Current Diagnosis & Treatment of Pain, 1st Edition

10. Neuropathic Pain

Norman R. Harden MD

Essential Criteria

  • Neuropathic pain can be divided into two types: stimulus-evoked and stimulus-independent.
  • Stimulus-evoked pain is characterized by signs of hyperalgesia and allodynia that result from mechanical, thermal, or chemical stimulation.
  • Stimulus-independent pain (ie, spontaneous pain) may be persistent or paroxysmal and is often described as shooting, stabbing, or burning.
  • Paresthesias (defined as abnormal sensations) and dysesthesias (defined as unpleasant abnormal sensations) may be spontaneous or evoked.

General Considerations

  1. Epidemiology

Chronic neuropathic pain is not uncommon, although estimates of its prevalence vary widely from 2 to 40% of all adults. The estimated 3.75 million cases of chronic neuropathic pain in the United States include conditions as diverse as cancer-associated pain, spinal cord injury, low back pain, and phantom pain. Recurrent and persistent pain ranging from back pain to facial pain was reported by 45% of enrollees in one health maintenance organization in the United States, and in the United Kingdom up to 25% of patients who attended pain clinics experienced neuropathic pain syndromes.

  1. Etiology

Neuropathic pain that is associated with such disorders as diabetes mellitus and herpes zoster is the most frequently described and studied. However, these disorders are certainly not the exclusive causes of neuropathic pain. Radiculopathy, which may be an underlying cause in many cases involving lower back pain, is probably the most frequent peripheral nerve pain generator. A partial list of causes of neuropathic pain is presented in Table 10-1.

  1. Diagnostic Issues

Chronic pain should be broadly categorized into two categories: nociceptive (ie, stimulus of pain peripheral receptors in skin, joint, and muscle) and neuropathic (also called neurogenic, implying that damage to the nervous system is generating the pain). Pain is often multifactorial; there may be multiple primary and secondary pain generators.

Neuropathic pain refers to pain caused by a clinically heterogeneous group of disorders that vary widely in etiology and presentation. It includes signs and symptoms that arise from a primary lesion in the peripheral nervous system or from dysfunction in the central nervous system which occur in the absence of nociceptor stimulation, such as postherpetic neuralgia.

In contrast, nociceptive pain is a response triggered by an unpleasant, damaging, or potentially damaging stimulus in the periphery. It can be acute, such as acute postoperative pain, or it may be chronic, such as the inflammation of arthritis.

This binary categorization has clinical significance; for instance, neuropathic pain may not respond as well to opioids or nonsteroidal anti-inflammatory drugs (NSAIDs) as nociceptive pain, which is usually easily managed with these classes of drugs at least in the short-term. Neuropathic pain may be treated more effectively by drugs that stabilize or modulate central nervous system function (eg, drugs indicated for seizures or depression) or by antiarrhythmic agents (such as sodium channel blockers). While the reasons for correctly diagnosing neuropathic pain are clear, the methods for effectively doing so are not.

At present, the diagnostic approach to neuropathic pain relies on antiquated classification systems based on the etiology of pain and its anatomic distribution. This is less than ideal for several reasons. First, most neuropathic disease states are associated with more than one mechanism of pain and that mechanism may change over time. Second, different syndromes may produce mechanistically identical types of neuropathic pain. Finally, presenting symptoms and signs and diagnostic testing are often diverse within a single neuropathic pain syndrome. All of these problems hinder accurate and efficient diagnosis of neuropathic pain and contrast the


descriptive diagnoses with the “mechanistic diagnoses” currently recommended in pain circles. However, exist problems with this “mechanistic” diagnostic approach as well.

Table 10-1. Causes of Neuropathic Pain.





Compression/ entrapment syndromes

Carpal tunnel syndrome Chronic radiculopathy Plexus disorders Spinal stenosis Tarsal tunnel


Fabry disease
Hereditary sensory neuropathies


Guillain-Barré syndrome
Multiple sclerosis


Herpes zoster
Infectious mononucleosis

Metabolic disturbance

Diabetes mellitus

Nutritional deficiency

Folic acid deficiency
Niacin deficiency
Pyridoxine deficiency
Thiamine deficiency


Chemotherapy agents, especially vincristine, cisplatin, oxaliplatin, and taxanes
Glue sniffing
Other drugs including hydralazine, isoniazid, nitrofurantoin, phenytoin, thalidomide


Amputation (phantom limb pain/stump pain)
CRPS type II
Crush injuries
Spinal cord injury

Vasculitis/ connective tissue disorders

Lupus erythematosus
Polyarteritis nodosa
Rheumatoid arthritis Sjögren syndrome

Postherpetic neuralgia can be used to illustrate the pitfalls in diagnosing neuropathic pain according to mechanism. In postherpetic neuralgia, at least three different mechanisms for pain have been identified, all of which are associated with direct neuronal damage to both the peripheral and central nervous systems (ie, infectious, inflammatory, and ischemic). Each of these mechanisms may be associated with different symptoms. For instance, some patients present with a profound sensory loss in an area of pain. Other patients will have pronounced allodynia and hyperalgesia with minimal or no sensory loss. Still others will have sensory loss and allodynia. This multiplicity of potential mechanisms, signs, and symptoms increases the potential for misdiagnosis and may result in complicated or conflicting “mechanistic diagnoses.” Consequently, the response to treatment will be unpredictable, and two different patients with postherpetic neuralgia may respond differently to the same treatment. However, it is clear that mechanistically based diagnoses are superior to descriptive diagnoses. With new insights being gained into the biologic mechanisms underlying neuropathic pain, a more valuable way to approach neuropathic pain is not only through a clinical framework that categorizes pain according to the presumed etiology or affected body part, but also according to presenting signs, symptoms, and electrodiagnostic and quantitative sensory testing. The combined power of these complementary tactics may yield a more precise and clinically useful diagnostic target, and this approach has been gaining some acceptance in the pain community.

  1. Treatment Issues

Adding to the clinical challenge of treating neuropathic pain is the fact that many currently prescribed therapies lack evidence-based support in the form of prospective, randomized, controlled clinical trials or, in the case of drugs, approval by the US Food and Drug Administration (FDA) for the indication of neuropathic pain. (Exceptions to the latter include carbamazepine, which is approved for trigeminal neuralgia; the lidocaine patch and gabapentin, which are approved for postherpetic neuralgia; and duloxetine, which is approved for painful diabetic neuropathy.) Therefore, medications indicated for the treatment of other syndromes (including depression, seizures, and cardiac arrhythmias) are used off-label for the treatment of neuropathic pain. Without rigorous clinical data to support safety and efficacy in patients with neuropathic pain, formal guidelines for dosage and administration of many of these off-label drugs are not adequately established. These limitations render the current


haphazard treatment approaches even more cumbersome and speculative; however, they do provide a modest rationale for interdisciplinary care (ie, with drug treatments of variable effectiveness, nonpharmacologic treatments are more likely to be used). However, it must be noted there is even less evidence for nondrug interventions; at least for interdisciplinary care, the risk part of the risk/benefit ratio is minimal.

Close analysis of the published data reveals some useful information regarding the clinical usefulness of commonly used drugs for specific neuropathic pain symptoms. Though the majority of these studies target descriptive diagnoses (that is, specific drugs were evaluated in patients in a traditional diagnostic model), understanding treatment mechanisms and deriving treatment (nonpharmacologic as well as pharmacologic) from symptoms, signs, and test results as well as the underlying mechanisms they reveal will result in more effective therapy and improved quality of life for the patient. However, the literature providing this type of information is sparse.

Dworkin RH et al. Advances in neuropathic pain: diagnosis, mechanisms, and treatment recommendations. Arch Neurol. 2003;60:1524 .

Mannion RJ et al. Pain mechanisms and management: a central perspective. Clin J Pain. 2000;16:S144 .

Merskey H, Bogduk N, eds. Classification of chronic pain: descriptions of chronic pain syndromes and definitions of pain terms. 2nd ed. Seattle: IASP Press; 1994.

Woolf CJ. Dissecting out mechanisms responsible for peripheral neuropathic pain: implications for diagnosis and therapy. Life Sci. 2004;74:2605 .

Types of Neuropathic Pain

Physiologically, neuropathic pain results from central or peripheral nervous system damage, threat of damage, or dysfunction. Although nervous system damage would logically be expected to cause a sensory loss (negative symptoms)-with the degree of loss approximating the amount of damage-a proportion of cases present with various kinds of pain and dysesthesias (or positive symptoms).

The two relevant types of neuropathic pain are stimulus-evoked pain and stimulus-independent pain (ie, spontaneous pain). Stimulus-evoked pain is characterized by signs of hyperalgesia and allodynia that result from mechanical, thermal, or chemical stimulation. Stimulus-independent pain may be persistent or paroxysmal and is often described as shooting, stabbing, or burning. Paresthesias (defined as abnormal sensations) and dysesthesias (defined as unpleasant abnormal sensations) may be spontaneous or evoked.

Stimulus-Evoked Pain

Within the category of stimulus-evoked pain, hyperalgesia and allodynia are the two main symptoms that may manifest via mechanical, chemical, or thermal stimulation. Hyperalgesia is an exaggerated pain response produced by a normally painful stimulus (eg, pin prick), while allodynia is pain produced by a stimulus that is not usually painful (eg, light touch).

  1. Hyperalgesia

Hyperalgesia can arise from peripheral or central mechanisms, or both. Peripherally, sensitization of primary afferent nociceptors (Aδ and C fibers) may occur due to release of inflammatory mediators, such as bradykinin, histamine, prostaglandins, and substance P. Another peripheral mechanism for stimulus-evoked pain involves formation of a neuroma, a tangled mass of regenerating nervous tissue embedded in scar and connective tissue at the site of nerve injury which may act as a mechanically sensitive site. Neuromas accumulate or “uncover” pathologic and nonpathologic ion channels (eg, sodium channels) and receptors (eg, norepinephrine) that result in foci of hyperexcitability and ectopic activity. The neuroma sign, or Tinel sign, may be elicited by mechanically stimulating the affected area, triggering exquisite, “electrical” pain caused by changes in afferent nerve membrane properties and mechanical threshold.

  1. Allodynia

Allodynia may also be evoked by what is usually perceived as innocuous stimuli (usually mechanical or temperature). This may be due to peripheral or central sensitization. Peripheral sensitization occurs due to persistent release and presence of inflammatory or algesic substances in the local environment. In response to ongoing nociception or overstimulation, changes in spinal cord dorsal horn cells can occur, resulting in the central sensitization and central reorganization that finally lead to allodynia. Central sensitization may cause an increase in the size of the sensory receptive field, a reduced threshold for sensory (pain) perception, and hypersensitivity to various innocuous stimuli.

At the molecular level, central sensitization occurs when the excitatory amino acids glutamate and as-partate and substance P bind to receptors located on spinal dorsal horn transmission cells (second-order neurons). Specific glutamate receptors include N-methyl-D-aspartate (NMDA) receptors and non-NMDA receptors (α-amino-3-hydroxy-5-methyl-4-isoazolepropionic acid [AMPA], kainate), which may enhance and prolong depolarization. This may increase the responsiveness of the nociceptive system and lead to long-lasting changes in the dorsal horn transmission cells. In addition, NMDA receptors may be involved in potentiating synaptic


transmission into the cerebrum, a process that may be responsible for “pain memory,” (eg, phantom limb pain). In fact, it is likely that there are pain-associated excitatory amino acid receptors throughout the neuroaxis. Activation of non-NMDA receptors, specifically the AMPA and kainate receptors and neurokinin-1 (substance P) receptors may act to further sensitize the NMDA receptor.

Central changes also occur through reorganization. As the damaged nerve regenerates or begins firing ectopically or ephaptically, Aβ-fiber sprouting into the pain layers (laminae I and II) may occur. When nerves that do not normally transmit pain sprout into these more superficial regions of the dorsal horn-regions where the first synaptic relay in pain transmission usually occurs-pain may result from non-noxious stimuli. Regeneration also causes sensory disorganization, so that the normal somatotropic organization of inputs becomes disordered (“spreading”).

Another central change that contributes to the development of allodynia is the loss of inhibitory controls projecting to the superficial spinal cord dorsal horn. This occurs when segmental inhibitory interneurons (mediated by neurotransmitters like γ-aminobutyric acid [GABA], glycine, and endogenous opioids [enkephalins]) or descending inhibitory pathways (mediated by neurotransmitters such as serotonin and norepinephrine) decrease their function. Because this inhibition normally acts as a spinal “gate” for sensory information, reduced inhibition increases the likelihood that the dorsal horn neuron will fire spontaneously or more energetically to primary afferent input. Thus, allodynia may result from any of these three central mechanisms for stimulus-evoked pain: central sensitization, reorganization, or loss of inhibitory controls.

Stimulus-Independent Pain

Stimulus-independent pain, or spontaneous pain occurs without provocation, so symptoms can occur constantly or at any time. Paresthesias and dysesthesias can originate peripherally via ectopic impulses along the Aβ, Aδ, and C fibers, arising as spontaneous activity due to processes such as damaged (“leaky”) sodium channels that accumulate along affected nerves, causing a drift toward threshold potential. Paroxysmal shooting or electrical pain (once thought to distinguish ectopic activity in myelinated fibers) as well as continuous burning pain (still thought to be caused by activity in unmyelinated nerves), most likely occur from ectopic or ephaptic discharges arising in any type of fiber. Stimulus-independent pain may also occur as a result of reduced inhibitory input from the brain or spinal cord.

In most neuropathic pain syndromes, stimulus-independent pain occurs along with stimulus-evoked pain; for example, spontaneous burning pain and mechanical allodynia present concurrently in complex regional pain syndrome (CRPS). In some syndromes, the activity at the site of injury seems to maintain the peripheral or central sensitivity in some fashion, and blocking the peripheral input may at least temporarily normalize the altered central processing. Thus patients' symptoms cease until peripheral input returns.

Assessing Pain

A careful history, physical examination, and a thoughtful use of testing are necessary to properly and fully define the putative mechanisms involved in a given neuropathic pain syndrome. Detailed medical and surgical histories are essential first steps in understanding pain etiology. A comprehensive physical examination allows the physician to integrate the patient's presenting symptoms and to begin to localize which elements of the neuroaxis are involved. It is particularly important to identify the location, quality, intensity, and pattern of pain. The neurologic examination uses simple bedside tests to assess the patient for the presence or absence of specific stimulus-evoked signs (Tables 10-2 and 10-3). Special attention should be paid to the sensory examination, especially searching for hypoesthesia (numbness) or hyperesthesia (hyperpathia or allodynia or both). A distinction between mechanical and thermal allodynia may have clinical relevance.

Testing of reflexes, a comprehensive motor examination, and autonomic examination are all essential to understanding neuropathies. Testing can complement and corroborate careful history and physical examinations and has the advantage of being quantitative, although all tests have their known limitations. A comprehensive list of diagnostic tests evaluating the motor, sensory, and autonomic systems is presented in Table 10-4. Some highly technical and invasive tests may be necessary, such as immunohistochemical staining of skin-punch biopsy specimens using antibodies specific for small diameter myelinated and unmyelinated peripheral nerves which can be used to quantify nerve fiber density in patients with peripheral neuropathy.

The physician should also be aware of any comorbid conditions affecting the patient's pain experience and quality of life, such as sleep disturbance, anxiety, or depression, which may help guide treatment decisions.


Once the patient has been thoroughly assessed and a putative mechanism (ie, a working diagnosis) has been developed, a treatment strategy should be formulated targeting each mechanism, with a primary goal of normalizing the underlying dysfunction. Sometimes, the underlying dysfunction can be corrected (eg, the compression


in compressive neuropathy can be relieved by postural correction or a toxic or metabolic insult, such as hyperglycemia). If a direct fix is not possible, then the source or nociceptive generator can be targeted using nonpharmacologic treatments (eg, ice to reduce inflammation) or medications (eg, normalizing pathologic sodium flux with sodium channel blockers).

Table 10-2. Bedside Tests for the Assessment of Allodynia.

Stimulus-evoked sign: Allodynia
Definition: Normally nonpainful stimulus evokes a painful sensation.
Control: Identical stimulus in unaffected skin does not evoke pain.



Pathologic response

Mechanical dynamic

Stroking skin with a brush, gauze, or cotton applicator

Sharp, burning, superficial pain

Mechanical deep somatic

Manual light pressure at the joints

Deep pain at the joints

Mechanical punctate

Light manual pinprick with a sharpened wooden stick or stiff von Frey hair

Sharp superficial pain

Mechanical static

Manual light pressure of the skin

Dull pain Burning pain

Thermal cold

Contact skin with objects at 20°Ca

Painful, often burning, temperature sensation

Thermal warm

Contact skin with objects at 40°Ca

Painful burning temperature sensation

a As a control, also contact skin with object at skin temperature.

  1. Nonpharmacologic Treatment

Patients with neuropathic pain often require a significant reduction in their pain, but traditional medical treatments, including nerve ablation, nerve blocks, and medication management, achieve only partial success in these goals. Clinicians understand that the efficacy of these interventions is sometimes limited; they also understand the tremendous direct and indirect costs in terms of pain and suffering, health care expenditures, and quality of life, not to mention the costs to society in lost productivity and vocational disability.

In attempts to address the associated emotional, social, and vocational sequelae of chronic pain, interdisciplinary and multidisciplinary pain management programs have been developed (see Chapter 6). Interdisciplinary programs are designed to help patients learn to cope more effectively with pain and help them maintain the highest possible functional level in order to alleviate pain and suffering. These interdisciplinary pain management programs have developed and introduced nonpharmacologic treatments that have actually had a dramatic impact on the biomedical issues.

More than just a physical sensation, a patient's pain is also the emotional, cognitive, and behavioral reactions to that sensation. Patients who have chronic neuropathic pain may become more disabled than their physical impairments can explain. Numerous studies have shown that a one-to-one correspondence does not necessarily exist between the amount of tissue damage and the person's experience of pain. Pain can be aggravated




by psychosocial factors, such as emotional state, previous experience, secondary gain, and expectations. Untreated chronic neuropathic pain can result in needless personal suffering for the patient; excess disability; co-morbid emotional problems, including increased risk of suicide; overuse or misuse of psychoactive medications; increased medical care utilization; iatrogenic complications secondary to inappropriate procedures, interventions, and surgeries; and increased economic and social costs. Complex, intractable neuropathic pain clearly requires an interdisciplinary approach that addresses psychosocial as well as biologic factors and that focuses on functional restoration in all areas of life.

Table 10-3. Simple Bedside Tests for the Assessment of Hyperalgesia.

Stimulus-evoked sign: Hyperalgesia
Definition: Normally painful stimulus evokes a more intense painful sensation.
Control: Identical stimulus in unaffected skin evokes a less painful sensation.



Pathologic response

Mechanical pinprick

Manual pinprick of the skin with a safety pin

Sharp superficial pain

Thermal cold

Contact skin with coolants, such as alcohol (evaporation) or cold metal

Painful, often burning, temperature sensation

Thermal heat

Contact skin with objects at 46°Ca

Painful burning temperature sensation

a As a control, also contact skin with object at skin temperature.

Table 10-4. Neurologic Tests Used in the Diagnostic Assessment of Neuropathic Pain.

Type of neurologic test

Functions evaluated

Fibers examined

Anticipated findings

Motor System

Electromyography (EMG) and motor nerve conduction studies (NCS)

Motor nerve conduction velocity; compound muscle action potential amplitude

Efferent large myelinated motor axons

Velocity and amplitude decreased with reduction in number of large myelinated motor axons or with interruption in myelination

Sensory NCS

Sensory nerve conduction velocity and action potential amplitude

Afferent large myelinated sensory axons (Aβ fibers)

Velocity and amplitude decreased with reduction in number of large myelinated sensory axons

Sensory System


Mechanical pressure and threshold tolerance

Aδ and C fiber activity arising from nociceptors and Aβ fiber activity arising from mechanoreceptors

Lower threshold tolerance or suprathreshold response to stimuli


Presence of ectopic impulses velocity and action potential amplitude

Single fiber activity arising from nociceptors (Aδ and C fibers) and mechanoreceptors (Aβ fibers)

Ectopic impulse generation along sensory axons

Quantitative sensory testing (Peltier device)

Sensory and pain threshold after stimulus with cool and warm temperature

Aδ and C fiber activity arising from nociceptors and mechanoreceptors

Lower threshold and suprathreshold response to stimuli

Vibrometer (QST)

Vibration perception thresholds

Aβ fiber activity arising from mechanoreceptors

Increase thresholds

Von Frey hairs

Mechanical pressure and threshold tolerance

Aδ and C fiber activity arising from nociceptors and Aβ fiber activity from mechanoreceptors

Lower threshold and tolerance or suprathreshold response to stimuli

Autonomic System

Heart rate

Heart rate variation in response to deep breathing

Autonomic efferent parasympathetic axons (eg, vagus nerve)

Less variation seen with polyneuropathy affecting vagal function

Quantitative sudomotor axon reflex test (QSART)

Sweat gland response to stimulation

Sympathetic postganglionic sudomotor axons

Excess or persistent sweat with reduced latency or reduced sweat volume consistent with peripheral neuropathy

Skin temperature and blood flow measurements with thermistor, thermography, and laser Doppler

Comparison of skin temperature of involved extremity to asymptomatic extremity

Sympathetic postganglionic vasoconstrictor axons

Early, warmer skin on involved side from vasodilatation; later, cooler skin from vasoconstriction

Role of Physicians & Nurses

The physician often coordinates patient care, determines diagnoses, identifies patient needs for the team members, and supervises overall patient needs and treatment strategies, either on a team or in a single provider practice. If also acting as a team leader, the physician ensures that the team remains focused on the patient's global level of functioning.

The physician and the nurse must provide reassurance regarding the safety and efficacy of physical and occupational therapy intervention. This is perhaps the most important role the physician and the nurse play for both staff and patient. Their reassurance is particularly important for patients with chronic neuropathic pain. These patients, being very somatically focused as a result of their chronic pain, may react with substantial apprehension to minor changes in physical sensations and patterns. The reassurance of physician and nurse can both inspire the patient to continue with the appropriate care of their pain management regimen and protect the patient from needless diagnostic work-ups and medical interventions.


Either in the context of a team or in private practice, a working relationship with a cognitive-behavioral psychologist is essential. Neuropathic pain creates stress; it can also be aggravated by stress as well as other psychological factors. Pain psychologists can ascertain whether or not psychological factors are contributing to excess pain and disability. Because certain coping strategies are more effective than others in managing pain, they can determine the type of coping strategies best suited for each patient and offer reliable forecasts for each strategy's efficacy.

Psychologists can also provide education, counseling, and training in cognitive-behavioral pain management techniques. Because depression and anxiety levels are high among patients with chronic pain, psychological evaluation and therapy should be a component of any comprehensive neuropathic pain treatment program. Increasing patient awareness about the stress model of chronic pain (eg, how stress, emotional distress, muscle tension, and deconditioning can increase pain) can be an effective motivator for getting patients to invest in the treatment process.

To achieve effective pain management, securing the patient's understanding and acceptance of the self-management approach is a paramount step in psychological pain management. Patient assent to such a model typically leads to increased motivation, increased dedication to pain management techniques, and an increased expectation of pain management success. Without acceptance of self-management principles, minimal progress is to be expected.

Several relaxation techniques have been shown to be useful in helping patients manage neuropathic pain, including progressive muscle relaxation, controlled (or diaphragmatic) breathing, imagery, autogenic training, biofeedback-assisted relaxation, and hypnosis. Progressive muscle relaxation is a technique that involves the methodical tension and relaxation of sets of voluntary muscles that travel through the body until the patient's whole body is relaxed. It is easily adapted for patients who are unable to tense certain muscle groups. Controlled breathing, often combined with progressive muscle relaxation, involves measured (8 breaths per minute) diaphragmatic breathing to stimulate a relaxation response.

Biofeedback-assisted relaxation provides the patient with feedback about physiologic responses that indicate a relaxed state while the patient uses both progressive muscle relaxation and controlled breathing. Feedback included data about such physiologic responses as galvanic skin response, fingertip surface temperature, muscle tension (with surface electromyography [EMG] biofeedback), heart rate, breathing rate, or electroencephalogram (EEG) responses. Biofeedback has been shown to be effective in a wide range of pain conditions, including neuropathic pain.

Hypnosis can be considered a combination of relaxation, distraction (see below), and suggestion (or placebo effect). It is “highly focused attention during which the alteration of sensations, awareness, and perceptions can occur.” Many studies indicate that hypnotic techniques assist in reducing both acute and chronic pain, including neuropathic pain. Because many patients object to hypnotic techniques on religious grounds or because they stand opposed to what they consider a loss of control, some clinicians prefer to use the term “imagery.”

Cognitive thinking techniques can aid patients who react to pain episodes with intense anxiety and episodes of catastrophizing. Some patients entertain catastrophic thoughts such as the worry that the pain will never cease,


that it will worsen, that they cannot withstand it, that they will have a stroke or heart attack because of the pain, and other such fears. Catastrophic thinking has demonstrated links to increased pain and decreased coping. Psychologists can assist patients in identifying their catastrophic thoughts and coach them in the use of techniques like thought-stopping and cognitive restructuring. These techniques help stem the flow of catastrophic thoughts, and patients can subsequently be taught how to replace them with more adaptive ways of thinking. Cognitive thinking techniques have been shown to be effective in helping patients manage their pain.

Other studies have shown that distraction is another valuable technique for pain management. Distracting activities, such as music and sensory techniques, can all be helpful in the short term.

Emotional distress techniques, including treatment of anger, anxiety, and cognitive-behavioral treatment of depression, are used in conjunction with antidepressants to break the pain-distress-pain cycle. Some antidepressants also have a well-known analgesic effect, particularly for neuropathic pain. Stress management training is another essential element of pain management programs, because many patients lack fundamental proficiency in problem solving, communication, and assertiveness, skills that they will need to return to full functioning. Deficiencies in these areas may be a risk factor for job dissatisfaction and a major risk factor for chronicity.

Family and friends sometimes reinforce pain behaviors and activity avoidance by discouraging a return to functioning and encouraging the sick role, most often in the belief that they are acting in the patient's best interest. Family counseling sessions endeavor to assist family reduction of reinforcement for pain behaviors. Through counseling, the family learns to encourage and buttress a patient's constructive pain management coping techniques and discourage detrimental behaviors.

Depression, anxiety, stress, and other psychological factors have been shown to aggravate some types of neuropathic pain, including phantom limb pain. Some patients state that stimuli that remind them of their injury or loss can trigger pain episodes; other triggers include emotional distress and stress in general. Also, patients who endure a traumatic injury are susceptible to post-traumatic stress disorder and persistent pain.

Treatments for stress and emotional distress may possibly assist in patient pain modulation, but research results are inconclusive. One patient survey indicated that psychological interventions, including antidepressants, did not reduce pain, whereas other studies reported that distraction and relaxation did relieve phantom pain in some patients.

In particular cases where psychiatric diagnoses are ubiquitous, such as multiple sclerosis, psychological interventions can help improve emotional functioning substantially. Two studies report that cognitive-behavioral therapy reduced depression and anxiety when compared with a standard treatment control group. One case study reported that hypnosis resulted in less pain.

Gonzales, in a review of the pathophysiology and treatment of central pain including poststroke pain, stresses the importance of providing psychological intervention in cases in which the pain is resistant to pharmacologic treatment. In the review, Gonzales points out that this population has a high incidence of depression and an increased risk of suicide.

Physical Therapy

Training in both active and passive modalities focused on correcting or modulating the factors that may be contributing to neuropathic pain, such as poor posture, spasm, contractures, or bony ankylosis, is provided by the physical therapist. The goal of physical therapy is to teach the patient stretching and strengthening exercises that enhance flexibility in those muscle groups that would tend to compress the nerve and to simultaneously strengthen the muscle groups that would tend to relieve the compression. For example, the physical therapist working with a patient who has low-back radiculopathy and hyperlordosis might introduce suitable stretching and lumbosacral stabilization exercises; under such instruction, the patient might be able to sustain improved posture and keep the foramen in a more open arrangement, thereby relieving radicular compression. The physical therapist enhances stretching programs by training the patient to apply heat before sustained stretching and ice afterward. Strengthening typically consists of a regimen of active and passive range of motion exercises, eventually progressing to isometric exercises. Later, isotonic exercise and, ideally, supervised weight training may be useful, provided they do not aggravate the compression.

Thermotherapies, ultrasound, and other passive therapies are of limited usefulness in treatment of neuropathic pain. Neuromuscular facilitation and other manipulation therapies can be particularly valuable, primarily in assuaging pressure on a compressed nerve. Physical therapists should be consulted in decision-making when orthotics are indicated. Although bracing may be indicated in the initial stages of rehabilitation, it is not commonly useful as a long-term modality because it creates dependency and causes atrophy in supporting muscles.

Even though they are commonly used in the management of painful neuropathies, electrostimulation methods have not been sufficiently investigated. It remains to be seen whether recent techniques (burst technologies, strength duration variation modalities, or high-frequency, low-amplitude stimulation) improve upon traditional square-wave transcutaneous electrical nerve stimulation. Optimal electrode placement (eg, over a


motor point, along the course of nerves, at acupuncture points) remains similarly unresolved. Until the results of more comprehensive studies are available, a patient, flexible, and pragmatic methodology developed by an experienced physical therapist, will provide the best outcome. A fear of general or specific movements may have developed in some patients with neuropathic pain, who expect motion to be painful. These fears may create patient noncompliance with physical therapy recommendations. The physical therapist can work with the psychologist to identify any such barriers to compliance and provide appropriate psychological treatment.

Occupational Therapy

Workstation assessment, ergonomic correction, and orthotics form the core of neuropathic occupational therapy. In addition, assessment and modification of sleep postures, activities of daily living, and recreational activities are equally important. The identification and application of specific workplace modifications can be very advantageous in some cases, such as wrist padding for keyboard operators. Common sense modification of the existing workplace may be sufficient, but special equipment must sometimes be identified, adjusted, and installed. A work site visit is commonly necessary to suitably evaluate job tasks, to devise a work simulation regimen, and to ensure that all modifications and devices are properly applied.

Occupational therapy should be the principal method for treating compressive-type neuropathies, particularly in a vocational setting. Poor ergonomic work environments may create or aggravate problems like carpal tunnel and repetitive motion or occupational microtrauma-type neuropathies.

The occupational therapist also teaches the patient exercise and flexibility procedures to treat contractures in more serious neuropathic lesions, and prevent their development in other lesions. The occupational therapist should work in conjunction with the orthotist to augment functional recovery, particularly in the context of vocational rehabilitation. Specific braces and tensioning appliances and sometimes serial casting may help prevent or treat contractures in certain patients.

The occupational therapist's role in the treatment of CRPS/reflex sympathetic dystrophy is pivotal. Patients who have this challenging syndrome identify optimal functional restoration as their primary goal, occupational therapies are often fundamental. Scrubbing and stress loading in particular seem to be crucial to optimal outcome, although it is still not known whether the normalized proprioceptive input or the motor output is most beneficial; it is probable that some measure of both are required. Desensitization using a range of textured surfaces or textiles (ie, going from light brushes to silk cloths to rough toweling) can be vital in conditioning patients for other aspects of therapy and functional rehabilitation. Hydrotherapy and contrast baths may also be essential. In addition, the occupational therapist is able to recommend specialized garments, especially those for edema control. Edema measurement by volumetrics as an indication of improvement and specialized methods of manual lymph drainage have been used effectively.

Vocation Counseling

The vocational counselor intervenes in cases where the patient is unable to return to previous employment. In some cases, job dissatisfaction and anger may be so considerable that the best resolution for patient and employer may be in helping the patient (and employer) acknowledge the need to locate alternative employment and develop methods for doing so. Practical case management can help avoid expensive and extended efforts to return the patient to work that he or she may only be able to handle physically but not emotionally.

  1. Pharmacotherapy

The best clinical approach to applied pharmacology currently incorporates empiric observation and identification of possible mechanisms of the neuropathic lesion (“targets”), and then uses the best available pharmacologic information to match these potential disease mechanisms with putative drug mechanisms. Although monotherapy is the ideal approach, rational polypharmacy is often pragmatically used. Rational polypharmacy requires an informed conjecture regarding the underlying pain mechanisms, and a rational combination of drugs that act at different sites in the neuroaxis to interfere and modulate the diagnosed mechanisms. Two basic classes of medications should be considered: prophylactic drugs (used on a regular basis) to manage pain and other symptoms, and abortive drugs (used as needed) for pain or symptom flares. The prophylactic drugs will often be selected by the presentation of the patient's symptoms. For example, if a patient is extremely depressed or anxious and has insomnia, the clinician may choose a tricyclic antidepressant with significant analgesic, sedative, and anxiolytic properties.


  1. Tricyclic Antidepressants

Drugs that are thought of traditionally as antidepressants may be used to treat neuropathic pain because they are analgesic as well. It should be noted, however, that randomized controlled trials evaluating the efficacy of these drugs in alleviating neuropathic pain or reducing specific neuropathic pain symptoms are currently limited. Drugs that have been shown in clinical trials to have a


beneficial impact on specific neuropathic pain symptoms are listed in Table 10-5 and dosages for selected agents are presented in Table 10-6.

Table 10-5. Drugs that have Shown Efficacy in Certain Neuropathic Pain Symptoms.




Ketamine IV or IM
Lidocaine IV Morphine IV

Burning pain

Phenytoin IV


EMLA cream
Lidocaine IV

Shooting, lancinating pain

Phenytoin IV

Tricyclic antidepressants are accepted choices in neuropathic conditions and a meta-analysis of randomized clinical trials indicate their efficacy in treating neuropathic pain. One of these studies reported that 30 of every 100 patients with neuropathic pain who received antidepressants obtained at least a 50% pain relief.

Table 10-6. Dosing for Selected Medications.


Total daily dose




100-1000 mg/d

bid to qid


900-3600 mg/d



150-500 mg/d




10-200 mg/d



10-200 mg/d

qd to bid


37.5-340 mg/d

bid to tid



0.25-2 mg/kg/d

Continuous IV


0.25-0.5 mg/kg/

q3h(IV or IM)




The antihyperalgesic effects of tricyclic antidepressants may be related to enhancement of noradrenergic descending inhibitory pathways and partial sodium channel blockade, mechanisms that are independent of their antidepressant effects. Moreover, the sodium blocking effect may be the more potent mechanism in this class (which technically includes carbamazepine). When pain is independent of a stimulus, central mechanisms can be reasonably targeted, because these mechanisms cause sensitization of primary somatosensory afferents. Tricyclic antidepressants that cause a balanced inhibition of reuptake of both serotonin and noradrenaline (eg, imipramine, amitriptyline) may be more effective for painful polyneuropathy than those with relative selectivity for noradrenaline reuptake (eg, desipramine).

The responsible clinician must have a repertoire of several tricyclic/quadricyclic drugs, because specific drugs have specific side effects associated with them, and these may sometimes be used to the patients' advantage. For instance, a patient who is in moderate pain, depressed, overweight, and hypersomnolent with psychomotor retardation could be prescribed a tricyclic antidepressant with more noradrenergic selectivity (eg, desipramine), which may be activating and can cause some anorexia, rather than a sedative agent associated with weight gain.

  1. Selective Serotonin Reuptake Inhibitors

The performance of selective serotonin reuptake inhibitors for neuropathic pain is not impressive. Certain newer antidepressant agents, including venlafaxine and mirtazepine, show some promise in clinics and have the advantage of a different, more benign side effect and toxicity profile.


The anticonvulsant compounds are some of the best-studied drugs for neuropathic pain, and there is substantial evidence for their efficacy based on meta-analyses and randomized clinical trials. Many of the newer anticonvulsants block sodium and calcium channels, which produces a decrease in neuronal excitability.

  1. Gabapentin

In fact, gabapentin, widely used for neuropathic pain, first drew the attention of the research community when its successful use in the treatment of CRPS was published in an anecdotal report. The mechanism of action of gabapentin (and now pregabalin) was thought to work primarily through enhancement of endogenous GABA systems that function in pain modulation (but it is not a GABA agonist). New evidence suggests this


may not be a primary mechanism, and current theory focuses on “synaptosomes” in the presynapse. In addition, gabapentin may have some effect on the suppression of excitatory amino acids, such as glutamate. In several large randomized clinical trials, gabapentin and pregabalin have demonstrated significant efficacy in postherpetic neuralgia and diabetic peripheral neuropathy.

  1. Membrane Stabilizers

Phenytoin and other membrane-stabilizing antiepileptic drugs (sodium channel blockers) may have some use in neuropathic pain, particularly in cases where there is a potential role for ectopic activity in the generation of pain. Carbamazepine is a membrane stabilizer and has a traditional and perhaps clinically important place in the treatment of neuropathic pain, especially trigeminal neuralgia. Oxcarbazepine may be as effective as carbamazepine and has fewer side effects, according to the results of an open-label trial of patients with painful diabetic neuropathy, but this was not borne out in the pivotal trial.

  1. Other Anticonvulsants

Many other anticonvulsants such as levetiracetam, topiramate, lamotrigine, and zonisamide have modestly compelling evidence suggesting they may be useful in neuropathic pain, and several large pivotal trials are now in progress.

Anti-Inflammatory Drugs

NSAIDs, corticosteroids, and free-radical scavengers are infrequently used in some neuropathic pain conditions, particularly those associated with considerable inflammation. In neuropathic pain, neuroimmune interactions may occur and provide the rationale for immunosuppressive therapy. Animal studies with cyclosporine, thalidomide, and methotrexate support this premise. NSAIDs inhibit cyclooxygenase (COX) and prevent the synthesis of prostaglandins, which induce inflammation and perhaps peripheral hyperalgesia. In addition to the peripheral anti-inflammatory action of NSAIDs, another proposed mechanism is the blockade of spinal nociceptive processing. However, in several clinical trials of neuropathic pain, NSAIDs have shown mixed results. Ketoprofen has detectable antibradykinin effects as well as the standard antiprostaglandin effect. Randomized clinical trials for COX-2 inhibitors have not been conducted. Corticosteroids can be particularly useful in the early/acute phases of certain types of neuropathic pain (such as radiculopathy) in which significant inflammation exists. A short course of corticosteroids may be indicated, but longer courses have a questionable risk-benefit ratio and numerous contraindications.

Free-radical scavengers (ie, dimethylsulfoxide [DMSO] and vitamin C; see below) may reduce the concentration of reactive oxygen species, which are an acknowledged agent in inflammatory processes that may be involved specifically in neurogenic inflammation.


Opioids may be useful, especially in acute stages, but their use for chronic pain management remains somewhat controversial. Several studies of opioids for neuropathic pain suggest their efficacy. In general, neuropathic pain appears to be less responsive to opioids than nociceptive pain; neuropathic pain thus requires higher doses, leading to an increased risk of side effects. To avoid these complications, a strategy that entails the use of nonopioid medications for prophylaxis and reserves the use of opioids for crisis management is indicated. The use of opioid therapy can be linked to increased function, and the use of an acute or subacute opioid protocol is therefore often used to allow the patient to begin to progress in nonpharmacologic therapies.

NMDA Receptor Antagonists

  1. Ketamine and Amantadine

NMDA receptor antagonists (eg, MK-801, ketamine, amantadine) have been considered for management of windup, sensitization, and tolerance to opioids but have been proven too toxic at effective dose levels for regular use. Ketamine has been evaluated in a small study of cancer patients with neuropathic pain who are unresponsive to morphine, and there is considerable ongoing interest in high-dose inpatient protocols of ketamine for CRPS, as well as interest in lower dose inpatient or outpatient protocols. Several delivery systems are also being studied. Amantadine has been evaluated in cancer patients with neuropathic pain and in patients with chronic neuropathic pain, with some support.

  1. Dextromethorphan

Plain dextromethorphan in pill form may be better tolerated than some of the other NMDA antagonists and may enhance the effect of other medications, specifically opioids. A study in rats demonstrated that combined oral administration of morphine sulfate and dextromethorphan can prevent the development of tolerance to the antinociceptive effects of morphine sulfate. However, dextromethorphan is not effective in low doses, is toxic in doses that are high enough to exhibit efficacy, and so far ineffective when taken on its own.

Other Drugs

Clonidine has been evaluated in its various forms: orally for postherpetic neuralgia, intrathecally in a rat


neuropathic pain model, and as a transdermal patch for diabetic polyneuropathy. Unfortunately, a large, randomized clinical trial for neuropathic pain conditions showed no overall efficacy. According to a recent systematic review, data regarding clonidine is not convincing.

Mexiletine, an orally administered antiarrhythmic drug with local anesthetic properties, has been used in some clinics to treat neuropathic pain, but results of a randomized trial in HIV-associated neuropathy exhibited no efficacy. In addition, mexiletine has many problematic side effects.

Systemic lidocaine, administered either intravenously or subcutaneously, may be effective for neuropathic pain but only provides temporary improvement in most trials.

Topical Treatments

Topical treatments for neuropathic pain differ from trans-dermal medications (eg, the fentanyl patch, transdermal clonidine). Topical treatments deliver medication locally to the affected skin and soft tissues. Topical treatments for neuropathic pain include the lidocaine patch 5%, eutectic mixture of local anesthetics (EMLA) cream, capsaicin, and DMSO.

The lidocaine patch is a non-woven patch containing 5% lidocaine. It is FDA-approved for the treatment of postherpetic neuralgia, and is used increasingly for other neuropathic pain conditions. The lidocaine patch may be useful in some very local or focal neuropathic pain phenomena, including allodynia.

Capsaicin, a vanilloid compound found in chili peppers, causes activation and the dying-back of nociceptive nerve endings by allowing unchecked cation influx. At the site of application, it often induces a painful burning sensation. In a randomized clinical trial, topical capsaicin was effective for the treatment of postherpetic neuralgia. In our experience, however, topical capsaicin has proven to be intolerably painful early on, messy, and associated with very poor compliance.

DMSO is a free radical-scavenging agent. In a high-quality study evaluated in a systematic review, DMSO (50% cream for 2 months) did not show significant pain reduction in persons with CRPS compared with placebo.

Interventional Treatments

Although nerve blocks have been used by physicians for generations for the treatment of neuropathic pain, there is actually very little evidence to support this approach. It also is counterintuitive in the treatment of chronic pain of any sort. Certainly, the transmission of pain can be stopped by local anesthetics, but they all have relatively short half-lives, and any beneficial effect soon wears off. Regional corticosteroid injection can be very helpful in inflammatory conditions and may be of some use in neurogenic inflammation but are not helpful per se in neuropathic pain. The principal utility of blocks is diagnostic, and the specific nerve involved in generating the neuropathic pain can be identified. They may also have some use in providing a pain free “window of opportunity,” so that patients can vest in uncomfortable nonpharmacologic therapy, such as physical therapy.

Managing Specific Pain Syndromes

Hyperalgesia & Allodynia

Because hyperalgesia probably depends on peripheral as well as central mechanisms, treatment can logically be initiated with local therapy. These therapies (ice, sodium ion block) include topical anesthetic agents, such as EMLA cream or lidocaine impregnated patches. Topical agents have been used with variable success in patients with neuropathic pain; however, these results include treatment of a variety of additional conditions besides hyperalgesia. One study monitored the impact of topical EMLA on patients with only hyperalgesia and reported significant efficacy. Moreover, some studies have found that a lidocaine patch has demonstrated efficacy in patients with postherpetic neuralgia, but these studies were not specifically designed to assess hyperalgesia. The 5% lidocaine patch has been approved by the FDA for the treatment of neuropathic pain in patients with postherpetic neuralgia.

The putative mechanism for the efficacy of capsaicin is the selective stimulation of unmyelinated C fiber afferent neurons, which causes the release of substance P. Prolonged application is thought to deplete substance P stores (see above) from sensory nerve endings; this eventually prevents or reduces the transmission of pain. However, repeated applications of capsaicin (3 to 4 times daily for 4 to 8 weeks) are required before clinical effectiveness can be assessed, and it is not always well tolerated.

Many centrally acting medications have been recommended for the management of allodynia. Local anesthetic blocks are effective in temporarily eliminating thermal and sometimes mechanical allodynia; part of this success may result from their ability to inhibit the continued nociceptive input that initiates and maintains central sensitization-one of the causes of allodynia. Topical lidocaine has been used successfully to treat patients with postherpetic neuralgia experiencing allodynia. The use of lidocaine gel or a 5% patch were both significantly more effective than placebo in relieving pain with only minimal increases in lidocaine serum concentrations.

Clinical trials in patients with painful diabetic peripheral neuropathy and postherpetic neuralgia have demonstrated that tricyclic antidepressants are effective in relieving neuropathic pain, but these studies do not differentiate between allodynia and stimulus-independent


symptoms such as burning and electrical pain. In addition to being excellent sodium channel blockers (peripheral), the tricyclic antidepressants are known to inhibit the reuptake of serotonin and norepinephrine (central). The analgesic properties of these drugs may be related at least partially to the restoration of inhibitory controls.

Although its mechanism of analgesic effect has not been specifically determined, experimental data suggest that gabapentin acts at multiple central sites. Gabapentin binds with high affinity to a unique site in the brain that is associated with an auxiliary subunit of calcium channels. Gabapentin most likely modifies and modulates first-and second-messenger calcium currents and ultimately may cause a decrease in firing of the transmission cell or a decrease in the release of certain monoamine neurotransmitters. These observed and hypothesized mechanisms might underlie the effect of gabapentin on allodynia. Gabapentin likely also works presynaptically to block synaptosome function, as has been demonstrated in the related drug pregabalin. In a pilot study of patients with various peripheral and central neuropathic pain syndromes, Attal and colleagues demonstrated that gabapentin (up to 2400 mg/d) was effective in reducing tactile and cold allodynia. Gabapentin had no effect on normal mechanical and thermal pain thresholds, suggesting a lack of direct antinociceptive effect. Other GABA-enhancing drugs, including baclofen (a GABAB agonist), have been shown to be effective in reducing tactile allodynia in rat models.

Traditionally, clinicians have been reluctant to treat pain with opioid analgesics because of multiple concerns, including the threat of diversion, psychological dependency, and cognitive impairment. These fears are variably overestimated in usual clinical practice and the approach has been changing; the clinical use of opioids is becoming more acceptable and perhaps they are now overprescribed. Although opioids may not be as effective in neuropathic pain as in nociceptive conditions, certain evidence supports the short-term use of opioids in patients with allodynia. In a randomized, double-blind, placebo-controlled trial, high-dose morphine (mean 19.2 mg infused over 1 hour) was effective in relieving allodynia in 11 of 19 patients with postherpetic neuraglia. Although adverse effects were common, respiratory depression and excessive sedation were not observed. When therapeutic response to opioids is suboptimal, other routes of administration should be undertaken or combination therapy with other analgesics, such as tricyclic anti depressants, should be considered. No trials of sufficient length allow comment on the comprehensive ramifications of long-term opioid therapy for central sensitization.

Allodynia may also be treated with drugs that antagonize the NMDA receptors responsible for central sensitization. The NMDA antagonist ketamine has demonstrated effectiveness in alleviating allodynia in patients with postherpetic neuralgia, chronic posttraumatic pain, and chronic neuropathic pain. NMDA antagonists have also been used in patients with phantom limb pain (ketamine), orofacial pain (ketamine), surgical neuropathic pain (amantadine), diabetic neuropathy (dextromethorphan), and postherpetic neuralgia (dextromethorphan), though effects on allodynia were not specifically evaluated.

Ectopic Activity at a Neuroma

Theoretically, the effects of the “neuroma sign” (Tinel sign) can be at least partially ameliorated by drugs that block ectopic firing secondary to accumulation of dysfunctional (“leaky”) pathologic sodium channels and dysfunctional sodium pumps. To date, supporting data are limited to animal studies. These data show that intravenous lidocaine, tocainide, and mexiletine given in subanesthetic concentrations stop the firing of spontaneously active fibers in the neuroma without blocking conduction. Carbamazepine and phenytoin may also be effective. Some studies suggest that, theoretically, other sodium channel blockers (such as lamotrigine or topiramate) could be useful, but the data are currently inconclusive. All of these drugs have additional and potentially salient effects.

There is some evidence from the rehabilitation literature to support the use of these agents for dramatic neuroma/ectopic activity in such diagnoses as postamputation pain.

Treatment of Stimulus-Independent Pain

Sodium channel blockers are the mainstay of treatment for chronic neuropathic pain syndromes arising from ectopic discharges in nociceptive fibers. Carbamazepine is traditionally the treatment of choice for the shooting pain of trigeminal neuralgia, and it was first proven effective for this condition in the early 1960s. There are many side effects, such as hematopoietic and hepatic toxicity and rash. When skin rash develops, oxcarbazepine can often adequately substitute for carbamazepine, or patients could simply begin with oxcarbazepine, which appears to have a lower incidence of skin rashes than carbamazepine. This is an example of a newer, second-generation drug that has a much improved risk-benefit ratio because the new compound is not metabolized to the toxic epoxide.

Like carbamazepine, lamotrigine has demonstrated higher efficacy than placebo in alleviating the sharp, shooting, or stabbing pain of trigeminal neuralgia when administered in conjunction with phenytoin or carbamazepine to treat refractory cases. However, in a separate placebo-controlled study, lamotrigine (200 mg daily) showed no effect on pain in 100 patients with neuropathic pain of various causes. In still another placebo-controlled trial, a single dose of phenytoin


(15 mg/kg infused intravenously over 2 hours) significantly relieved shooting pain in patients experiencing acute flares of neuropathic pain. In addition, tricyclic antidepressants may be effective for treating shooting pain, possibly because of their sodium channel blocking properties.

Several trials have demonstrated that tricyclic antidepressants are also effective in alleviating burning pain. Drugs evaluated include amitriptyline (2.5 to 150 mg/d), desipramine (12.5 to 250 mg/d), and imipramine (25 to 350 mg/d). However, side effects of tricyclic antidepressants include sedation and anticholinergic effects, which limit their usefulness. Gabapentin generated a moderate but significant relief of both continuous burning pain and paroxysmal (lancinating/shooting) pain.

Attal N et al. Effects of gabapentin on the different components of peripheral and central neuropathic pain syndromes: a pilot study. Eur. J. Neurol.1998;40:191 .

Harden RN. Chronic opioid therapy: another reappraisal. APS Bull. 2002; 12:1. McQuay H et al. Anticonvulsant drugs for management of pain: a systematic review. BMJ. 1995;311:1047 .