-fibres with glutamate is responsible for a prolonged slow depolarisation of the neurons and subsequent removal of the block. Not only do AMPA receptor antagonists have no effect on wind-up but the brief depolarisation produced by this receptor would not be expected to produce any prolonged removal of the block, unlike the long-lasting slow (several seconds) activations caused by peptides. The lack of peptides in large AS afferent fibres explains the lack of wind-up after low-threshold stimuli. This NMDA receptor activation has been clearly shown to play a key role in the hyperalgesia and enhancement of pain signalling seen in more persistent pain states including inflammation and neuropathic conditions.
There are a number of antagonists at the multiple regulatory sites found on the NMDA receptor and its channel, including the licensed drugs, ketamine, a potent channel blocker, and the weaker agents, dextromethorphan and memantine. These drugs have been shown to be antinociceptive in a number of animal models of inflammation and nerve damage and there are also data from volunteer and clinical studies to support this. Overall, these studies indicate that it is likely that aberrant peripheral activity is amplified and enhanced by NMDA-receptor-mediated spinal mechanisms in tissue damage and neuropathic pain and that the receptor is critical for both the induction and maintenance of the pain. Thus, therapy after the initiating damage can still be effective. Although there is much good clinical evidence for the effectiveness of agents acting as antagonists at the NMDA-receptor complex, especially ketamine, and although some individual patients get good pain relief in nerve injury situations, the majority cannot achieve complete pain control. This is partly because adequate dosing is prevented by the narrow therapeutic window of the existing drugs.
Ultimately, the broad use of glutamate receptor/channel antagonists in the treatment of pain will therefore depend on strategies that increase their therapeutic window over existing drugs. These may include drugs acting on subtypes of the receptor (NR2B receptor antagonists are analgesic but side-effects have not been fully evaluated), drugs with different use-dependent block of the channel or more practically, use of low-dose NMDA blockers in combination with another agent.
As neurons become more active, then ion channels, other than sodium channels, open in their membranes. There are a number of voltage-operated calcium channels (see Chapter 3) that are critical for both transmitter release and neuronal excitability. Successful results in animals with agents that block neuronal voltage-sensitive calcium channels would also suggest that there is an increase in central neuronal excitability after both inflammation and nerve damage. N-type channels, blocked by rn-conotoxin, a marine snail toxin, have been shown to play a key role in behavioural allodynia and the neuronal responses to low- and high-threshold natural stimuli after nerve damage, and in the C-fibre-evoked central hyperexcitability that follows inflammation. Blockers of this channel (SNX-111 or rn-conotoxin) are considerably more effective after nerve injury (spinal nerve ligation) and since the channel is voltage operated then these results again suggest increased excitability of the spinal cord after injury. Less is known about P-type channels but rn-agatoxin GVIA, a selective blocker, is effective against persistent inflammatory inputs through central spinal actions. Unfortunately, since calcium channels are extensively distributed in all excitable tissue it is necessary to give blockers used for analgesia by the spinal route.
Gabepentin is an antiepileptic drug that has analgesic activity in neuropathic pain states from varying origins. Two recent randomised controlled trials of gabapentin in patients, one group with postherpetic neuralgia and another with diabetic neuropathy, concluded that gabapentin was effective in the treatment of these pain states. It has also been reported that gabapentin is effective in pain due to peripheral nerve injury and central lesions, with particular effectiveness on paroxysmal pain and allodynia. How gabapentin works is not clearly established but it is thought the drug may interact with calcium channels in that it becomes attached to the so-called gabapentin-binding protein, itself associated with a subunit of the calcium channel. This action would fit with the evidence that N-type calcium channel blockers are more effective in reducing behavioural and electrophysiological responses to sensory stimuli after both nerve injury and tissue damage, conditions where it appears that N-type calcium channels are upregulated.
The influx of calcium through activation of the NMDA channel and also voltage-operated calcium channels may be a mechanism through which further profound changes in nociceptive processing occur. Rises in internal calcium in neurons is a key means by which genes can be activated. The protooncogene markers c-fos and c-jun can be observed in dorsal horn neurons only minutes after the application of noxious stimulation, either mechanical or thermal or from tissue damage. The one functional piece of evidence at present for the consequences of gene induction is the increase in the mRNA and dynorphin production in some dorsal horn cells, although the physiological consequences of this are unknown.
A comparatively new putative nociceptive transmitter is the gas nitric oxide (NO), and many studies have provided much indirect evidence for a spinal role of this gas during prolonged nociceptive events. NO therefore appears to have a role during prolonged chronic pain states which have been associated with NMDA-receptor activation. It has been proposed that NMDA-receptor activation and the associated Ca2+ influx results in the generation of NO by activation of the enzyme, nitric oxide synthase (NOS). The NOS antagonist, L-NAME, abolishes hyperalgesia in neuropathic animals, reduces pain-related behaviour after inflammation and blockers of the production of NO prevent wind-up. One proposed action of NO is as a retrograde transmitter feeding back from spinal neurons onto presynaptic sites to further increase transmitter release from C-fibres. The synthesis of inhibitors of the neuronal version of NOS which lack hypertensive effects yet are antinociceptive suggests possible therapeutic uses of NOS inhibitors.
This positive feedback may also be due to the spinal generation of prostanoids, following both NMDA- and substance P-induced activation of neurons. It is now recognised that in addition to the well-documented production of prostaglandins in peripheral tissues there can be central neuronal synthesis, again with calcium being the trigger. It is not yet known how important this central action is to the analgesic effects of systemic NSAIDs but, as mentioned earlier, COX-2 is constitutive in the spinal cord and further upregulated by peripheral inflammation.
There are important inhibitory systems built into the control of events following C-fibre stimulation. Thus, during peripheral noxious stimulation, spinal mechanism, driven by NMDA-receptor-mediated activity, can become active to damp down further neuronal responses, the purine, adenosine (see Chapter 13), appears to be involved in this type of control and has been reported to be effective in humans with neuropathic pain. It is thought that the depolarisations caused by activation of the NMDA receptor increase the metabolic demand on neurons and so ATP utilisation increases. ATP then is metabolised to adenosene and the purine then acts on its inhibitory A1 receptor in the spinal cord to reduce further neuronal activity — a negative feedback. Thus there are potential indirect targets for the control of NMDA events. These transmitter systems are summarised in Fig. 21.4.
y-Amino butyric acid (GABA) has been firmly established as the major inhibitory neurotransmitter in the central nervous system. The extensive distribution and influence of GABAergic terminals suggests the nervous system operates under considerable restraint, with GABA acting as a tonic controller of excitation. This is also true for the spinal cord where GABA is concentrated in interneurons of the superficial dorsal horn. About one-third of neurons in the superficial spinal cord, the main site of termination of AS- and C-fibre afferents, contain GABA. In addition, there is evidence that GABA can co-exist with either glycine, galanin, enkephalin or neuropeptide Y in separate populations of neurons. GABAergic terminals contact more AS-fibre terminals than C-fibre terminals, and in support of this anatomical data, the benzodiazepine (Bz), midazolam, has weak depressive effects on C-fibre-evoked responses, but marked effects on AS-fibre-evoked responses. In addition, the GABAA antagonist bicuculline facilitates C-fibre-evoked activity less than the profound potentiation of AS-fibre-evoked responses. Both presynaptic and postsynaptic GABAA receptor-mediated mechanisms are documented in the spinal cord. Several studies have demonstrated Bzs to be analgesic, whereas others have found no antinociceptive properties. In addition, there are contradictory reports of Bzs both potentiating and antagonising morphine analgesia. This diversity of results, however, is the product of many different experimental protocols, models of nociception and routes of administration. In addition, the sedative and myorelaxant effects of these compounds must be considered and these will always limit the usefulness of GABAergic agonists.
In the spinal cord GABA can also activate the G-protein-linked GABAb receptor, also found pre- and postsynaptically. Baclofen modulation of nociceptive transmission is seen under inflammatory conditions in animals but in humans the drug appears to lack any analgesic effect.
Almost all clinically used opioid drugs act on the mu opioid receptor, the receptor for morphine, and they can be highly effective analgesics in many patients unless the pain is due to nerve damage where some patients have inadequate control. The assessment of the analgesic effectiveness of opioids in both animals and in patients is complicated by the fact that the type of neuropathy and the extent, duration and intensity of the symptoms will vary. There is no real consensus from clinical studies on the efficacy of morphine in neuropathic pain states. Dose escalation with morphine was shown to produce good analgesia in one study and others have reported that, in general, morphine could be effective in a group of patients with neuropathy. Another study concluded that opioids were entirely ineffective and finally, opioid analgesia was less in
Was this article helpful?
This guide will help millions of people understand this condition so that they can take control of their lives and make informed decisions. The ebook covers information on a vast number of different types of neuropathy. In addition, it will be a useful resource for their families, caregivers, and health care providers.