PNS origin of phantom limb sensation and pain: reply to letter to the editor regarding Foell et al., peripheral origin of phantom limb pain: is it all resolved?

BDNF in Neuropathic Pain; the Culprit that Cannot be Apprehended

In males but not in females, brain derived neurotrophic factor (BDNF) plays an obligatory role in the onset and maintenance of neuropathic pain. Afferent terminals of injured peripheral nerves release colony stimulating factor (CSF-1) and other mediators into the dorsal horn. These transform the phenotype of dorsal horn microglia such that they express P2X4 purinoceptors. Activation of these receptors by neuron-derived ATP promotes BDNF release. This microglial-derived BDNF increases synaptic activation of excitatory dorsal horn neurons and decreases that of inhibitory neurons. It also alters the neuronal chloride gradient such the normal inhibitory effect of GABA is converted to excitation. By as yet undefined processes, this attenuated inhibition increases NMDA receptor function. BDNF also promotes the release of pro-inflammatory cytokines from astrocytes. All of these actions culminate in the increase dorsal horn excitability that underlies many forms of neuropathic pain. Peripheral nerve injury also alters excitability of structures in the thalamus, cortex and mesolimbic system that are responsible for pain perception and for the generation of co-morbidities such as anxiety and depression. The weight of evidence from male rodents suggests that this preferential modulation of excitably of supra-spinal pain processing structures also involves the action of microglial-derived BDNF. Possible mechanisms promoting the preferential release of BDNF in pain signaling structures are discussed. In females, invading T-lymphocytes increase dorsal horn excitability but it remains to be determined whether similar processes operate in supra-spinal structures. Despite its ubiquitous role in pain aetiology neither BDNF nor TrkB receptors represent potential therapeutic targets.

Neuropathic pain; what we know and what we should do about it

Neuropathic pain can result from injury to, or disease of the nervous system. It is notoriously difficult to treat. Peripheral nerve injury promotes Schwann cell activation and invasion of immunocompetent cells into the site of injury, spinal cord and higher sensory structures such as thalamus and cingulate and sensory cortices. Various cytokines, chemokines, growth factors, monoamines and neuropeptides effect two-way signalling between neurons, glia and immune cells. This promotes sustained hyperexcitability and spontaneous activity in primary afferents that is crucial for onset and persistence of pain as well as misprocessing of sensory information in the spinal cord and supraspinal structures. Much of the current understanding of pain aetiology and identification of drug targets derives from studies of the consequences of peripheral nerve injury in rodent models. Although a vast amount of information has been forthcoming, the translation of this information into the clinical arena has been minimal. Few, if any, major therapeutic approaches have appeared since the mid 1990's. This may reflect failure to recognise differences in pain processing in males vs. females, differences in cellular responses to different types of injury and differences in pain processing in humans vs. animals. Basic science and clinical approaches which seek to bridge this knowledge gap include better assessment of pain in animal models, use of pain models which better emulate human disease, and stratification of human pain phenotypes according to quantitative assessment of signs and symptoms of disease. This can lead to more personalized and effective treatments for individual patients. Significance statement: There is an urgent need to find new treatments for neuropathic pain. Although classical animal models have revealed essential features of pain aetiology such as peripheral and central sensitization and some of the molecular and cellular mechanisms involved, they do not adequately model the multiplicity of disease states or injuries that may bring forth neuropathic pain in the clinic. This review seeks to integrate information from the multiplicity of disciplines that seek to understand neuropathic pain; including immunology, cell biology, electrophysiology and biophysics, anatomy, cell biology, neurology, molecular biology, pharmacology and behavioral science. Beyond this, it underlines ongoing refinements in basic science and clinical practice that will engender improved approaches to pain management.

Reaffirming the link between chronic phantom limb pain and maintained missing hand representation

Phantom limb pain (PLP) is commonly considered to be a result of maladaptive brain plasticity. This model proposes that PLP is mainly caused by reorganisation in the primary somatosensory cortex, presumably characterised by functional degradation of the missing hand representation and remapping of other body part representations. In the current study, we replicate our previous results by showing that chronic PLP correlates with maintained representation of the missing hand in the primary sensorimotor missing hand cortex. We asked unilateral upper-limb amputees to move their phantom hand, lips or other body parts and measured the associated neural responses using functional magnetic resonance imaging (fMRI). We confirm that amputees suffering from worse chronic PLP have stronger activity in the primary sensorimotor missing hand cortex while performing phantom hand movements. We find no evidence of lip representation remapping into the missing hand territory, as assessed by measuring activity in the primary sensorimotor missing hand cortex during lip movements. We further show that the correlation between chronic PLP and maintained representation of the missing hand cannot be explained by the experience of chronic non-painful phantom sensations or compensatory usage of the residual arm or an artificial arm (prosthesis). Together, our results reaffirm a likely relationship between persistent peripheral inputs pertaining to the missing hand representation and chronic PLP. Our findings emphasise a need to further study the role of peripheral inputs from the residual nerves to better understand the mechanisms underlying chronic PLP.

Referred cramping phantom hand pain elicited in the face and eliminated by peripheral nerve block

Phantom limb pain is a restricting condition for a substantial number of amputees with quite different characteristics of pain. Here, we report on a forearm amputee with constant phantom pain in the hand, in whom we could regularly elicit the rare phenomenon of referred cramping phantom pain by touching the face. To clarify the underlying mechanisms, we followed the cramp during the course of an axillary blockade of the brachial plexus. During the blockade, both phantom pain and the referred cramp were abolished, while a referred sensation of "being touched at the phantom" persisted. Furthermore, to identify the cortical substrate, we elicited the cramp during functional magnetic imaging. Imaging revealed that referred cramping phantom limb pain was associated with brain activation of the hand representation in the primary sensorimotor cortex. The results support the hypothesis that referred cramping phantom limb pain in this case is associated with a substantial brain activation in the hand area of the deafferented sensorimotor cortex. However, this alone is not sufficient to elicit referred cramping phantom limb pain. Peripheral inputs, both, from the arm nerves affected by the amputation and from the skin in the face at which the referred cramp is evoked, are a precondition for referred cramping phantom limb pain to occur, at least in this case.

Increased excitability of medium-sized dorsal root ganglion neurons by prolonged interleukin-1β exposure is K(+) channel dependent and reversible

KEY POINTS

Neuropathic pain resulting from peripheral nerve injury is initiated and maintained by persistent ectopic activity in primary afferent neurons. Sciatic nerve injury increases the excitability of medium-sized dorsal root ganglion (DRG) neurons. Levels of the inflammatory cytokine interleukin 1β (IL-1β) increase and peak after 7 days. Five to six days of exposure of medium sized DRG neurons to 100 pm IL-1β promotes persistent increases in excitability which abate within 3-4 days of cytokine removal. This is associated with a profound attenuation of K(+) channel currents but only modest increases in function of cyclic nucleotide-sensitive hyperpolarization-activated channels (HCNs) and of voltage-gated Na(+) and Ca(2+) channel currents. It is unlikely, therefore, that direct interaction of IL-1β with DRG neurons is capable of initiating an enduring phenotypic shift in their electrophysiological properties that follows sciatic nerve injury. The findings also underline the importance of K(+) channel modulation in the actions of inflammatory mediators on peripheral neurons.

ABSTRACT

Chronic constriction injury of rat sciatic nerve promotes signs of neuropathic pain. This is associated with an increase in the level of interleukin 1β (IL-1β) in primary afferents that peaks at 7 days. This initial cytokine exposure has been proposed to trigger an enduring alteration in neuronal phenotype that underlies chronic hyper-excitability in sensory nerves, which initiates and maintains chronic neuropathic pain. We have shown previously that 5-6 days of exposure of rat dorsal root ganglia (DRGs) to 100 pm IL-1β increases the excitability of medium-sized neurons. We have now found using whole-cell recording that this increased excitability reverts to control levels within 3-4 days of cytokine removal. The effects of IL-1β were dominated by changes in K(+) currents. Thus, the amplitudes of A-current, delayed rectifier and Ca(2+) -sensitive K(+) currents were reduced by ∼68%, ∼64% and ∼36%, respectively. Effects of IL-1β on other cation currents were modest by comparison. There was thus a slight decrease in availability of high voltage-activated Ca(2+) channel current, a small increase in rates of activation of hyperpolarization-activated cyclic nucleotide-gated channel current (IH ), and a shift in the voltage dependence of activation of tetrodotoxin-sensitive sodium current (TTX-S INa ) to more negative potentials. It is unlikely, therefore, that direct interaction of IL-1β with DRG neurons initiates an enduring phenotypic shift in their electrophysiological properties following sciatic nerve injury. Persistent increases in primary afferent excitability following nerve injury may instead depend on altered K(+) channel function and on the continued presence of slightly elevated levels IL-1β and other cytokines.