Central Modulation of Nociception

Nociceptive transmissions are modulated at the spinal level by both local neuronal circuits and descending pathways originating in the brainstem through the dorsal horns and the spinothalamic projections. Intrasegmental and intersegmental projections arising from cells located in the Rexed laminae I and II modulate both presynaptic and postsynaptic elements of primary nociceptive afferent terminals in the spinal cord. Activation of non-nociceptive afferent fibers may suppress nociceptive transmission in the dorsal horn. This is the major component of circuitry models referred to as the gate control theory of pain transmission. The development and widespread use of the spinal cord stimulator is based on this theory.

Descending inhibitory systems appear to have three functionally interrelated neurotransmitter mechanisms: the opioid, the noradrenergic, and the serotonergic systems. Opioid precursors and their respective peptides (β-endorphin, methionine [met]-enkephalin, leucine [leu-] enkephalin, and dynorphin) are present in the amygdala, hypothalamus, PAG, raphe magnus, and the dorsal horn. Noradrenergic neurons project from the locus caeruleus and other noradrenergic cell groups in the medulla and pons. These projections are found in the dorsolateral funiculus. Stimulation of these areas produces analgesia, as does the administration (direct or intrathecal) of α₂-receptor agonists such as clonidine (Khodayar et al., 2006). Many serotonergic neurons are found in the raphe magnus. These neurons send projections to the spinal cord via the dorsolateral funiculus. Administration of serotonin to the spinal cord produces analgesia, and pharmacological blockade or lesion of the raphe magnus can reduce the effects of morphine. The antinociceptive effects of antidepressants such as tricyclics and newer serotonin-norepinephrine reuptake inhibitors such as duloxetine and milnacipran are believed to reduce pain by increasing serotonin and norepinephrine concentrations in descending inhibitory pain pathways.

Opioid Receptors

Opioids are the core pharmacological treatment for acute pain. They act via receptors on cell membranes. Opioid receptors are coupled to G proteins and are thus able to effect protein phosphorylation via the second messenger system and change ion channel conductance. Presynaptically, activation of opioid receptors inhibits the release of neurotransmitters involved in pain, including substance P and glutamate. Postsynaptically, activation of opioid receptors inhibit neurons by opening potassium channels that hyperpolarize and inhibit the neuron.

Currently, there are five proposed classes of opioid receptors: µ, δ, κ, σ, and ε. µ Receptors are the main functional target of morphine and morphine-like drugs; they are present in large quantities in the periaqueductal gray matter in the brain and the substantia gelatinosa in the spinal cord. µ Receptors are also found in the peripheral nerves and skin. Activation of µ receptors results in analgesia, euphoria, respiratory depression, nausea, vomiting, and decreased gastrointestinal (GI) activity, as well as the physiological syndromes of tolerance and dependence. Two distinct subgroups of the µ receptors have been identified: µ₁, found supraspinally, and µ₂, found mainly in the spinal cord. The µ₁ receptor is associated with the pain-relieving effects of opioids, whereas µ₂ receptors mediate constipation and respiratory depression.

The δ receptor has similar central and peripheral distribution as the µ receptors. Studies have shown that δ-opioid agonists can provide relief of inflammatory pain and malignant bone pain. Meanwhile, peripherally restricted κ-opioid agonists have been developed to target κ-opioid receptors located on visceral and somatic afferent nerves for relief of inflammatory, visceral, and neuropathic chronic pain. The potential analgesic effects, combined with a possible lower abuse rate and fewer side effects than µ-receptor agonists, makes δ- and peripherally restricted κ-opioid receptor agonists promising targets for treating pain (Vanderah, 2010).

Peripheral Sensitization

Neurological systems may change their responses to various incoming stimuli. This process is called neuroplasticity and occurs in both the peripheral and central nervous systems. Peripheral sensitization may result in decreased threshold and increased responses after repeated stimuli to the peripheral nervous system. Most of the inflammatory mediators sensitize the peripheral nerve endings through posttranslational phosphorylation of the key receptor channels. For example, phosphorylation of tetrodotoxin-resistant voltage-gated sodium channels by protein kinase A and protein kinase C increases the sodium current to depolarize the nerve terminal, thereby producing greater excitation in the peripheral nerve and lowering the activation threshold of the neurons. A damaged C fiber therefore may begin to fire spontaneously. Changes in non-nociceptive afferents such as Aβ fibers also contribute to the process of sensitization. These fibers normally convey information about non-noxious mechanical stimuli. Following an injury, a pain-related neurotransmitter such as substance P may be released from terminals of Aβ fibers. These alterations may explain the pain induced by non-nociceptive stimuli, such as allodynia.

Central Sensitization

Central sensitization plays a major role in the development of neuropathic pain syndromes. First, postsynaptic depolarization in the spinal cord in response to afferent stimulation can induce removal of magnesium blockade in N-methyl-d-aspartate (NMDA) receptors such that glutamate now induces a depolarization upon receptor binding. This process is short lasting and is called wind up (Katz and Rothenberg, 2005). It is responsible for the temporal summation of inputs. The second set of changes is related to phosphorylation of the NMDA receptor, which is a key process for longer-lasting changes in the excitability of the dorsal horn neurons that produce central sensitization. This posttranslational modification of the NMDA receptors results in dramatic changes in excitability due to removal of the voltage-dependent magnesium block in the absence of cell depolarization and also to changes in channel kinetics, such as channel opening time. NMDA receptor activation allows calcium influx into the cell, which further augments signal transduction within the dorsal horn neurons by activating a number of intracellular signal transduction kinases. As a result, relatively brief C-fiber inputs initiate very rapid changes in membrane excitability. This manifests both as a progressive increase in excitability during the course of the stimulus (wind up) and as post-stimulus changes that may last for several hours (central sensitization).

Clinically, the real meaning of peripheral and central sensitization is the enhanced and prolonged pain perception to minor stimulations, or sometimes without peripheral stimulation. Once peripheral and central sensitizations are involved, the pain is usually more difficult to treat. It is now believed that peripheral and central sensitization may be involved in a wide variety of chronic pain conditions such as reflex sympathetic dystrophy, tension headache, carpal tunnel syndrome, pain after spinal cord injury (Carlton et al., 2009), and even in pain conditions previously thought to be mainly nociceptive in nature such as fibromyalgia, epicondylalgia, and osteoarthritis (Gwilym et al., 2009). The challenge to the clinician is that when trying to make a diagnosis for a pain patient, we should not only try to localize the pain source—as most clinicians always do—we should also factor in the role of peripheral and central sensitization and what the best treatment strategy will be in each case.

Trigeminal Neuralgia

Trigeminal neuralgia (TN), or tic douloureux, is characterized by paroxysmal lancinating attacks of severe facial pain. TN has an incidence of approximately 4/100,000, with a large majority of cases occurring spontaneously. Both genders experience TN, but there is a slight female predominance, and the diagnosis is most common over the age of 50. Classic TN is characterized by abrupt onset and termination of unilateral brief electric shock-like pain. Pain is often limited to the distribution of one or two (commonly the second and third) divisions of the trigeminal nerve. Trivial stimuli including washing, shaving, smoking, talking, and/or brushing the teeth (trigger factors) can evoke the pain. Some areas in the nasolabial fold and/or chin may be particularly susceptible to stimulation (trigger areas). In individual patients, pain attacks are stereotyped, recurring with the same intensity and distribution. Most TN patients are symptom free between attacks, and clinical examination is usually normal. Attacks of TN occur in clusters, and remissions can last for months.

The cause of TN pain attacks is unknown. Compression of the trigeminal nerve by benign tumors and vascular anomalies may play a role in the development of clinical symptoms. Studies of surgical biopsy specimens from TN patients who had presumed vascular compression demonstrate evidence of inflammation, demyelination, and close apposition of axons (leading to the possibility of ephaptic transmission between fibers). The ignition hypothesis of Devor proposes that a trigeminal nerve injury induces physiological changes that lead to a population of hyperexcitable and functionally linked trigeminal primary sensory neurons. The discharge of any individual neuron in this group can quickly spread to activate the entire population, resulting in a sudden synchronous discharge and a sudden jolt of pain characteristic of a TN attack.

The diagnosis of TN is based primarily on a history of characteristic paroxysmal pain attacks. The White and Sweet criteria are still commonly used worldwide (Box 44.1). In the majority of TN patients, the clinical examination, imaging studies, and laboratory tests are unremarkable (classic TN). In a smaller group, TN is secondary to other disease processes affecting the trigeminal system (symptomatic TN). Because a significant percentage of patients have symptomatic TN resulting from other disease processes, diagnostic MRI studies should be part of the initial evaluation of any patient with TN symptoms. Special attention should be paid to MS plaques, tumor, and subtle vascular anomalies that may be the source of root compression. Recent studies found that high-resolution three-dimensional (3D) MRI reconstruction and magnetic resonance cisternography may provide alternative tools to better identify the presence of neurovascular compression and even measure the volume of neurovascular compression at the cerebellopontine angle and predict the prognosis after initial treatment (Tanaka et al., 2009).

Carbamazepine is the first choice for treatment of TN; both controlled and uncontrolled studies confirm its clinical efficacy. Carbamazepine monotherapy provides initial symptom control in as many as 80% of TN patients. Of those initially responding to the drug, approximately 75% will continue to have long-term control of pain attacks. Controlled studies demonstrate that baclofen and lamotrigine are superior to placebo for treatments of TN. In the experience of many clinicians, baclofen is just as effective as carbamazepine and often better tolerated. A recent study found that oxcarbazepine may be effective for those who were unresponsive to the treatment of carbamazepine (Gomez-Arguelles et al., 2008). Pregabalin may also be potentially effective. If a patient is not satisfied with single medication therapy, adding another oral medication may offer additional benefits. Intravenous (IV) lidocaine or phenytoin could be effective for some severe refractory cases of TN. However, these treatments carry additional risks and require close cardiovascular monitoring. Opioid analgesics have not been proven effective for TN and should be avoided.

Posterior fossa exploration and microvascular decompression (MVD) is assumed to directly treat the cause of TN. However, this is a complex and invasive therapy with a possibility of death. With the availability of other less-invasive procedures, MVD is infrequently used and is only reserved for younger and healthier patients. Several studies demonstrated trigeminal radiofrequency rhizotomy successfully controls symptoms in over 85% of TN cases. The technique is minimally invasive. To heat the gasserian ganglion, a radiofrequency needle is inserted through the foramen ovale under the guidance of fluoroscopy. The procedure can be finished in less than 30 minutes in experienced hands. A few patients experience sensory loss and dysesthesia (analgesia dolorosa) in the distribution of the damaged trigeminal fibers with this procedure. Stereotactic radiosurgery (SRS) employs computerized stereotaxic methods to concentrate ionizing radiation on the trigeminal root entry zone. Several studies have demonstrated the high clinical efficacy and relative safety of this new technique. It is currently recommended as a first-line noninvasive surgical technique in many pain centers, especially for frail or elderly patients (Zahra et al., 2009).

Low Back Pain

Low back pain (LBP) is the most common condition seen in pain clinics. Approximately 60% to 80% of the U.S. population will experience back pain some time during life. Neurologists are often consulted for the diagnosis and treatment of LBP. It is critical for clinicians to appropriately examine the patients and make a diagnosis before treatment is rendered. Common causes of LBP include muscle strain, lumbar disk herniation, lumbar radiculopathy, lumbar facet joint syndrome, sacroiliac joint syndrome, and lumbar spinal stenosis.

Patients with acute muscle strain in the low back often have histories of acute injury. Physical examination may reveal tenderness or muscle spasms. Nonsteroidal anti-inflammatory drugs (NSAIDs), muscle relaxants, massage therapy, physical therapy, or acupuncture often provide effective pain relief. However, many times muscle pain in the low back is secondary to injuries in deeper tissues, such as lumbar disk herniation or lumbar radiculopathy.

Acute lumbar disk herniation after injury may cause severe LBP. Patients often complain of severe shooting or stabbing pains in the low back, with frequent radiation pain down the dorsomedial part of the foot when the L5 nerve root is involved, or the lateral part of the foot or the small toe when the S1 nerve root is involved. The straight leg raising test is often positive. Detailed neurological examinations may find decreased sensation to pin prick in the area innervated by L5 and/or S1 nerve root(s). Patient may also have mild weakness on the tibialis anterior (L5), or peroneus longus and brevis muscles (S1). These patients usually have severe tenderness and spasm over the lumbar paraspinal muscles. Lumbosacral MRIs may reveal disk herniation at L4-5 and/or L5-S1 level(s). Electromyography (EMG)/NCV tests may not detect a lumbar radiculopathy. NSAIDs, muscle relaxants, and physical therapy may help some patients with acute disk herniation and lumbar radiculopathy. If patients fail these treatments, lumbar epidural corticosteroid injections may offer fast and effective pain relief if the nerve roots are not severely mechanically compressed. Surgery is suggested for those with severe focal weakness of relevant muscles or incontinence. Surgery may also be indicated for severe pain that lasts for more than 3 months and does not respond to aggressive pain management if disk herniation is demonstrated by MRI or computed tomography (CT) studies.

Lumbar facet joint syndrome is found in up to 35% of patients with LBP. It is frequently associated with arthritis or injuries in lumbar facet joints. Patients may complain of pain in the low back, often on one side only. Pain may radiate down the back or front of the thigh. Physical examination may find positive tenderness over the lumbar paraspinal muscles and facet joints. Back extension and lateral rotation to the side of the pain often increases the back pain. Results of a straight leg raising test should be negative. Neurological examination should be normal unless there is a coexistent lumbar radiculopathy or other neurological condition. Diagnosis of facet joint syndrome is clinical. MRI and CT reports of facet joint arthropathy do not correlate with clinical findings. Often these changes are age related. NSAIDs should be tried for patients with lumbar facet joint syndrome before they are considered for diagnostic medial branch blocks or intra-joint corticosteroid injections.

Sacroiliac (SI) joint syndrome is another major source of LBP. The patient may have pain in one side of the low back, with pain radiation down to the hip or thigh. Pain is often increased when these patients try to walk upstairs. Physical examination may find tenderness over the SI joint, and the Patrick test or single-leg standing often exacerbate SI joint pain. NSAIDs are the first-line medication for SI joint inflammation. SI joint corticosteroid injection can provide temporary pain relief. Radiofrequency lesions to denervate the SI joint have been reported effective; however, more studies are needed to confirm clinical efficacy of this treatment.

Lumbar spinal stenosis is a common age-related change. The majority of seniors older than 60 years of age have varying degrees of spinal stenosis due to disk herniation, osteophytes, or degenerative spondylolisthesis. Preexisting congenital lumbar canal stenosis predisposes to the development of this syndrome. Fortunately, fewer than 30% of those with spinal stenosis have clinical pain. Patients often have pain in the low back, with pain radiation down the back of both legs. Standing or walking may worsen pain. Patients often walk with a hunched back and sit down after walking a short distance to relieve pain (neurogenic claudication). The pain usually takes minutes to disappear, compared to seconds with vascular claudication. On physical examination, patients often have less tenderness over the lumbar spine than those with acute lumbar disk herniation. A straight leg raising test may be normal. The condition must be distinguished from vascular claudication. Patients may try NSAIDs first. Lumbar epidural corticosteroid injections may provide pain relief for this group of patient for weeks or even months. If a patient has severe pain and refuses surgery, chronic narcotic treatment often provides adequate pain control but runs a risk of the development of tolerance and addiction.

Cervicogenic Headache

Cervicogenic headache refers to head pain originating from pathology in the neck. It is believed that pain from the C2-C3 nerve dermatome can radiate to the head and face (Fig. 44.1). An earlier study found that pain from the C2-C3 and C3-C4 cervical facet joints also can radiate to the occipital area (Fig. 44.2). The term cervicogenic headache was first introduced by Sjaastad and colleagues in 1983. However, the concept of cervicogenic headache is controversial and not well accepted by the majority of neurologists. The International Headache Society (2004) published its first diagnostic criteria in 1998 and revised it 2004. Patients with cervicogenic headache often have histories of head and neck trauma. Pain may be unilateral or bilateral. Pain is frequently localized to the occipital area, but it may also be referred to the frontal, temporal, or orbital regions. Headaches may be triggered by neck movement or sustained neck postures. This headache is constant with episodic throbbing attacks like a migraine. Patients may have other symptoms mimicking a migraine, such as nausea, vomiting, photophobia, phonophobia, and blurred vision. Owing to significant overlap of the symptoms of cervicogenic headache and migraine without aura, cervicogenic headache is often misdiagnosed as migraine. Clinicians should always consider cervicogenic headache in the differential diagnoses when evaluating a headache patient. History of head/neck injury and detailed examination of the occipital and upper cervical area should be part of the evaluation for headache. Patients with cervicogenic headache may have tenderness over the greater or lesser occipital nerve, cervical facet joints, and muscles in the upper or middle cervical region. Cervicogenic headache does not respond well to migraine medications. Treatment should be focused on removal of the pain source from the occipital-cervical junction. Initial therapy is directed to physical therapy modalities and NSAIDs. Interventional treatment such as greater occipital nerve block, cervical facet joint block, deep cervical plexus block, and botulinum toxin injections may provide effective pain relief (Zhou et al., 2010).

Fig. 44.1 Pain referral from C2 and C3 nerve roots. The C2 pain dermatome consists of an occipital parietal area 6 to 8 cm wide, extending paramedially from the subocciput to the vertex. The C3 pain dermatome is a craniofacial area, including the scalp around the ear, the pinna, the lateral cheek over the angle of the jaw, the submental region, and the lateral and anterior aspects of the upper neck.

(From Poletti, C.E., 1996. Third cervical nerve root and ganglion compression: clinical syndrome, surgical anatomy, and pathological findings. Neurosurgery 39, 941-948.)

Fig. 44.2 Pain referral pattern from the cervical facet joints. Injection of contrast medium into the cervical facet joints in normal volunteers induced pain in the head, neck, and shoulder area. Of note, injection of C2-C3 facet joints can cause pain in the occipital area.

(From Dwyer, A., Aprill, C., Bogduk, N., 1990. Cervical zygapophyseal joint pain patterns. I: a study in normal volunteers. Spine 15, 453-457.)

Complex Regional Pain Syndrome

Terminology describing the complex regional pain syndromes has evolved over the last century. The term causalgia was first coined by Weir Mitchell in the 1870s for severe progressive distal limb pain with major nerve injury. In 1946, Evans introduced the term reflex sympathetic dystrophy (RSD); it was later defined by the International Association for Study of Pain (IASP) as “continuous pain in a portion of an extremity after trauma, which may include fracture but does not involve a major nerve, associated with sympathetic hyperactivity.” In 1994, the IASP introduced the term complex regional pain syndrome (CRPS), describing a painful condition that includes regional pain, sensory changes (e.g., allodynia), abnormalities of temperature, abnormal sudomotor activity, edema, and abnormal skin color changes that occur after an initiating noxious event such as trauma. Two types of CRPS have been recognized: CRPS I corresponds to RSD, in which no definable nerve lesion is found. CRPS II refers to the cases with a definable nerve lesion and corresponds to the earlier term of causalgia.

The mean age of CRPS patients ranges from 36 to 46 years, with women predominating (60%-81%). It is caused typically by an injury such as a fracture (16%-46%), strain or sprain (10%-29%), post surgery (3%-24%), and contusion or crush injury (8%-18%). Clinical features of CRPS often include pain, edema, autonomic dysfunction such as change in temperature or color in the involved limbs, motor dysfunction, and psychological abnormalities such as depression (Fig. 44.3). Schwartzman and Maleki reported the pattern of spreading of CRPS in three stages. In the early stage, CRPS often involves only one limb with pain, minor edema, and increased skin temperature. CRPS may spread from one limb to the others. In the later stage, CRPS could involve the full body and the four extremities with severe pain, edema, cold and cyanotic limbs, joint contracture, and atrophy of muscles and bones.

Fig. 44.3 Complex regional pain syndrome I. The right foot and ankle are mildly swollen and reddish in comparison with the left foot and ankle. The intravenous access was used for a Bier block.

Excruciating pain is the cardinal feature of CRPS. Pain is often described as burning, aching, pricking, or shooting. Severity of pain is not proportional to the initial injury, and pain is not limited to the area of the injury or a specific nerve distribution. Patients may feel severe pain to minor pain stimulation such as a safety-pin prick (hyperalgesia). A light touch to skin (innocuous stimulation) may cause severe long-lasting pain (mechanical allodynia). A cooling stimulus such as a drop of alcohol may be perceived as painful (thermal allodynia). Decreased temperature and pinprick sensations in the affected limb are common.

Edema of the affected limb is present in the majority of patients. It could be very mild in the early stage of CRPS, mimicking mild cellulitis. However, in the late stage, edema may be so severe that a Doppler test is needed to rule out the possibility of deep vein thrombosis.

Autonomic dysfunction may manifest as changes of skin color and temperature, as well as sweating abnormalities. The affected area may be reddish at one time and then become blue, purple, or pale over a course of minutes to hours. Livedo reticularis is common in CRPS. Livedo is a descriptive term used to describe the red, non-blanchable (i.e., does not turn white when pressed) network pattern (reticulated) in the skin. About 60% of patients may report excessive sweating in the affected limbs. Temperature asymmetry between the affected and unaffected sides may exceed 1°C.

Motor dysfunctions in CRPS include mild weakness, decreased range of motion, tremor, dystonia, and myoclonus. Dystrophic manifestations are seen in the form of increased or decreased nail and hair growth in the affected extremity, hyperkeratosis or thin glossy skin, and osteoporosis of the underlying bones.

Diagnosis of CRPS is clinical. According to IASP, if a patient has the above-mentioned features, a diagnosis of CRPS may be made if other clinical conditions such as infection or DVT are ruled out. EMG/NCV tests are not sensitive to CRPS and frequently cause severe pain to patients. A triple-phase bone scan may reveal abnormal absorption in the affected limbs (increased or decreased), though it is not a primary diagnostic procedure for CRPS.

The pathophysiology of CRPS is not completely understood. Multiple mechanisms are considered in the generation and maintenance of CRPS. Increased systemic calcitonin gene-related peptide (CGRP) levels may contribute to neurogenic inflammation, edema, vasodilatation, and increased sweating. Elevated neuropeptide concentrations may lead to pain and hyperalgesia. Immunological mechanisms (e.g., altered expressions of human leukocyte antigen [HLA], substance P, cytokines, and interleukins) are believed to contribute to the pathogenesis of clinical symptoms such as edema. Up-regulation of adrenergic receptors and functional coupling between sympathetic efferent and sensory afferent fibers may provide the basis of the sympathetic nervous system abnormalities in the pathogenesis of CRPS. The central mechanisms in CRPS may include central sensitization in the spinal cord, brainstem, or thalamus, cortical reorganization in the primary somatosensory cortex, and disinhibition of the motor cortex.

The goals of treatment for CRPS are pain relief, functional recovery, and psychological improvement. However, treatment of CRPS remains a challenge. There is little if any evidence for the efficacy of any treatment modality. In the early stages of CRPS treatment, occupational and physical therapies are often used. Occupational and physical therapies are supported by anecdotal data and have not been validated by randomized prospective trials.

Patients diagnosed with CRPS for over 2 months should also undergo a psychological evaluation—which includes psychometric testing—to identify and treat psychological disorders such as anxiety, depression, or personality disorders. Counseling, behavioral modification, biofeedback, relaxation therapy, group therapy, and self-hypnosis should be considered. The goal of psychotherapy is to improve patient motivation and coping skills.

Tricyclic antidepressants, antiepileptics, and narcotics such as methadone are commonly used empirically for CRPS, even though clinical controlled studies have not proven their efficacy. A recent review article summarized the evidence derived from randomized controlled trials pertaining to the treatment of CRPS. The review reported clinical improvement with dimethyl sulfoxide, steroids, epidural clonidine, and intrathecal baclofen. Only bisphosphonates appear to offer clear benefits for patients with CRPS (Tran de et al., 2010). NMDA receptor modulation is a major interest of current research. It has been reported that subanesthetic infusions of ketamine might offer a promising therapeutic option in the treatment of appropriately selected patients with intractable CRPS (Schwartzman et al., 2009). A recent preliminary study reported that IV immunoglobulin treatment could potentially decrease pain in CRPS patients (Goebel et al., 2010). However, more studies are needed to further establish the safety and efficacy of these novel approaches.

Minimally invasive techniques have been used extensively for the treatment of CRPS. Techniques include sympathetic block, intravenous regional block (IVRB), somatic nerve block, epidural drug administration, intrathecal drug delivery, and neurostimulation. Stellate ganglion blocks in early-stage CRPS may significantly decrease pain and hasten clinical recovery. It may also prevent the recurrence of CRPS after reoperation of the affected extremity. In a double-blind study, IVRB with bretylium provided significantly longer analgesia than lidocaine. Good pain relief is reported with the use of epidural delivery of clonidine and ketamine and also with intrathecal baclofen and morphine. An early study with 2-year follow-up reported that spinal cord stimulation (SCS) results in a long-term pain reduction and improvement in health-related quality of life. However, a more recent randomized study with 5-year follow-up found no extra benefit in terms of pain relief for those with a combination of SCS and physical therapy, compared to those with physical therapy alone (Kemler et al., 2008). The author shares the same experience and opinion with the cited report. It seems that most RSD patients feel better immediately after the SCS implantation. However, the SCS itself may have difficulty stopping the spread of RSD, and once RSD spreads out of the area initially covered by the SCS, the pain is no longer “under control.”

Post-Stroke Pain Syndrome

Lesions at any level of the neuroaxis (generally affecting spinothalamocortical afferent sensory pathways) including the medulla, pons, midbrain, thalamus, subcortical white matter, and the cortex may produce central post-stroke pain syndrome (PSP). However, the thalamus and brainstem are common sites for PSP; 8% to 16% of thalamic stroke may lead to chronic pain. The frequency of pain after a geniculothalamic artery stroke is even higher (13%-59%).

The pathogenesis of PSP is not yet known. However, it has been suggested that hyperexcitation in the damaged sensory pathways, damage to the central inhibitory pathways, or a combination of the two may be responsible for the onset of PSP. Pain is the cardinal symptom and is described as spontaneous, severe, paroxysmal, and burning. Patients with thalamic pain syndrome also have hyperalgesia and allodynia in the affected limbs. Right-sided lesions predominate among reported cases of the thalamic pain syndrome.

Patients reporting pain due to brainstem infarction usually have involvement of pontine or medullary structures. Patients with midbrain infarction seldom complain of pain. Transitory eye and nose pain may be an initial symptom of pontine infarction. About 25% patients with dorsolateral medullary infarction develop ipsilateral facial pain, especially when the lesion involves the spinal trigeminal tract. Facial allodynia is also common. Some patients may experience pain in the contralateral limbs and trunk.

Treatment of central post-stroke pain remains a challenge. Tricyclic antidepressants are still a choice of treatment. Gabapentin and lamotrigine have been used to treat central post-stroke pain syndrome in open-labeled studies. Selective posterior rhizotomy has been reported to decrease painful spasticity in the lower limbs of hemiplegic patients after a stroke. It has been reported that chronic motor cortex stimulation therapy provides pain relief for some post-stroke patients (Brown and Pilitsis, 2006). Stereotactic radiosurgery of the pituitary and deep brain stimulation (DBS) have been used to treat PSP syndrome with some success (Pickering et al., 2009).

Some 40% to 60% of patients develop shoulder pain after a stroke. The mechanism of shoulder pain is not clear, but a strong association exists between pain and an abnormal shoulder joint examination, ipsilateral sensory abnormalities, and arm weakness. These patients usually have significant tenderness over the shoulder joint. It is postulated that the pain is due to inflammation in the joint secondary to immobilization and joint contracture ( frozen shoulder syndrome). The majority of shoulder pain may be resolved or improved for 6 months following a stroke with intensive physical/occupational therapy. Antiinflammatory medications may be used. Suprascapular nerve or brachial plexus block can provide temporary pain relief to prepare for physical therapy. Proper positioning of the shoulder, range-of-motion activities, and avoidance of immobilization may further help prevent or alleviate shoulder pain.

Spinal Cord Injury and Pain

There are about 240,000 patients with spinal cord injuries (SCIs) in the United States; 86% of individuals with SCI report pain at 6 months post discharge, with 27% of these individuals reporting pain that impacts most of their daily activities. Patients can have pain both at and below the level of spinal injury. Pain intensity is not associated with the magnitude or location of the lesion, occurrence of myofascial pain syndrome, or onset of pain. However, pain is usually more severe in patients with gunshot injuries.

Pain after SCI originates from different sources including neuropathic, musculoskeletal, and visceral pain. Neuropathic pain after SCI is further divided into central and segmental pain. Central neuropathic pain often begins within weeks or months after injury. It is generally described as a burning, sharp, or shooting pain. Patients feel pain at or below the level of injury in areas where there is partial or complete loss of sensation to touch. Central pain is believed to be due to differentiation caused by SCI. Astrocytic activation in the spinal cord, up-regulation of chemokines, hyperexcitability of wide–dynamic range neurons in the spinal dorsal horn rostral to the lesion, and loss of γ-aminobutyric acid (GABA)ergic interneurons in laminae I-III of the spinal cord dorsal horn (Meisner et al., 2010) have been suggested to cause the neuropathic pain that follows SCI. Segmental pain often occurs around the border of injury and usually develops within the first few months after an injury. Allodynia and hyperalgesia are common. Nerve root entrapment could lead to severe segmental pain. Patients may describe stabbing or sharp pain or a band of burning pain at the level of injury. Syringomyelia with a cyst ascending from the level of the SCI may occasionally cause central pain.

Musculoskeletal pain in this group of patients may be due to muscle spasms below the level of SCI and arthritis in disused joints. Pain is generally described as dull or aching. It is usually worsened by movement and eased with rest. Visceral pain may begin a short time following SCI and could be related to constipation and urinary retention due to sphincter dysfunction. It may occur in the abdomen above or below the level of injury. This pain is often described as cramping, burning, and constant.

Pain management after SCI is difficult. Pharmacological and rehabilitative procedures are effective in only about 38% of patients. However, the initial workup should target identifying the pain source. Different kinds of pain may respond differently to treatments. For neuropathic pain, medications such as gabapentin, amitriptyline, and nortriptyline may ease the pain in some patients. Intravenous lidocaine may provide temporary pain relief. Intrathecal baclofen therapy may reduce chronic musculoskeletal pain associated with spasticity and improve the patient’s quality of life. Intrathecal morphine and clonidine offer limited help to relieve the pain. DBS has been reported to be effective in some cases, but there is insufficient evidence to validate its routine use. Limited evidence exists for use of motor cortex stimulation (Previnaire et al., 2009). SCS lacks long-term efficacy for the relief of spasticity and pain in SCI and is believed not to be cost-effective. Dorsal root entry zone lesions and dorsal rhizotomy have also been used with limited success. Appropriate management of bowel or bladder dysfunction may help ease visceral pain. If an ascending syrinx is present, surgical drainage may be effective in relieving the pain.

Pain in Multiple Sclerosis

Pain is a common symptom in multiple sclerosis (MS). The prevalence of pain in this disease is higher than what was initially expected; some studies estimate it to be up to 86% (Bermejo et al., 2010), depending on the sample and specific questions used to assess the incidence and severity of pain. Osterberg et al. studied pain syndromes in 429 patients with definite MS, and 58% reported pain during the course of their disease; 100 (28%) had central pain, including 18 patients (5%) with trigeminal neuralgia. The majority of patients (87%) with central pain had symptoms located in the legs, while 31% were in the arms. Pain was mostly bilateral (76%) and constant. Aching, burning, and pricking were common qualities. Other reported pain syndromes in MS include the Lhermitte sign, dysesthetic pain, back pain, headache, and painful tonic spasms. Chronic pain in MS was found to have no significant relationship to gender, age of onset, disease duration, or disease course. Chronic pain can have a significant negative impact on functions in persons with MS, such as the ability to engage in household work and psychological functioning. Chronic pain is significantly related to anxiety and depression in females. In the long-term care facility, residents with MS are more physically disabled and experience more frequent pain and a higher prevalence of pressure ulcers and depression than residents without MS.

Though pain affects a high percentage of patients with MS, its pathophysiology is unknown, and few studies have been conducted to investigate the treatment of pain in MS. The following principles are currently recommended for treatment of MS related pain:

Even though cannabis is not legally used in the United States to treat pain, European studies indicate that cannabis-based medicines are effective in reducing pain and sleep disturbance in patients with MS-related central neuropathic pain and are mostly well tolerated (Rog et al., 2005; Thaera et al., 2009). Oral ketamine, an NMDA receptor antagonist, has also been reported to be effective in the treatment of pain and allodynia associated with MS.

Phantom Limb Pain and Stump Pain

Phantom-limb pain describes the pain in a body part that is no longer present, which occurs in 50% to 80% of all amputees. Pain can have several different qualities, such as stabbing, throbbing, burning, or cramping. It seems to be more intense in the distal portions of the phantom limb. This pain may be related to a certain position or movement of the phantom and may be elicited or exacerbated by a range of physical factors (e.g., changes in weather or pressure on the residual limb) and psychological factors (e.g., emotional stress). It is more likely to occur if the individual had chronic pain before the amputation. Pain in the phantom is often similar to the pain felt in the limb before amputation. Phantom pain is most common after the amputation of an arm or leg, but it may also occur after the surgical removal of other body parts such as breast, rectum, penis, testicle, eye, tongue, or teeth. About 30% of persons with amputation report the feeling of telescoping, the retraction of the phantom towards the residual limb, and in many cases the disappearance of the phantom into the limb. This may be accompanied by a shrinking of the limb. Recent evidence suggests that telescoping is associated with more phantom-limb pain.

Phantom-limb pain is commonly confused with pain in the area adjacent to the amputated body part. This pain is referred to as residual-limb or stump pain. Patients may report severe “knife-stabbing” or sharp pain at the end of the amputated limb. Formation of a neuroma or pressure lesions of the stump may exacerbate stump pain. Physical examination may reveal the existence of a neuroma; it is usually very sensitive to touch or pressure. However, stump pain may coexist with phantom-limb pain.

Changes along the neuroaxis may contribute to the experience of phantom-limb pain. Spinal mechanisms are characterized by increased excitability of the dorsal horn neurons, reduction of inhibitory processes, and structural changes at the central nerve endings of the primary sensory neurons, interneurons, and the projection neurons. Supraspinal changes related to phantom-limb pain involve the brainstem, thalamus, and cortex. Reorganization of the somatosensory cortex of the human cerebral cortex in amputees has been supported by findings from several imaging studies. People with arm or hand amputations show a shift of the mouth into the hand representation in the primary somatosensory cortex (Woodhouse, 2005). Studies in human amputees have shown that reorganizational changes also occur at the thalamic level and are closely related to the perception of phantom limbs and phantom-limb pain. Neuroma in the stump may be more responsible for stump pain than phantom-limb pain. However, abnormal input originated from a neuroma in the residual limb may increase the amount of central reorganization, enhancing the chance of phantom-limb pain. Psychological factors play a role in the modulation of phantom-limb pain; the pain may be exacerbated by stress. Patients who lack coping strategies, fear the worst, or receive less social support tend to report more phantom-limb pain.

Treatment for phantom-limb pain is difficult. Although tricyclic antidepressants and sodium channel blockers are treatments of choice for neuropathic pain, no controlled studies exist of these agents for phantom-limb pain. Opioids, calcitonin, and ketamine have proven to be effective in reducing phantom-limb pain in controlled studies. Transcutaneous nerve stimulation (TENS) may have a minor effect. A maximum benefit of about 30% has been reported from treatments such as local anesthesia, far-infrared rays, sympathectomy, dorsal root entry-zone lesions, cordotomy, rhizotomy, neurostimulation methods, or pharmacological interventions such as anticonvulsants, barbiturates, antidepressants, neuroleptics, and muscle relaxants. Use of a myoelectric prosthesis may alleviate cortical reorganization and phantom-limb pain, and DBS has also been reported to treat phantom-limb pain. Mirror therapy has been studied, but to date, there is only circumstantial evidence for the effectiveness of mirror therapy in treating phantom pain; more studies are needed to support its clinical use.

Nonsteroidal Antiinflammatory Drugs

Nonsteroidal antiinflammatory drugs, including aspirin, are the most widely used analgesics. Traditionally NSAIDs are considered weak analgesics and used extensively for headaches, arthritis, and a wide range of minor aches and post-surgical pain conditions.

NSAIDs are powerful inhibitors of prostaglandin synthesis through their effect on cyclooxygenase (COX). Prostaglandins are not thought to be important pain mediators, but they do cause hyperalgesia by sensitizing peripheral nociceptors to the effects of various mediators of pain and inflammation such as somatostatin, bradykinin, and histamine. Thus, NSAIDs are used primarily to treat pain that results from inflammation and hyperalgesia. Table 44.1 lists commonly used NSAIDs.

Table 44.1 Commonly Used Oral Nonsteroidal Antiinflammatory Drugs

Acetaminophen is not strictly an antiinflammatory medication. Its peripheral and antiinflammatory effects are weak, but it shares many properties of NSAIDs. It readily crosses the blood-brain barrier, and its action resides primarily in the central nervous system (CNS), where prostaglandin inhibition produces analgesia and antipyresis.

Common side effects of NSAIDs include GI toxicity, stomach ulcers, and gastric bleeding. Renal dysfunction can occur with prolonged and excessive use of NSAIDs. Particularly at risk from excessive use of NSAIDs are elderly patients with renal dysfunction, congestive heart failure, ascites, or hypovolemia. Other adverse effects of NSAIDs include hepatic dysfunction or necrosis, asthma, vasomotor rhinitis, angioneurotic edema, urticaria, laryngeal edema, or even cardiovascular collapse. Because of the wide availability of acetaminophen and its potential toxicity (especially liver toxicity), in 2009 the U.S. Food and Drug Administration (FDA) proposed a decrease in the maximum daily dose of acetaminophen from 4000 mg to 3250 mg, reducing the maximum individual dose from 1000 to 650 mg. They relegated 500 mg tablets to prescription status and mandated new labeling on acetaminophen packaging (Krenzelok, 2009). Acetaminophen is a potential cyclooxygenase 2 (COX-2)–selective inhibitor. It may also increase cardiovascular risks.

Cardiovascular risks of NSAIDs, especially COX-2 inhibitors, have become a major focus of attention in recent years. Suggestions that the use of COX-2 inhibitors may decrease prostacyclin (PGI2) levels, with relatively unopposed platelet thromboxane A2 generation that may lead to increased thrombotic risk, have cautioned against the use of such agents. Rofecoxib (Vioxx) was withdrawn from the market in September 2004 owing to increased cardiovascular risks. A recent study found that the hazard ratio (95% confidence interval) for death was 1.70, 1.75, 1.31, 2.08, 1.22, and 1.28 for rofecoxib, celecoxib, ibuprofen, diclofenac, naproxen, and other NSAIDs, respectively (Gislason et al., 2009). Even though limited long-term data on cardiovascular risk associated with nonselective NSAIDs have been available, and some contradictory warnings and recommendations have been published recently by the American Heart Association, FDA, and independent experts (Gluszko and Bielinska, 2009), the general suggestion is that both NSAIDs and selective COX-2 inhibitors should be avoided or used with extreme caution if a patient has a high cardiovascular risks and a history of heart failure.

Antidepressants

Tricyclic antidepressants are probably the most commonly used adjunct analgesics in the management of chronic pain (Dworkin et al., 2010) (Table 44.2). The tertiary amines (amitriptyline, imipramine, doxepin, and clomipramine) and the secondary amines (nortriptyline and desipramine) both have analgesic properties. Amitriptyline is the prototype antidepressant used in this context. Clinical efficacy of tricyclics for neuropathic pain has been demonstrated by numerous well-controlled double-blind clinical studies for both neuropathic and somatic pain. Clinicians have to be familiar with the possible side effects of amitriptyline, especially in elderly patients. These adverse effects include sedation, dry mouth, constipation, urinary retention, glaucoma, orthostatic hypotension, and cardiac arrhythmias. Patients should be warned about the side effects before they start the medication. Amitriptyline should be avoided in patients with a history of heart disease (conduction disorders, arrhythmias, or heart failure) and closed-angle glaucoma. Amitriptyline should be started at a relatively low dose (10 mg) at bedtime and slowly titrated up as tolerated. Most patients report improved sleep after taking amitriptyline. The onset of pain relief may precede the anticipated onset of antidepressant effects. In general, pain relief may be expected in 7 to 14 days. The dosage required for pain management is usually lower than for depression; 75 to 100 mg at bedtime is often effective. If the patient cannot tolerate this dose or is not a good candidate for amitriptyline, other tricyclics such as nortriptyline or desipramine may be considered. These secondary amines generally have fewer anticholinergic effects and are therefore better tolerated than tertiary amines. However, their clinical efficacy is not as well established as that for amitriptyline.

Table 44.2 Tricyclic Antidepressants Commonly Used for Pain Management

The main advantage of the selective serotonin reuptake inhibitors (SSRIs) is the favorable side-effect profile. However, SSRIs are clearly less effective than tricyclic antidepressants. The NNT (number needed to treat to reach 50% pain relief) is 6.7 versus 2.4 (Coluzzi and Mattia, 2005). It seems that selective serotonin/noradrenaline reuptake inhibitors (SNRI) are relatively more effective for pain management than most of the SSRIs. Venlafaxine is an SNRI for which randomized controlled trials showed good pain relief effect for painful polyneuropathy and neuropathic pain following treatment of breast cancer. Duloxetine has also been demonstrated to have significant analgesic effects in diabetic polyneuropathy and fibromyalgia. Milnacipran is another SNRI; randomized double-blind placebo-controlled studies found that milnacipran is effective in controlling pain and improving global status, fatigue, and physical and mental function in patients with fibromyalgia (Arnold et al., 2010). Nausea, hyperhidrosis, and headache are the most common adverse events.

Anticonvulsants

Anticonvulsants are believed to be particularly useful in treating lancinating, electrical, or tic-like pain. These medications may be also beneficial in patients with neuropathic pain who do not respond to antidepressants. The older generation of anticonvulsants includes carbamazepine, valproic acid, clonazepam, and phenytoin. Carbamazepine was perhaps the most popular agent used for trigeminal neuralgia. However, carbamazepine may cause serious side effects such as sedation, nausea, vomiting, bone marrow suppression, hyponatremia, hepatic dysfunction, and serious drug-drug interaction. Carbamazepine should be started at 100 mg at night and titrated up slowly, especially for the elderly.

Valproic acid has been proven to be effective in reducing the frequency of migraine attacks (Vikelis and Rapoport, 2010). Some studies found that valproates may provide significant pain relief in patients with post herpetic neuralgia and diabetic neuropathy. However, negative results have also been reported. Common side effects include tremor, ankle swelling, sedation, and GI discomfort. Weight gain and hair loss may be a major cosmetic concern, especially for younger patients. Valproate should not be used for children younger than 2 years of age because of hepatotoxicity. Generally, valproate is not the first-line choice for neuropathic pain.

Gabapentin modulates the function of the α₂-δ subunit of voltage-dependent calcium channels in the dorsal horn of the spinal cord to decrease the release of excitatory neurotransmitters such as glutamate and substance P. The analgesic efficacy of gabapentin has been demonstrated in several types of nonmalignant neuropathic pain. Its high safety profile, few drug-drug interactions, and proven analgesic effect in several types of neuropathic pain have made gabapentin the recommended first-line co-analgesic for treating a variety of neuropathic pains, especially in the medically ill and in elderly patients. The most common adverse effects are drowsiness, dizziness, and unsteadiness. Gabapentin should be started at a dose of 100 to 300 mg at bedtime. If titrated carefully, gabapentin is usually well tolerated up to 3600 mg daily. However, gabapentin has a nonlinear pharmacokinetic profile: the rate of bioavailability decreases as the dose increases.

Pregabalin is a GABA analog with similar structure and mechanism of action as gabapentin. It has antiepileptic, analgesic, and anxiolytic activity. Pregabalin has been approved by the FDA for the management of neuropathic pain associated with diabetic neuropathy, postherpetic neuralgia, and fibromyalgia (Straube et al., 2010). Food does not significantly affect the extent of absorption. Pregabalin is not protein bound and exhibits a plasma half-life of about 6 hours. Hepatic metabolism is negligible, and most of the oral dose (95%) appears unchanged in the urine. At a dose of 300 mg/day, about 45% of diabetic neuropathy patients had 50% pain relief. This means that pregabalin has an NNT of 2.2 for diabetic neuropathy. Pregabalin seems to be more effective than gabapentin and other anticonvulsants for neuropathic pain. Common side effects of pregabalin include dizziness, sedation, dry mouth, and peripheral edema.

Oxcarbazepine is a ketoderivative of carbamazepine, with better tolerability. It can block sodium-dependent action potentials. The medication does not induce hepatic enzymes and has fewer drug-drug interactions than carbamazepine. Multiple open studies have suggested that oxcarbazepine may be effective for the treatment of neuropathic pain. However, a double-blind controlled study did not find significant difference between oxcarbazepine and placebo for the treatment of pain due to diabetic neuropathy (Grosskopf et al., 2006).

Lamotrigine is an antiepileptic drug that stabilizes neural membranes by blocking the activation of voltage-sensitive sodium channels and inhibiting the presynaptic release of glutamate. Multiple open studies have supported the use of lamotrigine in neuropathic pain. However, controlled studies found no efficacy of lamotrigine for the treatment of neuropathic pain (Breuer et al., 2007; Rao et al., 2008). Lamotrigine is ineffective for prevention of migraine.

Topiramate has proven its efficacy and safety in the prophylactic treatment of episodic migraine in a number of randomized controlled clinical trials (Naegel and Obermann, 2010). Even though open studies and case reports continue to support the use of topiramate in the treatment of various kinds of neuropathic pain, controlled studies failed to reveal any benefit of topiramate for the treatment of neuropathic pain. The mechanisms of action include blockade of sodium channels, enhancement of GABA inhibition, and attenuation of kainate-induced responses at glutamate receptors. The starting dose is usually small (e.g., 25 mg twice a day for an adult). It may be incrementally increased weekly by 50 mg up to 200 mg/day. Topiramate may induce memory loss, word-finding difficulties, disorientation, and sedation. The other common adverse affects are renal calculi, tremors, dizziness, ataxia, headaches, fatigue, and GI upset. Topiramate may also induce significant weight loss. This medication may be more helpful in obese pain patients.

Tiagabine, zonisamide, and levetiracetam are among the group of new anticonvulsants. Some uncontrolled and case studies have reported positive effects of these medications for neuropathic pain. However, controlled double-blind studies have not been reported.

Systemic Local Anesthetics

Systemic administration of local anesthetics has been used to treat neuropathic pain syndrome. Clinical trials have provided some evidence that lidocaine and mexiletine are superior to placebo for neuropathic pain (Carroll et al., 2008). Intravenous lidocaine is used for the treatment of neuropathic pain as a second-line therapy. If a patient has a positive response to IV lidocaine therapy, a trial of oral mexiletine may be considered. However, mexiletine has a relatively high rate of adverse effects such as nausea, vomiting, tremor, dizziness, unsteadiness, and paresthesias. Given the limited number of supportive studies, mexiletine and other oral local anesthetics should only be used as second-line agents for neuropathic pain that has failed to respond to anticonvulsants or antidepressants.

Topical Analgesics

Double-blind placebo-controlled studies have confirmed the efficacy of the 5% lidocaine patch for the treatment of postherpetic neuralgia (Lin et al., 2008) and for those patients with trigger points in myofascial pain syndrome (Affaitati et el., 2009). However, the lidocaine patch may not be effective in treating pain due to traumatic rib fractures (Ingalls et al., 2010). Minimal systemic absorption occurs. The patch is usually applied 12 hours per day, with minimal systemic side effects. Topical lidocaine ointment in various concentrations (up to a compounded formulation of 10%) may offer a cost-effective alternative.

Capsaicin is the spicy ingredient in chili pepper. It can deplete substance P from the terminals of afferent C fibers, potentially leading to decreased pain perception. Capsaicin creams are effective in reducing postsurgical pain in cancer patients. When applied topically, it may initially release substance P and cause severe burning pain. Pain related to the use of capsaicin gradually decreases over a few days if the cream is applied regularly. A lower-concentration cream (0.025%) or the application of a topical local anesthetic may help some patients decrease the initial burning pain and tolerate the medication better. A recent study found that topical capsaicin might effectively decrease pain in patients with chronic migraine (Papoiu and Yosipovitch, 2010). It is important to warn patients not to get any trace of the cream on mucous membranes, since this causes severe pain.

Miscellaneous Drugs

Baclofen is a GABA_B receptor agonist with powerful antinociceptive effects in experimental animal models and established efficacy for chronic pain. It may be most useful in blocking the lancinating or episodic types of pain and reducing allodynia. It is commonly used for trigeminal neuralgia, together with carbamazepine. Baclofen may be started in doses of 5 mg, 2 to 3 times a day, and may be escalated to a maximum dose of 200 mg given in divided doses. Common side effects include CNS symptoms such as dizziness and drowsiness, as well as GI symptoms. Baclofen is a highly hydrophilic agent and has poor penetration of the blood-brain barrier. Intrathecal baclofen could be a promising adjuvant therapy to enhance the effect of other intrathecal medications such as morphine or clonidine or SCS for chronic pain.

N-Methyl-D-Aspartate Receptor Blockers

NMDA receptors are involved in the development of central sensitization associated with chronic refractory pain syndromes. NMDA antagonists may modulate CNS function, offering a novel approach to treating chronic neuropathic pain. Intravenous anesthetic doses of ketamine may induce serious side effects such as vivid hallucinations and psychosis. However, double-blind placebo-controlled studies have confirmed that low-dose IV ketamine may provide significant pain relief for CRPS type 1 without significant psychomimetic side effects (Sigtermans et al., 2009). Methadone has the property of both µ-opioid receptor agonist and NMDA antagonist. Evidence indicates that methadone has similar analgesic efficacy to morphine, but adverse effects due to prolonged half-life—particularly respiratory depression, cardiac arrhythmia, and sudden death—make it critical for providers to be familiar with methadone’s pharmacological properties before considering methadone as an analgesic therapy for chronic pain. Amantadine is a noncompetitive NMDA antagonist. Dextromethorphan, the d-isomer of the codeine analog, levorphanol, is a weak, noncompetitive NMDA receptor antagonist. Memantine is an NMDA antagonist used for the treatment of Alzheimer disease. All three of these medications possess some analgesic properties. Current data are too scant or too weak, however, to recommend clinical use of any of these drugs for chronic pain management.

Opioid Analgesics

Opioids are the major class of analgesics used in the management of moderate to severe pain. These medications produce analgesia by binding to specific receptors both within and outside the CNS. However, their use in nonmalignant pain is still controversial. Opioid analgesics should be used with caution for chronic nonmalignant pain.

Opioids are classified according to the activity on the opioid receptors as full agonists, partial agonists, or mixed agonists-antagonists. Commonly used full agonists include hydrocodone, codeine, morphine, oxycodone, hydromorphone, methadone, and fentanyl. Buprenorphine is a partial agonist. It has lower intrinsic efficacy compared to other full opioid agonists and displays a ceiling effect to analgesia. Mixed agonist-antagonists include pentazocine, butorphanol tartrate, dezocine, and nalbuphine hydrochloride. These mediations block opioid analgesia at one type of receptor (µ) while simultaneously activating other opioid receptors (κ). Mixed agonist-antagonists should not be used together with full agonists, because they may cause withdrawal syndrome and increased pain. Table 44.3 lists commonly used narcotics and their equi-analgesic dosage.

Equi-analgesic dosage means the dose of different narcotics needed to reach the same analgesic effects. The middle two columns of Table 44.3, for example, indicate that 7.5 mg of oral hydromorphone every 3 hours may have analgesic effects equal to 1.5 mg of IV hydromorphone every 3 to 4 hours or 30 mg of oral morphine every 3 to 4 hours.

Narcotics are also classified as mild to strong according to their potency. Codeine is the prototype of the mild opioid analgesics. The duration of action (2-4 hours) is similar to that of aspirin and acetaminophen. It is commonly used together with NSAIDs when NSAIDs alone have proven ineffective. Hydrocodone, oxycodone, propoxyphene, and meperidine are other mild opioid analgesics. Meperidine is likely to cause dysphoria, or less commonly to cause myoclonus, encephalopathy, and seizures. These toxic effects result from metabolites such as normeperidine that accumulate with repeated doses. Meperidine should be avoided in patients who require chronic treatment. Morphine and hydromorphone are the prototypes of high-potency opioid analgesics. Morphine has a relatively rapid onset, especially when administered parenterally, and a short duration of action, about 2 to 4 hours. Sustained-release oral preparations (e.g., MS Contin and Kadian, with duration of action of 12 hours and 24 hours, respectively) are useful for patients requiring chronic opioid therapy.

Route of administration is important to consider when choosing opioids. Oral administration of opioids is the preferred route, because it is the most convenient and cost-effective. Oral opioids are available in tablet, capsule, and liquid forms and in immediate and controlled-release formulations. Patients should be informed not to break the controlled-release tablets, since this can cause immediate release and cause a potential overdose. If patients cannot take medication orally, other less-invasive routes such as transdermal or rectal routes should be tried. Intramuscular administration of narcotics should be avoided because this route is often painful and inconvenient, and absorption is unreliable. Intravenous administration may be more expensive and is not practical for most chronic pain patients.

The advantage of transdermal administration is that it bypasses GI absorption. Both fentanyl and buprenorphine are commercially available for transdermal administration. Fentanyl patches come in five sizes, delivering medication at 12, 25, 50, 75, and 100 µg/h. Each patch contains a 72-hour supply of fentanyl, passively absorbed through the skin during this period. Plasma levels rise slowly over 12 to 18 hours after the patch placement. This dosage form has an elimination half-life of 21 hours. Unlike IV fentanyl, transdermal administration of fentanyl is not suitable for rapid dose titration. It is often used for patients with chronic pain and already on opioids. As with other long-acting analgesics, all patients should be provided with oral or parenteral short-acting opioids for breakthrough pain.

Intrathecal analgesia may be considered when pain cannot be controlled by oral, transdermal, subcutaneous, or IV routes because side effects such as confusion and nausea further limit dose titration. Documentation of the failure of maximal doses of opioids and adjunct analgesics administered through other routes should precede consideration of intrathecal analgesia. For patients with chronic pain who have failed or cannot tolerate other treatment modalities, before implantation of a permanent pump, a trial is usually needed of single intrathecal injections, epidural injection, or continuous epidural administration. If there is significant pain relief without major side effects during the trial, the patient may be a candidate for permanent implantation of an intrathecal delivery system. Morphine is the most commonly used intrathecal drug used for pain relief. The main indication of the long-term intrathecal opioids is intractable pain in the lower part of the body. With proper selection and screening, good to excellent pain relief is expected in up to 90% of patients.

Physicians need to be familiar with side effects of opioids before prescribing these medications. Common side effects of opioids include constipation, sedation, nausea, vomiting, and respiratory depression due to overdoses. Occasionally, opioids may cause myoclonus, seizures, hallucinations, confusion, sexual dysfunction, sleep disturbances, and pruritus. Constipation is a common problem associated with opioid administration. Tolerance to the constipating effects of opioids hardly ever occurs during chronic therapy. Some patients are too embarrassed to tell the physician about constipation problems, so physicians should always ask patients about this. Mild constipation can usually managed by an increase in fiber consumption and the use of mild laxatives such as milk of magnesia. Severe constipation may be treated with a stimulating cathartic drug (e.g., bisacodyl, standardized senna concentrate, MiraLax, and similar drugs). Tapentadol is a novel centrally acting analgesic with two modes of action, µ-opioid agonist and norepinephrine reuptake inhibition. It was approved by the FDA for treatment of acute pain in the year 2008. Multiple double-blind controlled studies found tapentadol’s analgesic effects similar to morphine and oxycodone. However, tapentadol has fewer GI side effects such as nausea and vomiting (Daniels et al., 2009; Smit et al., 2010). Owing to its dual mechanism of action and better GI tolerability, there is potential for off-label use in chronic pain.

Transitory sedation is common if opioid doses are increased substantially, but tolerance also usually develops rapidly. Reducing the opioid dose, switching to another opioid, or use of CNS stimulants such caffeine, dextroamphetamine, or methylphenidate may help increase alertness. Nausea and vomiting may be managed with antiemetics chosen according to the modes of action (e.g., metoclopramide, chlorpromazine, haloperidol, scopolamine, hydroxyzine). Patients receiving long-term opioid therapy usually develop tolerance to the respiratory-depressant effects of these agents. However, respiratory depression is often due to an overdose, or when pain is abruptly relieved and the sedative effect of the opioid is no longer opposed by the stimulating effect of pain. To reverse respiratory depression, opioid antagonists (e.g., naloxone) should be given incrementally in doses that improve respiratory function but do not reverse analgesia, to avoid reoccurrence of severe pain.

Accumulation of normeperidine, a metabolite of meperidine, may cause seizures, especially in patients with chronic renal insufficiency. Therefore, meperidine is only indicated for acute use; chronic use should be avoided. Tramadol is a synthetic narcotic, most commonly used for mild pain. Tramadol may decrease the seizure threshold and induce seizures, so it should be avoided in patients with a history of seizures. It should not be not be used with tricyclic antidepressants. The recommended maximum dosage of tramadol is 400 mg/day.

Tolerance and physical dependence should be expected with long-term opioid treatment and not confused with psychological dependence or drug abuse, which is characterized by compulsive use of narcotics. Patients may crave narcotics and continue to consume it despite physiological or social damage consequent to their use. Tolerance of opioids may be defined as the need to increase dosage requirements over a period of time to maintain optimum pain relief. For most pain patients, the first indication of tolerance is a decrease in the duration of analgesia for a specific dose. Patients with stable disease do not usually require increasing doses. Increasing the dosage requirement is most consistently correlated with a progressive disease that produces more intense pain. Physical dependence on opioids is revealed when opioids are abruptly discontinued or when naloxone is administered; it typically manifests as anxiety, irritability, chills and hot flashes, joint pain, lacrimation, rhinorrhea, diaphoresis, nausea, vomiting, abdominal cramps, and diarrhea. The mildest form of the opioid abstinence syndrome may be manifested as viral flu-like syndromes. For short-acting opioids (i.e., codeine, hydrocodone, morphine, hydromorphone), the onset of withdrawal symptoms may occur within 6 to 12 hours and peak at 12 to 72 hours after discontinuation. For opioids with long half-lives (i.e., methadone and transdermal fentanyl), the onset of the withdrawal syndrome may be delayed for 24 hours or more after drug discontinuation. If a rapid decrease or a discontinuation of opioids is possible because the pain has been effectively eliminated, the opioid abstinence syndrome may be avoided by withdrawal of the opioid on a schedule that provides half the prior daily dose for each of the first 2 days and then reduces the daily dose by 25% every 2 days thereafter until the total dose (in morphine equivalent) is 30 mg/day. The drug may be discontinued after 2 days on the 30-mg/day dose, according to 1992 guidelines from the American Pain Society. Transdermal clonidine (0.1–0.2 mg/day) may reduce anxiety, tachycardia, and other autonomic symptoms associated with opioid withdrawal.

Diminishing opioid analgesic efficacy and increased pain during the course of opioid therapy is quite common. It is traditionally considered a result of opioid tolerance but could also be the result of opioid-induced hyperalgesia (OIH), which occurs when prolonged administration of opioids results in a paradoxical increase in atypical pain that appears to be unrelated to the original nociceptive stimulus. The mechanism of OIH is still unclear. However, opioid receptor desensitization, up-regulation of spinal dynorphin, and enhanced activity of excitatory transmitters such as NMDA are believed to be involved the pathogenesis of OIH (Silverman, 2009). Clinically, it is difficult to distinguish opioid tolerance and OIH. However, the issue of opioid-induced pain sensitivity should also be considered when an adjustment of opioid doses is being contemplated because opioid treatment is failing to provide the expected analgesic effects and/or there is an unexplainable pain exacerbation following a period of effective opioid treatment. Quantitative sensory testing of pain may offer the most appropriate way of diagnosing hyperalgesia. With OIH, an increased opioid dose is not always the answer. Office-based detoxification, reduction of opioid dose, opioid rotation, and the use of specific NMDA receptor antagonists are all viable treatment options for OIH.

Greater Occipital Nerve Block

Greater occipital nerve block is indicated for occipital neuralgia, commonly seen in patients after whiplash injury, falls on the back of the head, and other closed-head injuries. Patients are often misdiagnosed as having tension headache or migraine. These patients may have continuous headaches in the occipital, parietal, and sometimes the frontal region. The headaches may be increase several times a week and may be accompanied by nausea and vomiting. This condition is easily confused with migraine attacks, but physical examination may reveal positive tenderness over the greater occipital nerve. Palpation of the greater occipital nerve often makes the headache worse.

Greater occipital nerve block is the easiest interventional procedure for neurologists to perform in the office. For the procedure, one can palpate the posterior occipital protuberance, move 1.5 to 2 cm laterally, feel for the occipital artery pulsation and groove, then inject 2 to 3 mL of 0.5% bupivacaine with 20 mg of triamcinolone down to the bone and fan out (Fig. 44.4). According to this author’s data, for patients with occipital neuralgia after whiplash injuries, a greater occipital nerve block may provide immediate headache relief in 90% of patients and last for an average of 28 days. More rigorous clinical trials are needed to confirm the clinical efficacy of occipital nerve block for occipital neuralgia and cervicogenic headache (Ashkenazi et al., 2010). More research and education are warranted to increase clinician awareness of the existence of occipital neuralgia and cervicogenic headache, inasmuch as most neurologists seem more interested and well trained in examining the 12 pairs of cranial nerves than the greater occipital nerves.

Fig. 44.4 Occipital nerve block. For block of the greater occipital nerve, the patient sits in a chair resting the head on the arms above a treatment table.

Anecdotally, this author was called to consult on a headache patient with “normal neurological examination and MRI findings.” The patient was being treated with IV continuous hydromorphone, 2 mg/h, by a group of neurologists and residents at a major teaching university hospital in the United States. Examination revealed severe tenderness over the bilateral greater occipital nerves. Bilateral greater occipital nerve blocks immediately relieved the headache and made it possible to discontinue the IV hydromorphone. This patient reported that no one had touched her occipital area over several days of hospital stay, except for repeated MRI and CT scans, lumbar puncture, and multiple specialist consultations. The patient was discharged home with no headache. This case strongly suggests that examination of the greater occipital nerve should be a routine part of the physical examination of every headache patient.

Sphenopalatine Ganglion Block for Headache and Facial Pain

The sphenopalatine ganglion is a small triangular structure located in the pterygopalatine fossa, posterior to the middle turbinate and inferior to the maxillary nerve. It is covered by a thin layer (about 1 to 5 mm) of connective tissue and mucous membrane. Anesthetization of the sphenopalatine ganglion can be accomplished via the transnasal approach. The patient is placed supine on the treatment table with the nose pointed at the ceiling. A cotton applicator soaked with 2% to 4% lidocaine is inserted into the nose on the side of headache. To avoid mechanical discomfort, the cotton applicator should not be inserted deeply into the upper posterior wall of the nasopharynx. A slow drip of 2 to 4 mL of lidocaine over a 2- to 4-minute period into the nose through the cotton applicator often achieves the goal of a sphenopalatine ganglion block, with the local anesthetic flowing down to the back of nasopharynx by gravity. Sphenopalatine ganglion blocks have been reported to be effective in the relief of a wide variety of pain conditions of the head including acute migraine attacks, cluster headache, atypical facial pain, head and facial RSD, and postdural puncture headache (Cohen et al., 2009). Intranasal sphenopalatine ganglion block is safe and easy to perform in the clinic and may be helpful for neurologists without special training in interventional pain management techniques to treat an acute headache attack. Other methods for sphenopalatine ganglion block such as a lateral approach with fluoroscopic guidance or endoscopic sphenopalatine ganglion block have also been used. However, special training and equipment are needed.

Gasserian Ganglion Lesions for Trigeminal Neuralgia

The first choice for treatment of trigeminal neuralgia is carbamazepine. It can be used with other medication such as baclofen. Gasserian ganglion lesions are indicated when patients fail other medication treatments. These procedures include radiofrequency thermocoagulation, balloon compression, and glycerolysis. Radiofrequency thermocoagulation is the most commonly used procedure. This procedure is often performed by neurosurgeons, interventional pain specialists, or interventional radiologists with special training. The treatment requires inserting a radiofrequency needle through the face and foramen ovale into the base of the skull under the guidance of fluoroscopy, CT, or CT fluoroscopy. After the needle reaches the gasserian ganglion, radiofrequency energy is applied to induce thermocoagulation; 87% to 91% of patients experience immediate pain relief. In a 5-year follow-up, 50% of patients still had good pain relief. Common side effects include corneal anesthesia, masticator weakness, and anesthesia dolorosa. Recently, stereotactic radiosurgery for trigeminal neuralgia has been used more widely because of its noninvasive nature. Significant pain relief was achieved in 73% at 1 year, 65% at 2 years, and 41% at 5 years follow-up (Kondziolka et al., 2010). However, this procedure may be more costly than other procedures already mentioned.

Stellate Ganglion Block

The stellate ganglion is a sympathetic ganglion innervating the ipsilateral upper extremity, the neck, and the head. The structure is usually located in front of the junction between the C7 vertebral body and the transverse process. Stellate ganglion block is primarily indicated for complex regional pain syndrome of the head, neck, and upper extremities. Uncontrolled clinical reports indicate this procedure provides effective pain relief or may even reverse the course of early-stage CRPS type I. Other indications include vascular insufficiency of the arm and acute herpes zoster infection.

Technically, this block is achieved by inserting a needle through the neck to the front of the junction between the C7 vertebral body and the transverse process. Traditional hand palpation technique without guidance of fluoroscopy bears significant risks of injecting local anesthetics into critical structures in the neck such as the carotid and vertebral arteries or intrathecal space. Incorrectly located injections of local anesthetics may lead to loss of consciousness, seizures, paralysis, cardiac arrest, and death. Current use of fluoroscopic guidance for stellate ganglion block dramatically decreases the possibility of serious side effects and increases the rate of success.

Epidural Corticosteroid Injection

Pain specialists have used epidural corticosteroid injection (ESI) for decades to treat back and neck pain. The procedure is further divided into cervical, thoracic, and lumbar ESI (LESI), with the purpose of treating pain originating from different spinal regions. By 1995, there were at least 12 so-called double-blind placebo-controlled studies investigating the clinical efficacy of LESI for LBP. Of these studies, only six yielded positive results, while the other studies did not support the use of LESI for LBP. Actually, several of these studies exhibited the critical flaw of treating “low back pain” as a single entity. It is now realized that LBP is a clinical syndrome that may be caused by a variety of pathologies in the lumbar spine and adjacent organs. It is not reasonable to treat LBP with ESI, regardless of the cause. More recent well-designed placebo-controlled studies have provided clinical evidence that LESI decreases lumbar radicular pain caused by lumbar disk herniation (Roberts et al., 2009). The pain-relieving effect of LESI may last up to 3 months. Corticosteroids appear to speed the rate of recovery and return of function, allowing patients to reduce medication levels and increase activity while waiting for the natural improvement expected in most spinal disorders. Recent studies also support the use of LESI for pain relief in patients with spinal stenosis (Lee et al., 2010).

Past the age of 60, more than 90% of the normal population has a variety of degenerative spine changes including disk herniation, spinal stenosis, and foraminal stenosis. The majority of persons with these changes, however, do not have pain. It is now believed that the pain in patients with disk herniation and associated radiculopathy is not purely due to mechanical compression but is more likely due to chemical inflammation. A recent study provided convincing evidence for the role of inflammatory mediators in the pathogenesis of lumbar radicular pain and LBP in patients with lumbar degenerative diseases. In the study, the immunoreactivity of an array of cytokines was measured in lavage samples and compared with clinical response to the therapeutic injection. Ten subjects underwent repeated epidural lavage sampling 3 months after the steroid injection. It was found that interferon gamma (IFN-γ) was the most consistently detected cytokine. IFN-γ immunoreactivity was also highly correlated with reduction of pain 3 months after the epidural steroid injection. In subjects reporting significant pain relief (>50%) from the injection, mean IFN-γ immunoreactivity was significantly greater compared with patients experiencing no significant relief. The IFN-γ immunoreactivity in repeated lavage samples decreased to trace residual concentrations in patients who reported pain relief from the steroid injection. These results suggest that IFN-γ may be part of a biochemical cascade triggering pain in lumbar radicular pain (Scuderi et al., 2009). Other chemical substances such as phospholipase A₂, which is responsible for the liberation of arachidonic acid from cell membranes and starting the cascade of formation of inflammatory mediators such as prostaglandin E (PGE), is also believed to play a major role in pathogenesis of LBP. Epidural corticosteroid injection has been proven to suppress the functional activity of inflammatory mediators such IFN-γ and phospholipase A₂ (Scuderi et al., 2009) to decrease inflammation in the epidural space and surrounding nerve roots. With the support of evidence from both basic science and clinical studies, it is current common practice to offer patients with lumbar radicular pain due to disk herniation a trial of LESI before considering a surgical treatment for lumbar disk herniation. The procedure often prevents back surgeries. As long as pain is relieved and the patient is free of neurological deficits, a herniated disk should be left alone without further treatment.

Lumbar Facet Joint Block

The lumbar facet joint block procedure is indicated for lumbar facet joint pain syndrome. Lumbar facet joint syndrome may be found in up to 35% of patients with LBP. Clinically, this syndrome may mimic lumbar radiculopathy (sciatica). Patients may complain of LBP, often on one side, with pain radiating down the back or front of the thigh. Clinical examination may reveal tenderness on either or both sides of the lumbar spine over the lumbar facet joints. Lumbar spine extension and lateral rotation to the painful side may increase LBP because this maneuver increases pressure on the lumbar facet joints. The straight leg raising test is often negative. Traditionally, pain specialists have performed intrajoint corticosteroid injections, but over the last decade, this procedure has largely been replaced by a diagnostic medial branch (nerve innervating the lumbar facet joints) block with a small amount of local anesthetic. If the patient has significant pain relief (more than 50%) after two consecutive diagnostic medial branch blocks, facet joint rhizotomy with radiofrequency destruction of the medial branch will be performed to denervate the lumbar facet joints. A recent systematic literature review found moderate evidence to support the clinical efficacy and use of radiofrequency rhizotomy for lumbar facet joint syndrome (Datta et al., 2009).

Percutaneous Disk Decompression

Over 300,000 spine surgeries are performed each year in the United States. A majority of these surgeries are conducted for lumbar and cervical disk herniation. Traditional neurosurgical and orthopedic techniques for lumbar disk herniation include laminectomy, diskectomy, and fusion. A significant number of patients end up with so-called failed back surgery syndrome. Recurrent disk herniation, epidural abscess, scar tissue formation around nerve roots, facet joint syndrome, and muscle spasm may contribute to the clinical features of this syndrome. According to the recent literature, up to 100,000 new cases of failed back surgeries are produced every year in the United States alone as the result of spine surgeries. To avoid possible complications of open surgery, minimally invasive techniques for disk decompression have been developed. These techniques include chymopapain, the Nucleotome system, laser diskectomy, nucleoplasty, and disk Dekompressor.

Chymopapain is a proteolytic enzyme from the papaya fruit that may induce enzymatic decompression of the nucleus pulposus of a herniated disk. Initial clinical reports were highly positive, but serious side effects such as anaphylactic shock, transverse myelitis, and even death caused chymopapain to be largely replaced by other techniques.

Percutaneous Nucleotome was developed by a Japanese orthopedic surgeon, Dr. Hijikata, in 1975. This procedure inserts a 7-mm-diameter tube into the annulus and removes the disk material with specially designed forceps. The procedure has a reported success rate of 72%. However, because of the large diameter of the cannula, this technique is no longer commonly used. In 1986, Ascher and Choy introduced YAG laser diskectomy, a procedure still being used by spine surgeons, neurosurgeons, and some interventional pain specialists. This technique utilizes an 18-gauge probe and generates laser energy to evaporate part of the nucleus pulposus. It decreases the intradiscal pressure, with a reported success rate for back pain relief of 78% to 80%. Heat generated by the laser energy may cause patients to experience severe pain during the procedure and increased muscle spasm afterward.

Over the last decade, two new percutaneous disk decompression techniques have been reported. Introduced in 2000, DISC Nucleoplasty utilizes a unique plasma technology called Coblation to remove tissue from the center of the disk. During the procedure the DISC Nucleoplasty SpineWand is inserted into the center of the disk, where a series of channels are created to remove tissue from the nucleus. Disc DeKompressor was introduced in 2003. This procedure uses a 1.5-mm percutaneous lumbar diskectomy probe to aspirate the disk material. It is minimally invasive with less risk for nerve root damage. This technique is indicated for patients with contained disk herniation and lumbar radiculopathy. Observational studies suggest both Nucleoplasty and Disc DeKompressor may be potentially effective, minimally invasive treatments for patients with symptomatic contained disks. However, prospective randomized controlled trials are needed to confirm their clinical efficacy and to determine ideal patient selection for these procedures (Gerges et al., 2010).

Motor Cortex Stimulation

Motor cortex stimulation (MCS) has been used for the treatment of central and neuropathic pain syndromes since 1991. It has been used to treat medically unresponsive central and neuropathic pain including that due to thalamic, putaminal, and lateral medullary infarction, traumatic trigeminal neuropathy (not idiopathic trigeminal neuralgia), facial postherpetic neuralgia, brachial plexopathy, neuropathic pain after an SCI, phantom-limb pain, and CRPS. MCS has shown particular promise in the treatment of intractable neuropathic facial pain and central pain syndromes such as thalamic pain syndrome (Levy et al., 2010).

The MCS leads are surgically placed on the dura, with the target selected on the primary motor cortex based on somatotopic anatomical landmarks. The optimal stimulation level is that which provides the best pain relief yet does not cause a seizure, pain from dural stimulation, or EMG activity. Cortical stimulation is not indicated for patients with a history of seizures. Personality disorders such as severe depression or psychotic disorders must be screened out prior to using this procedure.

The precise mechanism for MCS in relieving pain remains unknown, but studies have demonstrated that it leads to an increase in cerebral blood flow in the ipsilateral thalamus, cingulate gyrus, orbitofrontal cortex, and midbrain. The extent of pain relief correlates best, however, with anterior cingulate gyrus blood flow. Rostroventromedial medulla (RVM) and the descending serotonergic pathway acting on the spinal 5-HT (1A) receptor may also contribute to spinal antinociception induced by M1 stimulation.

Spinal Cord Stimulation

Spinal cord stimulation (dorsal column stimulation) uses an array of electrodes placed in the epidural space immediately behind the spinal cord to stimulate the dorsal column of the spinal cord. The exact mechanism of SCS is unclear. However, it is believed that the gate-control theory of pain conduction plays a major role. When the dorsal column of the spinal cord is stimulated, it may attenuate the conduction of the pain signal on the spinothalamic tract through collateral inhibition. Inhibitory neurotransmitters such as GABA may also be involved.

As noted earlier in the chapter, patients should have a trial of SCS prior to permanent implantation. During the trial, a percutaneous lead is inserted through the skin into the epidural space. Once the tip of the lead reaches the appropriate level, it is connected to an external pulse generator. When the stimulator is turned on, the patient may feel tingling and numbness. If the painful area is covered by the stimulation, the pain is decreased by more than 50% and the patient is satisfied with the stimulation, a permanent implantation may be considered. The procedure of permanent implantation of the SCS is performed by pain specialists or neurosurgeons in an operating room. It requires percutaneous insertion of an electrode into the epidural space under the guidance of fluoroscopy. The tip of the electrode is threaded up to the T9-T11 level in the epidural space immediately behind the dorsal column for the treatment of low back and leg pain. The other end of the electrode is connected through a subcutaneous tunnel to an internal pulse generator buried under the skin in the low back or abdominal wall. The strength of the stimulation can be changed through a remote control. Common complications of SCS implantation include infection, migration of the electrodes, and failure of pain relief even after a “satisfied” trial. Serious complications such as spinal cord compression or epidural abscesses are rare.

SCS is indicated for failed back surgery syndrome, CRPS, and unremitting pain due to peripheral vascular disease. Multiple studies have found that SCS may also improve pain due to refractory angina and improve circulation in the coronary arteries. Some authors have reported treatment success with SCS for severe peripheral neuropathy, postherpetic neuralgia, chronic knee pain following total knee replacement, central pain in MS, and painful spasms of atypical stiff limb syndrome (Ughratdar et al., 2010). The value of SCS for amputation stump pain, phantom-limb pain, and SCI is yet to be established. Patients seeking SCS treatments usually have failed all other conservative treatments such as medication, physical therapy, and nerve blocks with anesthetics and/or corticosteroids. SCS is not indicated for severe depression and contraindicated for patients with a cardiac pacemaker or defibrillators.

Intrathecal Drug Delivery Systems

For patients with chronic severe pain, especially malignant pain, who are unable to tolerate the side effects of oral or IV medications, intrathecal delivery of medication offers a useful alternative. The technique of intrathecal delivery of medication has evolved since 1979. There are two kinds of pumps available in the United States: Codman and Medtronic intrathecal pumps. The pump is usually implanted subcutaneously in the abdominal wall. The pump contains about 18 to 50 mL of medication. It is connected to one end of a small-diameter tube that runs to the intrathecal space. The pump continuously delivers small amounts of medication directly into the lumbar cerebrospinal fluid. The Codman pump has fixed delivery rates of 0.5 ml or 1 ml/day. The concentration of medication has to be changed in order to change the daily dose of medication. Medtronic pumps are programmable with an external magnetic control to adjust the dosage and time of medication delivery.

Commonly used medications for pain management include morphine, hydromorphone, bupivacaine, clonidine, and ziconotide, a novel peptide that functions as a calcium channel blocker. Ziconotide was approved by the FDA in 2004 for treating intractable severe chronic pain, but its serious side effects have called the clinical use of this medication into question (Ziconotide, 2008). Baclofen is a GABA_B agonist. It has been used through an intrathecal delivery system for the treatment of severe spasticity and may also decrease the pain related to spasticity. Even though intrathecal opioid treatment was initially approved by the FDA for the treatment of patients with malignant pain, over the last decade, intrathecal opioids have been used extensively for nonmalignant pain such as failed back surgery syndrome. A retrospective cohort study with 3-year follow-up found a favorable outcome for intrathecal opioids. Some patients are able to eliminate oral opioids, although some increase in intrathecal opioid dosing may be required (Atli et al., 2010).