The International Association for the Study of Pain (IASP) definition of neuropathic pain is; 'Pain initiated or caused by a primary lesion or dysfunction in the nervous system. Peripheral neuropathic pain occurs when the lesion or dysfunction affects the peripheral nervous system. Central pain may be retained as the term when the lesion or dysfunction affects the central nervous system'.
In all the literature on the subject of neuropathic pain (NeuP) there are several over-riding themes (if not frustrations) which emerge.
· NeuP has a complex and misunderstood pathophysiology.
· There is a deficiency of effective remedies.
Over half a million people in the UK are affected. However, to put it into context, NeuP is not a given sequel to nerve injury and most patients do not develop NeuP after nerve injury.
Recognition & treatment of neuropathic pain in people
Pain which is not well controlled by conventional analgesics (opioids & NSAIDs) should raise the index of suspicion of NeuP.
NeuP is often described as burning, shooting or stabbing.
Features of NeuP include
-Allodynia- a painful response to an innocuous stimulus-assessed by lightly stroking the skin
-Hyperalgesia- an increased response to a painful stimulus- assessed by the response to a pin prick
-Altered skin sensation
-Dysaesthesia (an unpleasant abnormal sensation (can be spontaneous or evoked).
Painful diabetic neuralgia is the main cause of NeuP in humans in the UK with 30% of diabetics affected at some stage. Painful neuralgias such as trigeminal and post-herpetic neuralgia are common in clinical practice. Pain after discectomy and post surgical pain syndromes are also increasingly acknowledged as having neuropathic components.
Diagnostic tools are available in human medicine which are designed to differentiate nociceptive pain from neuropathic pain (Leeds Assessment of Neuropathic Symptoms & Signs) however classification is still difficult and often presumed based on the underlying clinical cause. Already one can appreciate how difficult assessment of NeuP is in animals.
Mechanisms of neuropathic pain
The pathophysiology of NeuP is due to complex changes in the physiology of the nerve and not due to peripheral nociceptor stimulation (as in inflammatory pain). 90% of NeuP arises from the peripheral nervous system although there is growing evidence that many chronic pain conditions such as low back pain have a central neuropathic component. A discussion of the mechanisms of NeuP gives an understanding of potential drug targets as well as explanations of how current drug therapies achieve their effects.
Simply put, NeuP is due to ion channel changes and electrical events. This next section explores those in greater detail.
Peripheral Mechanisms
Hyperexcitability
Voltage gated sodium channels (Nav), both peripherally and centrally play a role in inflammatory and NeuP. In NeuP changes in activity in these channels, notably oscillations of membrane potential, abnormal firing and ectopic activity, produces and maintains a level of hyperexcitability which causes central hyperexcitability. Selective depression of this hyperexcitability by targeting specific sodium channels is valuable in the search for new therapies.
-Nav 1.7 mutations are responsible for the burning pain of erythromelalgia
-Loss of Nav 1.7 channels leads to complete ablation of pain perception in humans
-Nav 1.8 is down regulated in small injured axons but upregulated in adjacent uninjured C fibres – an important mechanism for generation of spontaneous activity – so not only are injured neurons contributing to ectopic activity but also their uninjured neighbours.
So why not use local anaesthetics then? Yes, lidocaine has shown to be of benefit in NeuP treatment but is unable to discriminate between specific sodium channels, hence the targeting of specific channels. These sodium channels represent a huge area of research into remedies for NeuP.
Axonal Sprouting
Collateral fibres have been shown to sprout into denervated areas following nerve injury, however the degree of new growth is not proportional to the degree of hyperalgesia and so is thought not to contribute significantly to pain. Although axonal sprouting seems an easy explanation to account for the symptoms of phantom (limb) pain it is certainly not the only theory, with peripheral (regional blood flow) and central (cortical reorganization) factors appearing as more likely explanations for the pathophysiology (Flor et al, 2000).
Sympathetic-Sensory Coupling
In some NeuP cases the degree of pain is related to sympathetic (SNS) activity – a feature characteristic of another pain condition - complex regional pain syndrome. Abnormal contact between the two systems has been demonstrated following peripheral nerve injury with research suggesting that this occurs as the level of the dorsal root ganglion. This is not well documented in our veterinary patients.
Central Mechanisms
Hyperexcitability
With the constant barrage from ectopic discharge in the periphery, a state of hyperexcitability is induced in the dorsal horn neurons known as central sensitization. The NMDA receptor plays a significant role here, hence the potential for ketamine and other NMDA antagonists in the treatment of NeuP.
As well as sodium channels peripherally, voltage gated calcium channels play a role in generation and maintenance of abnormal excitability centrally. The alpha 2 delta 1 subunit of the calcium channel is over expressed in the dorsal root ganglia and spinal neurons in NeuP. This unit is the target for gabapentin and pregabalin although exactly how these drugs interact with the channel is unclear, however both have been shown to be effective in reducing allodynia and hyperalgesia in animal and human models and are licensed in people for NeuP.
Glutamate is another neurotransmitter responsible for maintenance of hyperexcitability. Glutamate acts on NMDA receptors, one of the targets for ketamine – a drug limited in the field of NeuP management by its side effects. Amantadine is an oral NMDA antagonist used to treat central sensitisation associated with osteoarthritis. Read about amantadine .
From the above discussion it can be seen that central hyperexcitability is particularly important for the development of hyperalgesia, allodynia and the ongoing pain associated with NeuP.
Spinal Cord Re-Organisation
Following peripheral nerve injury, Abeta fibres, which normally transmit innocuous sensory information are able to transmit pain signals, a role normally assumed by C fibres. This occurs due to spinal reorganization – the areas where Abeta and C fibres terminate, which were previously distinct, now overlap. This means that the second order neurons which transmit information to the higher centres are inundated with pain impulses from the periphery – known as a widened receptive field – essentially collecting more information on pain.
Endogenous Analgesics
The lack of opioid efficacy in NeuP is widely recognised. During acute inflammation opioid receptors are expressed in the periphery however by contrast during nerve injury it is thought that there is a consequential loss of axonally-expressed opioid receptors, accounting for a lack of analgesic efficacy. As well as this, a decrease in opioid binding at a spinal level and an increase in the opioid antagonist neuropeptide cholecystokinin account for a decreased opioid effect in NeuP. Studies do however suggest that spinal delivery of morphine is more effective in NeuP than systemic administration.
The cannabinoids form part of an endogenous modulatory system acting on CB1 receptors (central & peripheral) and CB2 receptors (peripheral only). Current evidence demonstrates an analgesic effect in NeuP with targeting of the CB1 receptor, but with associated side effects. Selective targeting of the peripheral CB1 receptor appears to reduce CNS side effects but maintain analgesia. Evidence shows that paracetamol has a cannabinoid effect as one of its modes of action although efficacy in NeuP is uncertain.
Descending Control
Descending inhibition plays a role in modulation of incoming nociceptive signals and is regulated by serotonin and noradrenaline, hence the role of the serotonin and noradrenaline reuptake inhibitor amitriptyline in treatment of NeuP. This group of drugs is the most commonly prescribed for NeuP and their use is supported by several randomized controlled clinical trials, hence their role as first line treatments. By blocking reuptake of noradrenaline and serotonin pain signals are down regulated in the spinal cord to reduced transmission to and therefore the perception of pain by the cerebral cortex.
So it appears that many mechanisms are involved in the generation and maintenance of NeuP (there are several more beyond the scope of this review) The reader will appreciate the gap between the previous discussion of mechanisms of NeuP and how treatment is actually directed on a clinical level. The review of emerging treatments by Dray cites 25 drugs currently in various phases of clinical trials, all of which are target specific.
Treatment
NeuP has been described as an area of unmet need pharmacologically. The mainstays of treatment in people are the tricyclic antidepressants (amitriptyline) and anticonvulsants (gabapentin), which are only effective in 50% of cases.
Patients with the same diagnosis show wide variation in response to treatment.
Reports of neuropathic pain in dogs
Three cases of NeuP were reported by Cashmore et al (2009) which may be useful in assisting the reader to recognize similar clinical presentations. Symptoms in people which are characteristic of NeuP include burning, shooting pain. The first case detailed by Cashmore et al describes a dog with mechanical allodynia over the antebrachium and palmar forepaw. ‘Stroking with a pen consistently caused marked agitation in the dog, displayed by it pulling its leg away while concurrently vocalising and attempting to bite the
examiner. The limb was held in a flexed position at the shoulder and elbow joints, which was thought to be a consequence of the abnormal sensation experienced by the dog’.
Electromyography suggested a lesion of the median and ulnar nerve roots and MRI failed to demonstrate a nerve root lesion. The dog was treated with amitriptyline 1.4mg/kg PO BID for a month. Within two weeks the owners reported no significant change in the dogs presenting signs and several adverse effects including altered mentation. Amitriptyline was discontinued and gabapentin 14mg/kg BID prescribed. One month later the lameness had improved and the mechanical allodynia had resolved. Following 3 months treatment with gabapentin the lameness has resolved and the dog was using the limb normally.
This case is a useful addition to the literature as it describes clinical signs in a dog which appear similar to those described by people with NeuP. It also reinforces the human perspective regarding treatment whereby the first line choice may be ineffective or rather the beneficial effects limited due to side effects.
The second case report is of a 3.5 year old Cairn Terrier with a progressive history of intermittent spontaneous and inducible scratching of the right side of its face. Each episode lasted 20s during which the episodes could not be stopped by the owner nor the clinician. No response was seen to NSAIDs or corticosteroids. Neurological and dermatological exams revealed no further abnormalities, except mechanoallodynia over the right side of the face located to the dermatome supplied by the infraorbital nerve. MRI was unremarkable. A diagnosis of NeuP was presumed. A one month trial of gabapentin did not improve clinical signs. One month of amitriptyline produced an immediate and dramatic response until the owner stopped treatment having considered the dog to be cured. Signs recurred within 3 days and were controlled again by amitriptyline.
The third case in the series was a 12 year-old Fox Terrier presented for back pain. The owners were unable to touch the dog’s lower back without eliciting severe discomfort, agitation and sometimes aggression. Radiographs taken by the referring vet were unable to account for clinical signs and trial treatment with meloxicam failed to show improvement. Upon presentation at the referral centre diffuse mechanical allodynia was noted over the lumbar region. Otherwise neurological, orthopaedic, radiographic and MRI & EMG exams were considered normal. A presumptive diagnosis of NeuP with no inciting factor was made. The dog showed a marked improvement in response to amitriptyline, with marked deterioration upon cessation of treatment.
These reports document methodical diagnosis of what appear to be NeuP, based on history, diagnostics and successful response to treatment. In each case the authors quite rightly use the words ‘presumed diagnosis’. Without a clear, easily recognizable example of NeuP in clinical practice it is difficult to construct a randomized, blinded clinical trial to evaluate one treatment versus another, although such case reports provide a starting block.
One case report describes the use of amantadine to treat neuropathic pain in a dog. Amantadine was chosen in this case based on the pathophysiology of the condition and the rationale that an NMDA antagonist was required. The dog responded positively to the addition of the amantadine to NSAID therapy. When the amantadine was stopped the signs of pain reverted – but control was gained a second time once amantadine was reintroduced. (Madden, Gurney, & Bright, 2014).
Pain is reported in 35% of dogs with syringomyelia and the most likely explanation of this pain appears to be central neuropathic pain. Pain is more likely in dogs with a wide syrinx and where the syrinx is deviated into the dorsal horn of the spinal cord. One theory suggests that the pain is a result of an imbalance of input to and output from the dorsal horn due to either anatomical or neurochemical alterations. Impaired dorsal horn function has been demonstrated in people with syringomyelia however the exact mechanism remains unclear. First line treatment options include NSAIDs (response can be variable) plus gabapentin or pregabalin. Examination of lifestyle features is important here to avoid factors (examples = harnesses, unexpected contact with other dogs during play, brushing) which can exacerbate the condition.
For treatment options and more details on syringomyelia click here.
Arthritis Pain
Typically we class OA pain as nociceptive. Thakur et al (2014) argue that a neuropathic component may exist because some OA pain is poor to control with conventional analgesics – a classic feature of neuropathic pain. So is there a lesion or disease of the somatosensory system to fit the NeuP classification? To fully evaluate this we need to perform QST (quantitative sensory testing) – which is well established in the research laboratory setting but perhaps less accessible to a practice setting.
So are peripheral nerves damaged in OA? Studies in rats do suggest upregulation of markers on neuropathy in OA models. In humans with OA where synovitis is present there are documented abnormalities in the innervation of the synovial layers. Combined with this there is an increase in cartilage innervation (cartilage is usually aneural) – an ectopic innervation arising from subchondral bone. These changes are referred to as plasticity – changes in the way pain is sensed within the diseased joint.
Regarding treatment, there is evidence that amantadine in conjunction with meloxicam is effective in management of OA and plenty of anecdotal reports of the use of gabapentin alongside NSAIDs.
Research into Neuropathic Pain
So why is progress in research so slow? It is recognized that the underlying mechanisms of NeuP are varied and so far treatment has been symptomatic. The drive for research for new treatments will ideally be targeted to the underlying mechanisms. At present there are no human volunteer models for NeuP, unlike inflammatory pain. As vets it is easy for us to empathise with pain researchers who struggle to determine what is perceived by the animal models. NeuP can occur after nerve injury, but not always, so will full or partial ligation of a rat’s sciatic nerve necessarily produce NeuP? And whilst animals in research may be studied for weeks, is it ethical to study such models for years – years being the normally course of NeuP?
Conclusion
Whilst it would seem obvious that correct diagnosis is the key to tailoring treatment to the underlying pathophysiology it is clear from the human literature that not only can the diagnosis be elusive, but the response to treatment can be far from predictable. This is the driving force behind research into neuropathic pain which has a large focus on the mechanisms of neuropathic pain which are multiple and complex.
References
Allen S. (2005) Pharmacotherapy of neuropathic pain. Continuing Education in Anaesthesia, Critical Care & Pain 5 (4) 134-137
Bennett M. (2001) The LANSS Pain Scale. The Leeds assessment of neuropathic symptoms and signs. Pain 92 147-157
Benrath J., Scharbert G., Gustorff B., Adams H.A., Kress H.G. Long-term intrathecal S(+)-ketamine in a patient with cancer-related neuropathic pain. British Journal of Anaesthesia 95 (2) 247-9
Bridges T., Thompson S.W.N., Rice A.S.C. (2001) Mechanisms of neuropathic pain. British Journal of Anaesthesia 87 (1) 12-26
Cashmore R.G., Harcourt-Brown T.R., Freeman P.M., Jeffery N.D., Granger N. (2009) Clinical Diagnosis and treatment of suspected neuropathic pain in three dogs. Australian Veterinary Journal 87 (1) 45-50
Dray A. (2008) Neuropathic pain: emerging treatments. British Journal of Anaesthesia 101 (1) 48-58
Flor H., Birbaumer N., Sherman R.A. (2000) Phantom Limb Pain. Pain. Clinical Updates VII (3) (available free via www.iasp-pain.org)
Rusbridge C., Jeffery N.D. (2008) Pathophysiology and treatment of neuropathic pain associated with syringomyelia. The Veterinary Journal 175 164-172
Thakur, M., Dickenson, A. H., & Baron, R. (2014). Perspectives. Nature Reviews Rheumatology, 10(6), 374–380. http://doi.org/10.1038/nrrheum.2014.47