eMedicine - Pathophysiology of Chronic Back Pain : Article by Anthony H Wheeler

You are in: eMedicine Specialties > Neurology > Headache and Pain

Pathophysiology of Chronic Back Pain

Article Last Updated: Jul 9, 2007

Chronic low back pain (LBP) is the most expensive benign condition in industrialized countries and the most common cause of activity limitation in persons younger than 45 years. It is defined as pain that persists longer than 12 weeks and is often attributed to degenerative or traumatic conditions of the spine. Fibrositis, inflammatory spondyloarthropathy, and metabolic bone conditions are also cited as causes. Although acute LBP has a favorable prognosis, the effect of chronic LBP and its related disability on society is tremendous. Unlike acute LBP, chronic LBP serves no biologic purpose. However, it is a disorder that evolves in a complex milieu influenced by endogenous and exogenous factors, and it alters the individual's productivity to an extent beyond what the initiating pathologic dysfunction would have.

Approximately 80% of Americans experience LBP during their lifetime. An estimated 15-20% develop protracted pain, and approximately 2-8% have chronic pain. Every year, 3-4% of the population is temporarily disabled, and 1% of the working-age population is disabled totally and permanently because of LBP. LBP is second only to the common cold as a cause of lost work time; it is the fifth most frequent cause for hospitalization and the third most common reason to undergo a surgical procedure. Productivity losses from chronic LBP approach $28 billion annually in the United States.

LBP is defined as chronic after 3 months because most normal connective tissues heal within 6-12 weeks unless pathoanatomic instability persists. A slowed rate of tissue repair in the relatively avascular intervertebral disk may impair the resolution of chronic LBP.

Traumatic or degenerative conditions of the spine are the most common causes of chronic LBP. Although disk protrusion and herniation have been popularized as causes of LBP and sciatica, asymptomatic disk herniations on CT and MRI are common. Furthermore, the relationship between the extent of disk protrusion and the degree of clinical symptoms is not clear. A strictly mechanical or pathoanatomic explanation for LBP and sciatica has proved inadequate; therefore, the role of biochemical and inflammatory factors remains under investigation. In fact, this failure of the pathologic model to predict back pain often leads to an ironic predicament for the patient with LBP.

If diagnostic studies are unrevealing of a structural cause, physicians and patients alike question whether the pain has a psychologic, rather than physical, cause. Physical and nonphysical factors, interwoven in a complex fashion, influence the transition from acute to chronic LBP. The identification of all contributing physical and nonphysical factors enables the treating physician to enact a comprehensive approach with the best likelihood for success.

The estimated yearly prevalence of LBP is 5-20% in the United States and 25-45% in Europe. About 2% of American workers have compensable back injuries each year, ie, a staggering 500,000 cases of work-related spinal injuries. LBP accounts for 19% of all Workers' Compensation claims in the United States. According to the Bureau of Labor and Statistics, metal workers generated 76% of all claims of back strain and/or sprains in the United States. Jobs that require manual-handling activities accounted for more than half of all back pain reports. Injuries to the back were highest among truck drivers, operators of heavy equipment, and construction workers.

An estimated 4.1 million Americans had symptoms of an intervertebral disk disorder between 1985 and 1988, with an annual prevalence of about 2% in men and 1.5% in women. A study of 295 Finnish concrete workers aged 15-64 years revealed that 42% of men, and as many as 60% of those aged 45 years or older, reported having sciatica. When interviewed approximately 5 years later, the lifetime prevalence had increased from 42% to 59%.

Sciatica due to lumbar intervertebral disk herniations usually resolves with conservative treatment. However, it leads to surgery more often than does back pain alone. In a published review of more than 15,000 disk operations, the most common surgical level was L4-5 (49.8%), followed by L5-S1 (46.9%); only 3.4% were performed at levels higher than these. Surgical treatment for lumbar diskogenic syndromes is most common in the United States, where the estimated rate is at least 40% higher than that of in other countries and more than 5 times higher than rates in Scotland and England.

In the United States, the estimated cost of LBP in 1980 was $85 million dollars. Between 1971 and 1981, the number of Americans disabled by LBP grew at 14 times the rate of population growth. Resultant disability in Western culture has reached endemic proportions, with enormous socioeconomic consequences.

LBP is most prevalent in industrialized societies. Genetic factors that predispose persons of specific ethnicity or race to this disorder have not been clearly identified with respect to mechanical, diskogenic, or degenerative causes. Men and women are affected equally, but in those older than 60 year, women report low-back symptoms more often than men.

The incidence of LBP peaks in middle age and declines in old age, when degenerative changes of the spine are universal. Sciatica usually occurs in patients during the fourth and fifth decades of life; the average age of patients who undergo lumbar diskectomy is 42 years.

The lumbar spine forms the caudal flexible portion of an axial structure that supports the head, upper extremities, and internal organs over a bipedal stance. The sacrum forms the foundation of the spine through which it articulates with the sacroiliac joints to the pelvis. The lumbar spine can support heavy loads in relationship to its cross-sectional area. It resists anterior gravitational movement by maintaining lordosis in a neutral posture.

Unlike the thoracic spine, the lumbar spine is unsupported laterally and had considerable mobility in both the sagittal and coronal planes. The bony vertebrae act as specialized structures to transmit loads through the spine. Parallel lamellae of highly vascularized cancellous bone form trabeculae, which are oriented along lines of biomechanical stress and encapsulated in a cortical shell. Vertebral bodies progressively enlarge in cross-sectional area because gravitational loads increase from cephalic to caudal segments. Bony projections from the lumbar vertebra, including the transverse processes and spinous processes, maintain ligamentous and muscular connections to the segments above and below them.

The intervertebral disk is composed of the outer annulus fibrosis and the inner nucleus pulposus. The outer portion of the annulus inserts into the vertebral body and accommodates nociceptors and proprioceptive nerve endings. The inner portion of the annulus encapsulates the nucleus, providing the disk with extra strength during compression. The nucleus pulposus of a healthy intervertebral disk constitutes two thirds of the surface area of the disk and supports more than 70% of the compressive load.

The nucleus is composed of proteoglycan megamolecules can imbibe water to a capacity approximately 250% of their weight. Until the third decade of life, the gel of the inner nucleus pulposus is composed of approximately 90% water; however, the water content gradually diminishes over the next 4 decades to approximately 65%. Nutrition to the inner annulus fibrosis and nucleus pulposus depends on the diffusion of water and small molecular substances across the vertebral endplates because only the outer third of the annulus receives blood supply from the epidural space.

Repeated eccentric and torsional loading and recurrent microtrauma results in circumferential and radial tears in annular fibers. Some annular tears may cause endplate separation, which results in additional loss of nuclear nutrition and hydration. Coalescence of circumferential tears into radial tears may allow nuclear material to migrate out of the annular containment into the epidural space and cause nerve root compression or irritation.

Throughout youth (at least the first 2 decades), 80-90% of the weight of the trijoint complex of the lumbar spine is transmitted across the posterior third of the disk; however, as disk height decreases and the biomechanical axis of loading shifts posteriorly, the posterior articulations (ie, facet joints) bear greater percentages of the weight distribution. Bone growth (osteophytes) compensates for this increased biomechanical stress to stabilize the trijoint complex.

Over time, hypertrophy of the facets and bony overgrowth of the vertebral endplates contribute to progressive foraminal and central canal narrowing. In addition to relative thickening of the ligament flavum and disk herniation, these changes contribute to reduction of the anteroposterior canal diameter and foraminal patency with neural compression. Spinal stenosis reaches a peak later in life and may produce radicular, myelopathic, or vascular syndromes such as pseudoclaudication and spinal cord ischemia.

LBP is most common in the early stages of disk degeneration, in what Kirkaldy-Willis called the stabilization phase. Impaired healing of the intervertebral disk due to its poor and peripheral blood supply serving only the external third of the outer annulus has been introduced as a possible explanation for the divergent behavior of this structure, which can produce chronic nociception. Also, elucidation of biochemical changes that may sensitize the disk and other structures capable of nociception within the trijoint construct may contribute to this discrepancy.

Many studies have demonstrated that the intervertebral disk and other structures of the spinal motion segment can cause pain.

Kuslich et al used regional anesthesia in 193 patients who were about to undergo lumbar decompressive surgery for disk herniation or spinal stenosis.¹ Pain was elicited by using blunt surgical instruments or an electrical current of low voltage in 30% of patients who had stimulation of the paracentral annulus fibrosis and in 15% with stimulation of the central annulus fibrosis. However, it is unclear why mechanical back pain syndromes commonly become chronic, with pain persisting beyond the normal healing period for most soft-tissue or joint injuries in the absence of nonphysical or operant influences.

In 1987, Mooney proposed that this LBP chronicity was best explained by a tissue component of the spine that obeyed physiologic rules different from those of other connective tissues in the body.² This divergent behavior is best illustrated in the intervertebral disk with its composition of large, unique, water-imbibing proteoglycan molecules. During adulthood, these large molecules break into small molecules that bind less water than the original molecule. Repair by means of proteoglycan synthesis is slow. Fissuring and disruption of annular lamellae further exacerbate molecular breakdown and dehydration of the disk. Arterial blood supply to the peripheral one third of the outer annulus is meager and inadequate to prevent subsequent internal degeneration.

The annulus and nucleus pulposus are similarly compromised, as they receive nutrition only by means of diffusion through adjacent vertebral endplates. Although sluggish healing of the intervertebral disk may partially account for the tendency of a spinal lesion to lead to chronicity, a direct concordance between structural degeneration and spinal pain does not exist.

Recent elucidation of biochemical behaviors and neurophysiologic factors affecting the disk and other regional pain-sensitive tissues may account for this discrepancy. In humans, painful disks have a lower pH than that of nonpainful disks. Also, experimental lowering of the pH in animal models induced pain-related behaviors and hyperalgesia. Diskography of canine disks that were normally or experimentally deformed seemed to show increased concentrations of neuropeptides, such as substance P (SP), calcitonin gene-related peptide (CGRP), and vasoactive intestinal peptide (VIP) in the dorsal root ganglion (DRG), implicating their possible role in the transmission or the modulation of pain. SP probably modulates initial nociceptive signals received in the gray matter of the dorsal spinal cord.

Somatostatin is another neuropeptide in dense concentration the dorsal gray of the spinal cord. Somatostatin is released from the DRG after noxious thermal stimulation and likely plays a role in pain transmission and in producing neurogenic inflammation. Therefore, the release of neuropeptides like SP, VIP, and CGRP may occur in response to noxious biochemical forces and environmental factors (eg, biomechanical stress, microtrauma, vibration), stimulating the synthesis of inflammatory agents (eg, cytokines, prostaglandin E₂) and degradative enzymes (eg, proteases, collagenase). These factors cause progressive deterioration of the motion segment structures, especially the intervertebral disk.

Inflammatory factors may be responsible for pain in some cases in which epidural steroid injections provide relief. Corticosteroids inhibit the production of arachidonic acid and its metabolites (prostaglandins and leukotrienes), inhibiting phospholipase A2 (PLA2) activity. levels of PLA2, which plays a role in inflammation, are elevated in surgically extracted samples of human herniated disks. Furthermore, PLA2 may play a dual role, inciting disk degeneration and sensitizing annular nerve fibers. Afferent nociceptors in nerve roots may be sensitive to various proinflammatory mediators, which are inhibited by corticosteroids, such as prostanoids produced from arachidonic acid and released from cell membrane phospholipids by PLA2.

Research suggests that proinflammatory cytokines may also contribute to diskogenic pain by sensitizing nociceptors and their effect on disk degeneration by suppressing proteoglycan synthesis and increased diskal matrix degradation. Cytokines are produced in response to neural injury in the CNS and may play a role in spinal neural hypersensitization and chronic neuropathic pain. Cytokines known to play a role in nociception include nerve growth factor, interleukin (IL)-1, IL-6, IL-10, and tumor necrosis factor-alpha (TNF-alpha).

Corticosteroids can inhibit activity of TNF-alpha, which induces IL-1 and prostaglandin E₂ production. Once released, these substances contribute to early and late effects of the inflammatory process and stimulate nociception. A nociceptive role for nitric oxide (NO) in diskogenic pain syndromes is under investigation. NO levels are elevated in human disk herniations and when the hydrostatic pressure of the disk is increased due to biomechanical stressors. NO inhibits proteoglycan synthesis in cells in the nucleus pulposus, leading to proteoglycan loss, reduced water content, and disk degeneration.

Neurotransmitters and biochemical factors may sensitize neural elements in the motion segment so that the normal biomechanical stresses induced by previously asymptomatic movements or lifting tasks cause pain. Furthermore, injury and the consequential neurochemical cascade may initiate the degenerative and inflammatory changes described above, which mediate additional biochemical and morphologic changes and which modify or prolong the pain stimulus. Whether the biochemical changes that occur with disk degeneration are the consequence or cause of these painful conditions is unclear. However, chemical and inflammatory factors may create the environmental substrata through which biochemical stress and forces cause axial or limb pain with various characteristics and degrees.

The pathophysiology of spinal nerve-root or radicular pain is unclear. Proposed etiologies include neural compression with axonal dysfunction, ischemia, inflammation, and biochemical influences. Spinal nerve roots have unique properties that may explain their proclivity to produce symptoms. Unlike peripheral nerves, spinal nerve roots lack a well-developed intraneural blood-nerve barrier, a lack which makes them more susceptible to symptomatic compression injury than peripheral nerves and vulnerable to endoneural edema formation.

Increased vascular permeability caused by mechanical nerve-root compression can induce endoneural edema. Furthermore, elevated endoneural fluid pressure due to intraneural edema can impede capillary blood flow and may cause intraneural fibrosis. Also, spinal nerve roots receive approximately 58% of their nutrition from surrounding CSF. Perineural fibrosis, which interferes with CSF-mediated nutrition, renders the nerve roots hyperesthetic and sensitive to compressive forces.

Research has elucidated several vascular mechanisms that can produce nerve-root dysfunction. Experimental nerve-root compression showed that venous blood flow can be stopped at low pressures, ie, 5-10 mm Hg. The occlusion pressure for radicular arterioles is substantially higher than this, approximating the mean arterial blood pressure and showing a correlation with systolic blood pressure; this factor increases the potential for venous stasis.

Some investigators postulate that venous then capillary stasis causes congestion that, in turn, may induce symptomatic nerve-root syndromes. Nerve-root ischemia or venous stasis may also generate pathologic biochemical changes that cause pain, unlike the progressive sensory then motor dysfunction typically seen with peripheral nerve compression. Studies of ischemia experimentally induced with low-pressure nerve-root compression demonstrated that compensatory nutrition from CSF diffusion is probably inadequate when epidural inflammation or fibrosis is present. Rapid onset of neural and vascular compromise is more likely than slow or gradual mechanical deformity to produce symptomatic radiculopathy.

Research has revealed other possible causative mechanisms for symptomatic radiculopathy. A 1987 animal study showed that autologous nucleus pulposus placed in the epidural space of dogs produced a marked epidural inflammatory reaction that did not occur in the comparison group, which received saline injections. Similar studies have shown myelin and axonal injury to nerve roots, as well as, reduced nerve conduction velocities exposed to autologous nucleus pulposus. Recent studies have demonstrated that experimental radicular exposure to degenerative nucleus pulposus and annulus fibrosis does not produce the same dysfunctional nerve-root changes. Cells of the nucleus pulposus can induce local neural dysfunction and generate algogenic agents, such as metalloprotease (eg, collagenase, gelatinase), as well as IL-6 and prostaglandin-E₂.

Other biochemical substances, including TNF, have been implicated as causes. TNF increases vascular permeability and appears to be capable of inducing neuropathic pain. When injected into nerve fascicles, TNF produces changes similar to those seen when nerve roots are exposed to the nucleus pulposus. In addition, a still-unanswered issue is whether an autoimmune response occurs when nucleus pulposus is exposed to the systemic circulation because it is sequestered by the annulus fibrosis and because the immune system may not be recognized as normal. Indeed, research to date suggests that the cause of symptomatic radiculopathy is multifactorial and more complex than just neural dysfunction due to structural impingement.

The superior and inferior articular processes of adjacent vertebral laminae form the facet or zygapophyseal joints, which are paired diarthrodial synovial articulations that share compressive loads and other biomechanical forces with the intervertebral disk. Like other synovial joints, the facets react to trauma and inflammation by manifesting pain, stiffness, and dysfunction with secondary muscle spasm leading to joint stiffness and degeneration. This process is borne out, as previously described, through the degenerative cascade of the trijoint complex. Numerous radiologic and histologic studies have shown that diskal and facet degeneration are linked and that, over time, degeneration of the segment leads to osteoarthritis of the facets.

Studies of provocative intra-articular injection techniques demonstrated local and referred pain into the head and upper extremities from cervical facets, into the upper midback and chest wall from thoracic facets, and into the lower extremity from the lumbar facets. The fibrous capsule of the facet joint contains encapsulated, unencapsulated, and free nerve endings.

Immunohistochemical studies have demonstrated nerve fibers containing neuropeptides that mediate and modulate nociception (eg, SP, CGRP, VIP). SP nerve fibers have been found in subchondral bone and degenerative lumbar facets subjected to aging and cumulative biomechanical loading. In fact, SP levels are correlated with the severity of joint arthritis. The infusion of SP into joints with mild disease reportedly accelerated the degenerative process. Furthermore, these chemicals and inflammatory mediators have been linked to proteolytic and collagenolytic enzymes that cause degradation of the cartilaginous matrix and osteoarthritis. Therefore, evidence of nociceptive afferents and the presence of algogenic neuropeptides, such as SP and CGRP, in facets and periarticular tissues support a role for these structures as spinal pain generators. Clinical research has demonstrated facet pain in 54-67% of patients with neck pain, 48% of patients with thoracic pain, and 15-45% of patients with LBP.

The sacroiliac joint is a diarthrodial synovial joint that receives its primary innervation from the dorsal rami of the first four sacral nerves. Arthrography or injection of irritant solutions into the sacroiliac joint provokes pain with variable local and referred pain patterns into regions of the buttock, lower lumbar area, lower extremity, and groin. Determined by using variety of blocking techniques, the reported prevalences of sacroiliac pain have been widely variable (2-30%) in patients evaluated for chronic LBP.

Pain receptors in muscle are sensitive to a variety of mechanical stimuli, including pressure, pinching, cutting, and stretching. Pain and injury occur when the musculotendinous contractual unit is exposed to single or recurrent episodes of biomechanical overloading. Injured muscles are usually abnormally shortened, with increased tone and tension due to spasm or overcontraction. Injured muscles often meet diagnostic criteria for myofascial pain (MP) syndrome, a condition that Drs Janet Travell and David Simons originally described.

MP is characterized by muscles that are in a shortened or contracted state, with increased tone and stiffness, and that contain trigger points (TrPs). TrPs are tender, firm, 3- to 6-mm nodules that are identified on palpation of the muscles. TrP palpation provokes radiating, aching pain into localized reference zones. Mechanical stimulation of the taut band, a hyperirritable spot in the TrP, by needling or rapid transverse pressure often elicits a localized muscle twitch.

Sometimes, TrP palpation can elicit a jump sign, an involuntary reflex or flinching disproportionate to the palpatory pressure applied. MP may become symptomatic as a result of direct or indirect trauma, exposure to cumulative and repetitive strain, postural dysfunction and physical deconditioning. MP can occur at the site of tissue damage or due to radicular and other neuropathic disorders at sites where pain is referred. Muscles affected by neuropathic pain may be injured due to prolonged spasm, mechanical overload, or metabolic and nutritional shortfalls.

The pathogenesis of MP and TrPs remains unproven. To date, research suggests that myofascial dysfunction with characteristic TrPs is a spinal segmental reflex disorder. Animal studies showed that TrPs can be abolished by transecting efferent motor nerves or infusing of lidocaine; however, spinal transsection above the level of segmental innervation of a TrP-containing muscle does not alter the TrP response. Simons postulates that abnormal, persistently increased, and excessive acetylcholine release at the neuromuscular junction generates sustained muscle contraction and a continuous reverberating cycle. This cycle has been postulated to result in painful and dysfunctional extrafusal muscle contraction that forms the basis for MP and possibly the actual structural substrate of the TrP.

Neurophysiology of spinal pain and injury

Nociception is the neurochemical process whereby specific nociceptors convey pain signals through peripheral neural pathways to the CNS. In the setting of spinal injury, acute tissue damage to the spinal motion segment and associated soft tissues activates these pathways. When the peripheral source of pain persists, intrinsic mechanisms that reinforce nociception influence the pain. The nervous system can enhance a pain stimulus generated by tissue damage to levels far greater than any threat it signifies to the human organism; this is a common clinical scenario in cases of chronic spinal pain.

Peripheral transmission of pain stimuli through A-delta fibers and C-fibers leads to the release of excitatory amino acids, such as glutamine and asparagine, which then act on N-methyl-D-aspartic acid (NMDA) receptors, causing the release of the neuropeptide SP. Neuropeptides, such as SP, CGRP, and VIP are transported to the endings of nociceptive afferents, which inflammation and other algogenic mechanisms sensitize. These neurochemical changes exaggerated, adverse sensitize nociceptors so that they respond to mild or normal sensory stimuli, such as light touch or temperature change (allodynia).

Transduction is the process whereby noxious afferent stimuli are converted from chemical to electrical neural messages in the spinal cord that communicate cephalad to the brainstem, thalamus, and cerebral cortex. Noxious mechanical, thermal, and chemical stimuli activate peripheral nociceptors that transmit the pain message through lightly myelinated A-delta fibers and unmyelinated C-fibers. Nociceptors are present in the outer annular fibrosis, facet capsule, posterior longitudinal ligament, associated muscles, and other structures of the spinal motion segment.

Algogenic substances that are typically involved in tissue damage and that can induce peripheral transduction include potassium, serotonin, bradykinin, histamine, prostaglandins, leukotrienes, and SP. Transduction leads to transmission, which is the conduction of afferent pain signals to the DRG and dorsal horn of the spinal cord. The DRG contains cell bodies of primary afferent nociceptors, including the neuropeptides SP, VIP, and CGRP. The DRG is mechanically sensitive and capable of independent pain transduction, transmission, and modulation.

Nociceptive modulation first occurs in the dorsal horn, where nociceptive afferents converge to synapse on a single WDR neuron. WDR neurons respond with equal intensity regardless of whether the neural signal is noxious (hyperalgesia). Hyperalgesia and allodynia initially develop at the injury site; however, when central sensitization occurs by means of WDR neural activity, the area of pain expands beyond the initial region of tissue pathology.

Finally, a phenomenon termed wind-up results from repetitive activation of C-fibers sufficient to recruit second-order neurons that respond with progressively increasing magnitude; NMDA receptor antagonists can block this effect. Wind-up probably also contributes to CNS sensitization, including hyperalgesia, allodynia, and persistent pain. These nociceptive mechanisms, which reinforce the pain signal, frequently recruit sympathetic nervous system involvement. Elevated norepinephrine levels in injured areas increase pain sensitivity by means of regional vasomotor and sudomotor changes. Also, high acetylcholine levels can augment ongoing local and regional involuntary muscle contraction and spasm.

Chronic LBP is not simply the same as acute LBP that persists for a greater duration. Usually 6-7 weeks is sufficient for healing to occur in most soft-tissue or joint injuries; however, 10% of LBP injuries do not resolve in this period. The evolution of chronic LBP is complex, with physiologic, psychological, and psychosocial influences. These influences can be divided into 3 major categories, with subcategories, as follows:

Peripheral mechanisms may reinforce nociception when the source of pain persists. If an ongoing pathologic condition causes the peripheral pain stimulus, continuous nociception may induce repetitive stimulation or sensitization of pain receptors and nerve fibers so that they adversely respond to even mild or normal sensory stimuli (ie, allodynia). Furthermore, liberation of algogenic and other substances from damaged tissues may induce changes in the microenvironment by means of neuroactive, biochemical, inflammatory, or vasoactive effects that activate or increase the sensitivity of nociceptors.

Peripheral-to-central processing may also modify nociception. Persistent tissue damage may stimulate afferent nerve fibers that project to internuncial neurons in the spinal cord and thereby set up neuronal loops of continuous self-sustaining abnormal reverberating nociceptive activity. Peripheral inhibition, a mechanism for reducing the intensity of an afferent pain signal, may be impaired owing to persistently malfunctioning or diseased large peripheral myelinated fibers, which normally dampen nociception (eg, peripheral neuropathy, epidural scarring, chronic herniated disk material).

Ectopic impulse generation is a theoretical mechanism Wall and Gutnick proposed.³ Damaged sensory nerves, such as neuromata or demyelinating lesions in peripheral nerves, produce aberrant signals. Deafferentation hypersensitivity also purportedly causes abnormal and chronic nociceptive firing patterns.

CNS bias of the signal may occur in the spinal cord, brainstem reticular formation, or cortex. The brainstem reticular formation acts to direct the attention of the CNS toward or away from central and peripheral stimuli. Depending on the degree of focus, or the lack thereof, transmission of pain signals may be enhanced or inhibited. Furthermore, cortical influences, such as cognitive and affective disorders, may affect the intensity of the processed pain signal.

Psychological manifestations are 3-fold; they include behavioral, cognitive-affective, and psychophysiological mechanisms. Guarded movement, nonverbal and verbal expression of pain, and inactivity are called pain behaviors. Normal well-behavior patterns may become extinguished when these verbal and nonverbal pain behaviors are reinforced by environmental factors.

Cognitive-affective mechanisms often contribute to the perception of chronic pain. Pain complaints are common in depressed individuals, and patients with chronic pain frequently become depressed. Depression acts though biochemical processes similar to those that are operative in chronic pain; this may enhance symptoms through a synergistic relationship. Depressed patients may illogically interpret and distort life experiences, further complicating the feasibility of treatment or employment.

Psychophysiologic mechanisms naturally triggered by pain and injury can lead to generalized muscle overactivity, increased fatigue, and other pain problems (eg, tension myalgia, headache). The emotional stress that pain induces tends to heighten norepinephrine activity and activity of the sympathetic nervous system, which may further amplify nociception by means of peripheral or central mechanisms.

Barriers to recovery may be premorbid, caused by traumatic injury, or they develop over time as a result of psychological and environmental influences. These barriers strongly influence chronicity and the patient's prognosis. For example, medical problems, such as diabetes or heart disease, may make the patient a poor candidate for rehabilitation or surgery. Failed back surgery may create permanent physical and psychological impasses.

Patients differ in their inherent capacity to exercise. Deconditioning syndrome, a term Mayer coined, is caused by prolonged reduction of physical activity due to chronic LBP. This syndrome is associated with gradual reduction in muscle strength, joint mobility, and cardiovascular fitness, which over time may become a self-sustaining and independent component of the individual's musculoskeletal illness.

Preexisting psychological factors may combine with low back injuries to create a pain syndrome with predominantly psychiatric features. Psychiatric interviews of 200 patients with chronic LBP entering a functional restoration (FR) program revealed that 77% met lifetime diagnostic criteria for psychiatric syndromes, even when the category of somatoform pain disorder was excluded. In addition, 51% met criteria for at least 1 personality disorder.

Personality disorders or traits often affect the prognosis. People with borderline personalities may acquire pain as a method for structuring an otherwise empty existence, whereas narcissistic patients may acquire pain and seek medical attention as a way of preventing more serious illness. Those with an antisocial personality are often exploitative and prone to complications, and they may easily adopt game-playing roles. Patients with somatizing and hypochondriacal conditions are most likely to develop pain as a symptom and least likely to respond to treatments aimed at a presumed organic cause. Individuals with depression are prone to chronic pain or to have pain as a symptom. Other personality disorders or traits that may influence chronic pain include those paranoid, passive-aggressive, and avoidant conditions.

Previous learning and role models also affect the patient's prognosis and treatment outcome. An individual's cognitive or attribution style (ie, the patient's tendency to catastrophize, overgeneralize, personalize, or selectively attend to negative aspects of the pain experience) heavily influence his or her prognosis. The physical and emotional trauma that occurred during the injury or that was encountered during the ordeal of convalescence may contribute to the psychosocial milieu and create a host of emotional responses, including anxiety and fear. Psychophysiologic responses may be reinforced and include nightmares, palpitations, diaphoresis, headaches, dizziness, irritability, and fatigue. Patients are often overwhelmed and have feelings of abnormal dependence. They perceive a loss of control and look to their physician, attorney, or family for guidance. Some advisors may be oversolicitous or encourage compensation or litigation, creating further barriers to recovery.

Enduring prolonged pain also may cause emotional disturbances. Depression has already been mentioned as a common partner to chronic pain and is enhanced by loss of physical function, low self-esteem, loss of employment, and financial insecurity. Heightened anxiety may occur secondary to continued pain and the associated life disruption. Fear of injury and panic symptoms may also enhance anxiety and complicate the person's recovery. Anger or hostility directed at the workplace or perceived ineffective medical care may hinder communication with physicians, employers, family, and friends. As the length of time from injury increases, the aggregate of posttraumatic emotions becomes increasingly complex; avoidance learning and deactivation further complicate the situation.

As these barriers accumulate, the probability of a poor prognosis is enhanced. Neuropsychological factors may preexist or occur as a result of the injury. Limited cognitive dysfunction, either premorbid or from brain injury, may limit the patient's capacity to make decisions or succeed in a rehabilitation program.

Environmental and social influences may play the strongest role in determining the patient's prognosis for recovery. Job dissatisfaction or conflict is a key predictor of chronic LBP with disability. Compensated unemployment may reinforce chronicity in these cases. Family, financial, and legal issues also affect chronicity. The patient with chronic LBP may be unable to return to a previous strenuous or heavy job and may be poorly equipped to pursue alternative vocational options because of a lack of education. Older individuals may have reduced capacity for work and less vocational potential; therefore, loss of compensation becomes an overriding issue.

History

In most cases, chronic LBP has been investigated with appropriate physician evaluation and perhaps imaging studies. Characterization of the pain as mechanical is a primary goal when a history is obtained from a patient with chronic LBP and sciatica. Mechanical or activity-related spinal pain is most often aggravated by static loading of the spine (eg, prolonged sitting or standing), long-lever activities (eg, vacuuming or working with the arms elevated and away from the body), and levered postures (eg, forward bending of the lumbar spine). Pain is reduced when multidirectional forces balance the spine (eg, walking or constantly changing positions) and when the spine is unloaded (eg, reclining). Patients with mechanical LBP often prefer to lie still in bed, whereas those with a vascular or visceral cause are often found writhing in pain, unable to find a comfortable position.

Unrelenting pain at rest should suggest a serious cause, such as cancer or infection. Imaging studies and blood workup are usually mandatory in these cases and in cases of progressive neurologic deficit. Other historical, behavioral, and clinical signs that should alert the physician to a nonmechanical etiology requiring diagnostic evaluation are outlined below.

Nonphysiologic or implausible descriptions of pain may provide clues that operate; other psychosocial influences coexist.

Physical examination

Physical examination is important to confirm a mechanical or benign cause for the patient's LBP. Observations of verbal and nonverbal behaviors suggesting symptom magnification should be noted. Inspection of the spine requires the patient to disrobe. Open-back gowns give the physician only 1 view of the spine; therefore, swimming attire is often appropriate for complete, 360° inspection. Leg-length discrepancy and pelvic obliquity, scoliosis, postural dysfunction with forward head and shoulders, or accentuated kyphosis should be noted. Physicians' preferences vary in regard to the importance of testing range of motion; however, just asking the patient to bend forward often enables the most worthwhile observations.

The patient is asked to drop his or her head and shoulders forward and then move slowly into forward bending. Normal forward bending is revealed when the patient recruits from each cephalic segment to the level below, and so on, progressing from the cervical spine through the thoracic and lumbar region, where flexion of the hips completes the excursion into full flexion. Patients with clinically significant mechanical back pain or lumbar segmental instability usually stop cephalic-to-caudal segmental recruitment on reaching the thoracolumbar junction, or sometimes the involved lumbar level. To continue forward bending, they then contract their lumbar muscles to brace the mechanically compromised segment and then continue recruitment in a reverse direction, beginning with motion through the hips, then proceeding cephalad, level to level, completing the excursion of the spine to the erect posture.

In cases of severe mechanical back pain and segmental instability with regional muscular spasm, the patient often reports an inability to perform any flexion below a thoracic spinal level. Any soft-tissue abnormalities and tenderness to palpation should be recorded. Palpation of lumbar paraspinal, buttock, and other regional muscles should be performed early in the examination. The examiner should palpate and note areas with superficial and deep-muscle spasm, and he or she should identify TrPs and small, tender nodules in a muscle that elicit characteristic regional referred pain.

Dissociation of physical findings from physiologic or anatomic principles is the key in patients in whom psychological factors are suspected to be influential. Examples of this phenomenon include nondermatomal patterns of sensory loss, nonphysiologic demonstration of weakness (give-way weakness when not caused by pain, or ratchety weakness related to simultaneous agonist and antagonist muscular contraction), and dissociation between the lumbar spinal movements found during history taking or counseling sessions from movements observed during examination.

Assessment of Waddell signs have been popularized as a physical-examination technique to identify patients who have nonorganic or psychogenic embellishment of their pain syndrome. Examination techniques that Waddell proposed consist of simulated rotation of the hips en masse with the lumbar spine without allowing for spinal rotation; this maneuver normally does not cause pain. Another is the application of light pressure on the head, which should also be painless. Likewise, gentle effleurage of superficial tissues is unlikely to cause pain. Other techniques are striking dissociation between sitting and supine straight leg raising and the demonstration of nonphysiologic weakness and sensory deficits by the patient, as mentioned already.

Straight leg raising with the patient supine should produce ipsilateral leg pain between 10° and 60° to be declared positive. Straight leg raising that produces pain in the opposite leg carries a high probability of disk herniation, and investigation should be considered, especially if neurologic evidence for radiculopathy is present. Nonspecific complaints, overtly excessive pain behavior, patient contraction of antagonist muscles that limit the examiner's testing, or tightness of buttock and hamstring muscles are commonly mistaken for positive results on straight leg raising. Reverse straight leg raising may elicit symptoms of pain by inducing neural tension on irritated or compressed nerve roots in the mid-to-upper lumbar region. In addition, this maneuver helps the astute physician identify tightness of the iliopsoas muscle, which commonly contributes to chronic lumbar discomfort.

Neurologic evaluation is performed to determine the presence or absence and levels of radiculopathy or myelopathy. Anatomic localization is determined by muscle and reflex testing combined with historical information obtained during the interview and the absence of neurologic symptoms or signs that implicate cerebral or brainstem involvement. Consistent myotomal weakness and sensory findings that at least seem to coincide with segmental radiculopathy or polyradiculopathies should not be ignored.

The neurologist should identify syndromes of the lower motor neurons versus the upper motor neurons and the level of spinal dysfunction. Hyperreflexia in caudal spinal levels may change to reduced or absent reflexes in the upper extremities, determining radicular or spinal-cord localization of dysfunction. Rectal examination is indicated in patients in whom myelopathy, especially cauda equina syndrome, is a diagnostic concern. Tone of the anal sphincter; presence or absence of an anal wink; and correlation with motor, sensory, and reflex findings are appropriate in these cases.

Causes

Epidemiologic data suggest that risk factors include cigarette smoking; morbid obesity; occupations that require repetitive lifting, especially in forward bending and twisting positions, particularly when lifting requirements exceed the worker's physical capacity; and exposure to vibration caused by motor vehicles or industrial machinery. Studies indicate that smoking is most likely to be a risk factor for LBP in people with jobs that require heavy physical exertion. Fitness may be correlated with recovery and return to work after LBP; however, in prospective studies controlled for age, isometric lifting strength and degree of cardiovascular fitness were not predictive of back injury.

Occupational risk factors are difficult to define because exposures to specific causative influences are unclear, mechanisms of injury may be confusing, and supporting research findings are variable and conflicting for most environmental risks. Furthermore, job dissatisfaction, work conditions, legal and social factors, financial stressors, and emotional circumstances heavily influence back disability. Although many experts agree that heavy physical work, lifting, prolonged static work postures, simultaneous bending and twisting, and exposure to vibration may contribute to back injuries, the medical literature provides conflicting support for most of these proposed risk factors.

Extreme height and morbid obesity may predispose an individual to back pain. However, research studies have not clearly demonstrated that height, weight, or body build are directly related to the risk of back injury. Weakness of the trunk extensor muscles, compared with flexor strength, may be a risk factor for sciatica.

When LBP persists beyond 3 months, into the chronic phase, appropriate clinical and diagnostic information supporting a benign or mechanical cause should be accrued, if it has not been already. Also, prompt physician evaluation, including reasonable radiographic, laboratory, and electrophysiologic testing, is indicated in patients with persistent severe neurologic deficit, intractable limb pain, suspected systemic illness, or changes in bowel or bladder control. The spectra of mechanical (or activity-related) and nonmechanical causes of LBP are outlined below.

Differential diagnosis can include many neurological and systemic disorders, as well as referred pain from viscera or other skeletal structures such as the hip.

As indicated in the last section, unrelenting pain at rest should generate suspicion for cancer or infection. The appropriate imaging study is mandatory in these cases and in cases of progressive neurologic deficit. Plain anteroposterior and lateral lumbar spine radiographs are indicated for patients older than 50 years and for those with pain at rest, a history of serious trauma, or other potential conditions (eg, cancer, fracture, metabolic bone disease, infection, inflammatory arthropathy). The yield for discovering a serious condition with radiography outside these parameters is minimal, and the cost savings are substantial.

When LBP and sciatica persist into the subacute phase (pain lasting 6-12 wk), appropriate consultation and diagnostic imaging should be considered. Referring the patient to a physician with expertise in spinal disorders may be the most appropriate procedure for initial evaluation rather than relying on expensive diagnostic testing.

CT scanning is an effective diagnostic study when the spinal and neurologic levels are clear and bony pathology is suspected.

MRI is most useful when exact spinal and neurologic levels are unclear, when a pathologic condition of the spinal cord or soft tissues is suspected, when postoperative disk herniation is possible, or when an underlying infectious or neoplastic cause is suspected.

Myelography is useful to elucidate nerve-root pathology, particularly in patients with previous lumbar spinal surgery or with a metal fixation device in place. CT myelography provides accurate visual definition to elucidate neural compression or arachnoiditis when patients have undergone several spinal operations and when surgery is considered for the treatment of foraminal and spinal-canal stenosis.

When leg pain predominates and imaging studies provide ambiguous information, clarification may be gained by performing electromyography (EMG), somatosensory evoked potential (SSEP) testing, or selective nerve root blocks. When the cause of sciatica is related to neural compression by bony or soft-tissue structures in the spinal canal, surgical consultation should be considered. If diagnostic information is inadequate to explain the degree of neurologic deficit, pain, and disability, multidisciplinary evaluation may provide insight into perpetuating physical and psychosocial factors (see Media file 1).

Overview

The rationale for nonoperative treatment has been based on findings from clinical and autopsy studies, which demonstrate that protruding or extruded disk material can be resorbed over time.

In an uncontrolled study, Saal et al demonstrated that 90% of patients with a definite herniated lumbar disk and radiculopathy (criteria for surgical intervention), were successfully treated with aggressive rehabilitation and medical therapy.

A German neurologist, Henrik Weber, performed a longitudinal study of patients who had sciatica and confirmed disk herniations. The patients were randomly assigned to operative or nonoperative treatment. At 1-year follow-up, improvement was significantly greater in the surgically treated group than in the other. However, improvement in the groups did not significantly differ at 4-10 years of follow-up.⁴

Medical treatment

Treatment of spinal disorders can be divided into 3 phases based on the duration of symptoms. Primary nonoperative care consists of passively applied medical and physical therapy during the acute phase of soft-tissue healing. Secondary treatment includes spine-care education and active exercise programs that bridge the acute or the postoperative phase with return to work at previous level of function. Tertiary treatment focuses on interdisciplinary care to address LBP and disability, as well as acquired physical and psychological deconditioning that has developed as a result of chronic dysfunction.

Bed rest is usually appropriate for acute LBP and sciatica. Two days of bed rest is more effective than 7 days, resulting in less time lost from work. Prolonged bed rest can lead to progressive inactivity and avoidance, which reinforces abnormal illness behaviors. Such inactivity can also have deleterious physiologic effects, leading to shortened muscles and other soft tissues, joint hypomobility, reduced muscle strength, reduced cardiopulmonary endurance, and loss of mineral content from bone. For these reasons, bed rest is usually not recommended as a treatment for chronic LBP.

Principles for medical management in the chronic stages of LBP and disability differ in several ways from those in the acute phase.

Tricyclic antidepressants are useful in chronic LBP to alleviate insomnia, enhance endogenous pain suppression, reduce painful dysesthesia, and eliminate other painful disorders (eg, headaches). In addition, these medications may improve the patient's ability to cope, and they may reduce depression, anxiety, or fatigue associated with chronic LBP.

Calcium channel blockers and alpha-adrenergic antagonists are useful for treating LBP when it is associated with a complex regional pain syndrome.

On occasion, narcotics may be used to maintain function and mobility in a patient who has an acute exacerbation of chronic pain. However, continuous use of opioid analgesia for LBP is generally reserved as a final treatment option. When necessary, long-acting opioids should be given on a time-contingent dosing schedule rather than as needed. Improved function for achievement of vocational, recreational, and social goals are better measures of medication efficacy than subjective estimates of pain relief.

Local anesthetics, corticosteroids, or other substances may be directly injected into painful soft tissues, facet joints, nerve roots, or epidural spaces. They may also be given intrathecally. Therapeutic injections have been advocated to alleviate acute pain or an exacerbation of chronic pain, to help patients remain ambulatory outpatients, to allow them to participate in a rehabilitation program, and to decrease their need for analgesics, and to avoid surgery. Local injections into paravertebral soft tissues, specifically into myofascial trigger points, are widely advocated. However, a double-blind study to compare local anesthetic with saline injections and a prospective randomized double-blind study to compare dry needling with acupressure spray applications of lidocaine, corticosteroids, and vapor coolants revealed no statistically significant difference in therapeutic effects. Other procedures use "provocate [sic] and ablate" methods to identify potential and contributing generators of spinal pain, which can then be treated with chemical or thermal neurolysis. In some cases, selective radiofrequency ablation of the lesion is performed.

Intra-articular injections of the facet joints are advocated by many experts as a method for diagnosis and treatment of spinal pain.^{5, 6} Four studies of intra-articular corticosteroid injections in lumbar spine facet joints,^{7, 8, 9, 10} and one study in cervical spine joints¹¹ were performed using comparison groups that were similar demographically to the treatment group, but received another treatment. Two trials were randomized, one by Carette et al involving lumbar facet injections,⁷ and another by Barnsley et al involving cervical facet injections¹¹. Carette et al randomized 101 patients who reported more than 50% pain relief following a single intra-articular lidocaine block into a group that also received methylprednisolone. At 1-month follow-up, 42% of the methylprednisolone group and 33% of the saline group reported significant pain relief; however, at 6-month follow-up, a statistically significant comparison was identified. 46% of patients in the methylprednisolone group continued to experience pain relief compared with 15% of the patients in the saline group.

An analysis and synthesis of the evidence by Manchikanti et al excluded other referenced studies that demonstrated significant methodological flaws.⁵ Nonrandomized trials and observational studies have shown better results with reports of significant long-term pain relief (6 mo) ranging from 28% to 38% to 54%.

A retrospective evaluation by Lippitt showed initial relief in 50% of participants, however, only 14% of participants claimed relief at 6 months, and only 8% at 12 months.¹² Lau et al, in a retrospective review, reported initial relief in 56% of the patients, 44% with continued relief at 3 months, and 35% after 6 months.¹³ Although some physicians advocate facet injections as a treatment method, a large prospective study¹⁴ showed no long-term benefit. Therefore, intra-articular facet injections appear to have dubious therapeutic value when used in isolation. Although opinions and success rates of facet injections vary widely, the use of intra-articular facet injections is widely supported primarily for their diagnostic use.

Medial branch blocks have traditionally been used for both diagnostic and prognostic purposes but have demonstrated limited use as a therapeutic tool. The therapeutic role of medial branch blocks was evaluated in 3 randomized clinical trials^{15, 8, 16} and 3 nonrandomized clinical trials.^{17, 18, 19} Only one of the randomized trials used appropriate criteria to diagnose facet joint pain and showed adequate long-term follow-up of outcome data.¹⁵ Patients with chronic low back pain who failed standard nonoperative therapies were randomized into a treatment or comparison group. Both groups received blocks with anesthetic and Sarapin; however, the treatment group also received methylprednisolone as part of the injectate. All patients reported significant relief in the first 3 months, 82% between 4-6 months, and 21% between 7-12 months. Improvements were also noted in physical, functional, psychological, and return-to-work status. In the previously cited evidence-based review by the same author, median branch blocks were strongly supported for short-term pain relief and moderately supported for long-term relief of facet joint pain.⁵

Percutaneous radiofrequency (RF) neurotomy of the medial branches causes temporary denaturing of the nerves to the painful facet, but this effect may wear off when axons regenerate. In a 2000 review, Manchikanti et al cite strong evidence that RF denervation provides short-term relief (<6 mo) and moderate evidence for long-term relief (>6 mo) of chronic cervical, thoracic, and lumbar spinal pain of facet origin.²⁰ A randomized trial by Lord compared 12 patients receiving medial branch RF lesions of the cervical dorsal rami to the same number of patients receiving a sham procedure.²¹ Seven patients in the treatment group and one in the control group remained free of pain. Overall, patients receiving medial branch neurotomies had a long-term success rate of 75%. In another randomized trial, 47% of the treatment group showed sustained improvement following RF denervation at 12 months. Improvement measures included reduction of pain, functional disability, and physical impairment. These and other studies show strong support for both a short and long-term benefit from RF medial branch neurotomy for the treatment of lumbar facet syndrome in chronic low back pain patients.⁵ Potential side effects of RF denervation have included painful cutaneous dysesthesia or hyperesthesia, pneumothorax and deafferentation pain.²²

Epidural injections have been widely used by direct placement near the involved nerve root or by midline presentation, including caudal entry, and combining corticosteroid and local anesthetic of varying volumes. An intralaminar entry is directed more closely to the site of assumed pathology and requires less volume of the injectate than a caudal route. However, the caudal entry is usually considered a safe approach with small risk for inadvertent puncture of the dura or a neural structure. Transforaminal corticosteroid injections are more target-specific and require the least volume of injectate to reach the presumed pathoanatomic site or primary pain generator by an approach through the ventral lateral epidural space.

When considering an epidural injection, each approach has its advantages and disadvantages. The caudal approach requires a large fluid volume, thus greater dilution of the active ingredient within the injectate. Because the needle cannula is initially threaded at a relatively parallel plane to the spinal canal, the risk of intravascular, subcutaneous, subperiosteal, or interosseous needle puncture is greater.

Disadvantages of the intralaminar approach can include overdilution of the injectate, extra-epidural or intravascular placement of the needle, preferential cranial and posterior flow of the solution, and dural puncture. The intralaminar approach is also more difficult in postsurgical patients and below the L4-5 level.⁵ The transforaminal approach is difficult in the presence of a surgical osseous fusion or when spinal instrumentation is present. Other risks include intraneural or intravascular injection and spinal cord trauma. The use of fluoroscopy to direct needle placement and observe contrast flow should be considered requisite to reduce these potential adverse events.^{5, 6}

Evidence synthesis by Manchikanti involved review of 8 randomized or double blind trials. Five supported short-term relief^{23, 24, 25, 26, 27} (defined as <3 months) and 5 also supported long-term relief (defined as �3 months) when a caudal injection approach was used.^{23, 25, 26, 27, 28} In addition, 3 prospective trials^{29, 30, 31}, and 4 retrospective trials^{32, 33, 34, 35}, demonstrated support for short and long-term pain relief when epidural injections were performed in a series, rather than a single injection. An evidence synthesis for intralaminar epidural injections by Manchikanti et al showed 7 of 10 randomized trials to be positive for short-term relief, and 3 for long-term relief.⁵ Numerous nonrandomized trials showed patients benefiting from cervical or lumbar intralaminar epidural steroid injections.⁵ At present, the literature strongly supports the use of intralaminar corticosteroid epidural injections for providing short-term pain relief when treating cervical or lumbar radicular syndromes, even chronic cases; therefore, this treatment is best reserved for use as an adjunctive therapy or during a flare-up of symptoms.⁵ Also, transforaminal epidural injections have shown positive short- and long-term results in multiple randomized trials.^{5, 6}

Despite the literature support just cited, the debate regarding the use and benefit of epidural steroid injections for spinal pain patients still continues. Reviews by Koes et al in 1995 and 1999 supported the usefulness of lumbar and caudal epidural injections for low back pain and sciatica.³⁶ Meta-analyses in 1995 by Watts and Silagy³⁷ and in 1998 in by van Tulder et al³⁸ report conflicting evidence and inconsistent findings regarding the effectiveness of epidural steroids. A 1998 review of the literature³⁹ concluded that epidural corticosteroid injections were effective for back pain and sciatica, and, subsequently, a 2000 review by Vroomen et al cited epidural steroids as beneficial for some patients with nerve root compression and sciatica.⁴⁰

In summary, the spinal interventionalist must use clinical judgment as to the rationale and safety of performing the chosen procedure in combination with clinical experience when deciding how to proceed with treatment. Therefore, in some patients, epidural injections may be considered useful as a method of pain control and may provide benefit that is adjunctive to other therapies.⁶

Percutaneous adhesiolysis with or without spinal endoscopy is another interventional technique used to manage chronic refractory low back pain.^{41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54} This procedure is performed to disrupt presumed epidural adhesions, which may affect nerves or other pain sensitive tissues. Percutaneous lysis of epidural adhesions may also enable improved delivery of injected drugs to targeted painful structures. Epidural adhesiolysis with direct deposition of corticosteroids in the spinal canal can be achieved by a 3-dimensional view using a spinal or epidural endoscope. In a synthesis of the evidence the clinical use of percutaneous epidural adhesiolysis using a spring-guided catheter with or without hypertonic saline, Manchikanti et al reported their findings with short-term relief defined as less than 3 months and long-term relief as lasting longer than 3 months. Two randomized trials^{53, 55} were positive for both short and long-term relief. The effectiveness of spinal endoscopic adhesiolysis was summarized by an evidence synthesis by Manchikanti et al⁵ with review and comparison of 2 prospective^{56, 57} and 4 retrospective studies^{41, 53, 54, 55}. The effectiveness of spinal endoscopic adhesiolysis was further evaluated by reviewing 2 prospective^{56, 57} and 2 retrospective studies^{41, 58}. Short-term relief was defined as less than 6 months and long-term relief as more than 6 months. All studies showed support for short-term improvement, but none demonstrated any evidence to support long-term benefit.

Complications of adhesiolysis with spinal endoscopy may include dural puncture, spinal cord compression, catheter sheering, infection, injury from the endoscope, and administration of high volumes of fluid. Epidural infusion of high volumes of fluid, especially hypertonic saline, can potentially cause excessive epidural hydrostatic pressure, resulting in spinal cord compression, elevated interspinal or intracranial pressure, epidural hematoma, bleeding, increased intraocular pressures with resultant visual deficiencies including blindness, and dural rupture.⁵ Unintended subarachnoid or subdural puncture with injection of local anesthetic or hypertonic saline can also occur with resultant neural catastrophe.⁵ Hypertonic saline injection into the subarachnoid space has been reported to cause cardiac arrhythmia, myelopathy, and loss of sphincter control.^{5, 59} Arachnoiditis and sheering of the catheter with retention has also been reported with epidural adhesiolysis and hypertonicsaline.^{60, 61, 62} In summary, these procedures should only be performed under fluoroscopic control by well-trained, experienced spinal interventionalists.

A variety of intradiskal procedures have been developed because disk is often presumed causative for many painful spinal and radicular syndromes. A prospective randomized double-blind study of interdiskal injections into diskography-confirmed painful disks showed no statistically significant benefit between corticosteroids and local anesthetics.⁶³ Other therapies include chymopapain injections to achieve nucleolysis, percutaneous manual nucleotomy with nucleotome, thermal vaporization with laser, and percutaneous decompression with nucleotomy using Coblation technology (nucleoplasty). Intradiskal electrothermal therapy (IDET) is a minimally invasive technique in which the annulus is subjected to thermomodulation.⁵

The use of diagnostic diskography has been combined with therapeutic percutaneous intradiskal procedures in patients who demonstrate a concordant pain response. These include IDET, percutaneous laser disk decompression (PLDD), percutaneous radiofrequency annular neurolysis, and nucleoplasty.⁶⁴ These procedures are postulated to shrink collagen fibers and coagulate neural tissues, thereby alleviating nociception produced by mechanical loading upon a painful disk.

IDET is performed using radiographic placement of a 17-gauge introducer needle through a posterior annular wall into the nucleus of symptomatic disks as determined by diskography. A navigable catheter with a temperature-control, thermal-resistant coil is passed through the needle so that it curls along the posterior inner annulus. Catheter temperatures are slowly raised to 90°C, causing thermocoagulation of intradiskal and inner annular collagen, as well as associated nociceptors. Reduction in pain symptoms may result from denervation or shrinking and remodeling of the diskal structure, or both.⁶⁴

Karasek and Bogduk introduced a flexible electrode into a "discography symptomatic disc" with internal disk disruption.⁶⁵ The electrode was passed circumferentially along the inner annulus to heat and coagulate annular collagen and nociceptive nerve fibers. Of the 35 treated patients, 23% achieved complete pain relief and 60% improved. Improvements were sustained at 6 and 12 months. Seventeen patients comprising a parallel comparison group did not benefit from a physical rehabilitation program alone, except 1 patient who had dramatic pain reduction.⁶⁵ At a 2-year follow-up, 54% of the patients had achieved at least 50% relief with concomitant functional improvement.⁶⁶ A randomized double-blind study by Pauza et al was presented in 2002 and published in a review by Wetzel et al.^{67, 68} In fact, multiple prospective and retrospective studies examining the efficacy of this procedure have been published.^65</

Pathophysiology of Chronic Back Pain

AUTHOR AND EDITOR INFORMATION

INTRODUCTION

PATHOPHYSIOLOGY

TYPES OF PAIN

Neurophysiology of spinal pain and injury

EVOLUTIONARY MECHANISMS IN CHRONIC LBP

HISTORY, PHYSICAL EXAMINATION, AND CAUSES

History

Physical examination

Causes

DIAGNOSTIC STRATEGIES

MEDICAL CARE

Overview

Medical treatment

THERAPEUTIC SPINAL INTERVENTIONAL TECHNIQUES