Treatment of Painful Peripheral Neuropathies
footpadpatch: November 6, 2012
INTRODUCTION
Peripheral neuropathy is a common problem and for some patients, pain is a disabling component of the neuropathy. The first goal in evaluating patients with painful peripheral neuropathies is to identify the cause, with the hope of identifying treatment to reverse nerve damage. There are a large number of disorders that may be associated with a painful peripheral neuropathy. (Table 1) When neuropathic pain is present, it is appropriate to treat this concurrently, during the evaluation, as well as during the treatment of the neuropathy, if identified is in process. For some types of neuropathy, there are no available treatments for the underlying disorder, and therapy for neuropathic pain, if present may be the only treatment available.
Pain can be classified broadly as two types. Nociceptive pain is protective and is a normal response to tissue injury, serving to warn of the presence of injury. There is also sensitization of peripheral nociceptors and central nervous system changes, which protect the damaged area by avoiding contact.
Neuropathic pain is a pathologic or maladaptive pain, which results from damage to the nervous system, producing pain in the absence of stimulation of nociceptors or inappropriate response to stimulation of nociceptors. Nociceptive and neuropathic pain are not synonymous with acute and chronic pain. For instance, rheumatoid arthritis is a chronic pain, which is nociceptive pain. A herniated disc can cause acute sciatic pain, which is neuropathic.
Patients with neuropathic pain typically describe burning, lancinating, stabbing, cramping, aching and sometimes vice-like pain. It can be paroxysmal or constant. On examination hyperalgesia, allodynia or hyperpathia are sometimes elicited.
The normal pathways involved in the transmission of pain begin with stimulation of nociceptors including those that respond to chemical irritant stimuli such as VR1, DRASIC, P2X3 and noxious heat stimuli such as VR1 and VRL1. Signals resulting from intense mechanical and thermal stimulate A delta fiber nociceptors and intense mechanical ,thermal and chemical stimuli stimulate polymodal C nociceptors. Afferent fibers synapse in Rexed’s lamina I, II and V in the spinal cord, which is the first level of modulation. Opiate receptors and interneurons are present at the dorsal horn. There are are also descending inputs from the hypothalamus, periaqueductal gray. Opiods, norepineprhine and serotonin have modulatory effects on pain transmission.
MECHANISMS OF NEUROPATHIC PAIN
Why certain neuropathies cause pain is unknown, but there has been increasing knowledge about the mechanism of pain. The gate theory of pain proposed that the substantia gelatinosa acted as a gate allowing pain transmission to proceed and that the inhibitory affect of the substantia gelatinosa is increased by large diameter fibers. Central descending influence was also postulated. [Melzack and Wall 1965]Though specifics of this theory have been shown to be incorrect, the basic premise of central modification of pain perception at the dorsal horn and other parts of the central nervous system is still held. Dyck et al [Dyck et al. 1976], noted that neuropathic pain was related to the rate and kind of nerve fiber degeneration. It is not related simply to the ratio of remaining large and small nerve fibers. Brown et al noted that unmyelinated and small myelinated fibers were most prominently involved and unmyelinated nerve fiber sprouting was evident in painful diabetic neuropathy. [Brown et al. 1976]
Animal models of pain have added to the understanding of neuropathic pain. There are 4 commonly used animal models, which include the chronic constrictive injury (CCI) model, the partial nerve transection, spinal nerve transection model and the spared nerve model. [Kim and Chung 1992;Seltzer et al. 1990;Bennett and Xie 1988;Decosterd and Woolf 2000]The chronic constrictive injury model is produced by a loosely constrictive ligature around the sciatic nerve. [Bennett and Xie1988] Almost all of A-β fibers die, as do the majority of Aδ fibers. A large percentage of C fibers persist.
The partial nerve transection model involves tightly ligating and transecting 1/3 to ½ of of the rat sciatic nerve. The spinal nerve transection model involves tight ligation and transection of the L5 and L6 nerve roots. This affects 50% of the sural and saphenous nerve fibers. The spared nerve model involves a lesion of 2 of the 3 terminal branches of the sciatic nerve (tibial and peroneal) intact, leaving the sural nerve intact. [Decosterd and Woolf2000]There is partial deafferentation but not differential involvement of nerve fibers, in these three models.
Heat hyperalgesia, mechano-hyperalgesia, mechano-allodynia and cold-allodynia may be observed and behavioral changes consistent with spontaneous pain such as limping and guarding the hind paw are seen. Both peripheral and central mechanisms have been observed in animal models.
After nerve injury there are spontaneous discharges of nerve fibers, neuromas and dorsal root ganglion. This has been seen in traumatic nerve injury models , as well as models of peripheral neuropathy from diabetes and heavy metal intoxication. Increased expression of sodium channels are seen in neuroma’s in humans and animal models. The density of sodium channel expression correlates with the degree of pain. [England et al. 1996] Changes in sodium channel expression are seen and may contribute to nerve excitability and spontaneous pain. [Waxman et al. 1999;Craner MJ et al. 2002].
These spontaneous discharges by themselves can cause pain, but they also have additional affects. C fiber afferents trigger cell death of neurons in the dorsal horn, where inhibitory interneurons are concentrated, possibly through an excitotoxic mechanism. This may result in increased pain transmission.. ([Sugimoto et al. 1989;Sugimoto et al. 1990;Laird and Bennett 1992] C fiber afferents release glutamate and synapse on 2nd order neurons and have excitatory affects. Glutamate synapses at AMPA receptors which depolarizes the membrane. This depolarization releases the inhibition of the NMDA receptor by the magnesium ions and there is an influx of calcium. Second order neurons are gradually depolarized and responses are amplified and this changes the response of neurons to subsequent input.
Two processes that are distinct occur at the dorsal horn which are designated “windup” and “central sensitization.” Windup up results from repetitive C fiber firing at low frequencies that result in a progressive buildup of the amplitude of the response of the dorsal horn neuron, only during the repetitive train. Central sensitization is an abnormal sensitivity with a spread of hypersensitivity to uninjured sites and pain resulting from stimulation of low threshold A mechanoreceptors. Central sensitization follows a brief high frequency input and the increased response to subsequent inputs is prolonged. Both can be blocked by NMDA receptor antagonists. Central sensitization can result from windup. This is a result of the calcium influx through the NMDA receptor following depolarization of the dorsal horn membrane. The intracellular calcium activated a number of kinases of which protein kinase C (PKC) is likely important. PKC enhances the NMDA receptor , which results in subsequent glutamate binding of the NMDA receptor generating an inward current. Though windup can result in central sensitization it is not necessary for central sensitization to occur. [Woolf and Salter 2000;Woolf 1996]
Similar observations had previously been noted by Denny-Brown who described an enlarged and hypersensitive dermatomal region in primates when severing the surrounding nerve roots distal to the dorsal root ganglion compared with severing proximal to the DRG. This suggested plasticity of the dorsal horn neurons secondary to input from the DRG. [Kirk and Denny-Brown 1970;Kirk and Denny-Brown1970;Denny-Brown et al. 1973]
Nerve injury also results in sprouting of myelinated fibers into lamina II of the dorsal horn, which under normal circumstances receives only C fiber input. [Mannion et al. 1996]This may result in allodynia.
There is a genetic influence on the experience of neuropathic pain though it is poorly delineated. There are variations on the expression of pain behaviors seen in different strains of animals. [Devor and Raber 1990] A recent study noted that 56% of patients with painful diabetic neuropathy had a relative with painful diabetic neuropathy suggesting a genetic component.[Galer et al. 2000]
The relative importance of these various mechanisms is not clearly known, however there is potentially multiple sites for intervention in treating painful neuropathies.
Medications
Tricyclic anti-depressants have been beneficial in controlled studies of neuropathic pain. [Max 1995]They block the reuptake of norepinephrine and serotonin and are thought to modulate descending inhibitory pathways. There benefit with neuropathic pain is independent of their effect on depression. [Max et al. 1987] Patients who are not depressed can respond and lower doses than are used to treat depression are effective to treat neuropathic pain. Tricyclic anti-depressants can block voltage dependent sodium channels and this may contribute to their efficacy to treat neuropathic pain. [Brau et al. 2001] They are typically started at a dose of 10-25 mg at night and titrated as tolerated up to a dose of 150 mg if necessary. Side effects include dry mouth, cardiac arrhythmias, urinary retention and sexual dysfunction. Venlafaxine has fewer side effects and has also been reported to benefit neuropathy pain. [Ansari 2000]
The selective serotinin reuptake inhibitors (SSRI) have been less effective for treatment of neuropathic pain. Fluoxetine was not effective in clinical trials. [Max et al. 1992]Paroxetine and Citalopram been reported to be effective. [Sindrup et al. 1992;Sindrup et al. 1990]
Anticonvulsants have also been studied in neuropathic pain. Both phenytoin (Dilantin) and carbamezapine (Tegretol) have been beneficial in trials of diabetic neuropathy. [Chadda VS 1978;Rull et al. 1969]They act as sodium channel blockers. Both medications have frequent side effects that are well known.
Gapapentin (Neurontin), binding site alpha delta Ca channel. It is though to act at a spinal site of action. [Xiao and Bennett 1996] Controlled studies in diabetic neuropathy, post-herpetic neuralgia and other neuropathic pain states. [Backonja et al. 1998] Starting doses are usually100mg tid or 300mg qhs. Unless the drug is not tolerated or a dose of 600 mg tid has been tried it should not be considered a treatment failure. Often higher doses provide more benefit. Side effects include drowsiness, fogginess, leg edema. Gabapentin is cleared by the kidneys and there are not significant drug interactions. The dosing should be adjusted in renal insufficiency and failure.
Lamotrigine (Lamictal) is a sodium channel blocker, which also inhibits glutamate release. It has reduced cold allodynia in the CCI model. [Hunter et al. 1997]Lamotrigine is effective in controlled studies in diabetic, HIV associated painful neuropathy, post-herpetic neuralgia, and trigeminal neuralgia.[Simpson et al. 2000;Eisenberg et al. 2001]. I’ve found lamotrigine to cause less drowsiness or dizziness than most other medications used for neuropathic pain and though these symptoms have been seen during epilepsy trials, they were not seen more commonly than placebo in 2 recent trials of neuropathy pain. Rash is a side effect and a Stevens Johnson syndrome may occur. This is less frequent when the medication is titrated slowly. The rash is less common with a slow titration. The following titration schedule has been used for neuropathic pain: 25mg qd for 2 weeks, 25 mg bid for the next 2 weeks, 50 mg bid for the next 2 weeks, 100mg bid for 2 weeks. If the patient is also taking valproate which inhibits the metabolism of lamotrigine, a dose of 25 mg qod is recommended as an initial dose.
Oxcarbazepine (Trileptal) is similar in structure to carbamezapine, but lacks the 10,11 epoxide, which is thought to be responsible for better tolerability. Autoinduction does not occur and rash and drug interactions are less frequent than with carbamezapine. Hyponatremia does occur. [Zakrzewska and Patsalos 1989]
Double blind placebo controlled studies of mexilitine have been negative, however subgroups with stabbing and burning pain had benefit. Another study noted benefit in night time pain with the 675 mg dose, but not the lower doses of 450 mg or 225 mg.
Topiramate (Topamax) showed benefit in animal models and preliminary studies and anecdotal reports of diabetic neuropathy and neuropathic pain. [Edwards et al. 2000;Potter and Edwards 1998]Though the results have not been published Johnson & Johnson announced that the results of clinical trials in diabetic neuropathy pain were not positive (Reuter 9/18/2001).
Dextromorthorphan is a low affinity NMDA antigonist. It has been beneficial in animal studies and painful neuropathies. Side effects are frequent and it is poorly tolerated. [Nelson et al. 1997] Dextromethorphan in combination with other medications as well as more selective NMDA receptor antagonists that may have less side effects are being pursued.[Boyce et al. 1999]
Clonidine an alpha-2 agonist, as a transdermal patch was successfully used in a subset of patients with painful diabetic neuropathy using an enrolled enrichment design. Tizanidine, also an alpha-2 agonist, reduces thermal hyperalgesia in the CCI rat model and has been successful in open label studies in neuropathic pain.[Hord et al. 2001;Semenchuk and Sherman 2000b]
Tramodol (Ultram) has been effective in studies of painful diabetic neuropathy.[Harati et al. 1998] Side effects include nausea, headache, constipation, somnolence and seizures. Lidoderm is beneficial for post-herpetic neuralgia. The dose of 1-3 patches for 12 hours, with 12 hours off the patch. It has been beneficial in other forms of neuropathic pain. [Devers and Galer 2000]
Based on small preliminary and anecdotal series, as well as animal data, other medications including zonisamide and levetiracetam may be beneficial. [Backonja 2002;Semenchuk and Sherman 2000a;Hord et al2001]
HIV painful neuropathy is a particularly difficult syndrome to treat. Controlled studies of amitriptyline, mexilitine, and acupuncture have been negative. [Shlay et al. 1998;Kieburtz et al. 1998]Gabapentin has been reported to be successful, though requiring higher doses. [Newshan 1998] Phenytoin and carbamezapine as P450 inducers should be avoided as they can induce the metabolism of the protease inhibitors and make them ineffective. [Romanelli et al. 2000]Lamotrigine has been successful in blinded placebo controlled studies. Two consecutive studies however provided conflicting data with the first suggesting that there was no benefit above placebo in patients taking neurotoxic dideoxynucleotide anti-retrovirals compared with those patient not taking neurotoxic antiretrovirals and the second study showed the opposite. [Simpson et al. 2002;Simpson et al2000]
Since several different mechanisms are involved in neuropathic pain, treatment directed against the mechanism of pain would be desirable and has been proposed. Unequivocally identifying mechanisms at work to date has been difficult in patients. For instance the symptom of allodynia may occur from peripheral sensitization or central sensitization. [Woolf et al. 1998] It is a commonly held belief that tricyclic anti-depressants are more effective against burning pain and anti-convulsants are more effective against paroxysmal stabbing pain. In clinical trials however this has not been demonstrated.. Selective affects in animal models have been demonstrated. For example dextrorphan reduces heat hyperalgesia, but has no effect on mechanical allodynia in the CCI model. [Tal and Bennett 1994] How this correlates to treatment of people with neuropathic pain is still evolving. In the future cocktails targeting different mechanisms at work in patients with peripheral neuropathies may be possible.
http://geenabhud.wordpress.com/2012/11/06/d461/
