Chapter 30. Failed Back Surgery Syndrome

Failed back surgery syndrome (FBSS) was defined by North and Campbell in 1991 as persistent or recurring low back pain, with or without sciatica, following one or more lumbar spine operations.1 Van Goethem and colleagues describe it as a syndrome characterized by intractable pain and various degrees of functional incapacitation, following spine surgery.2 Rowlingson uses the term failed back surgery syndrome for patients with chronic debilitating low back pain occurring in a patient after back surgery of a variety of types, such as discectomy, laminectomy, and lumbosacral fusion, that was unsuccessful in relieving the patient’s symptoms.3 Fiume and colleagues consider FBSS to be a severe, long-lasting, disabling, and relatively frequent (5%–10%) complication of lumbosacral spine surgery.4 Although disability and chronic pain are commonly seen in the lumbosacral region and the lower extremity, similar mechanisms, pathophysiology, diagnostic dilemmas, and management options can be seen and extrapolated to the cervicothoracic region and the upper extremity.

In 1934, Mixter and Barr demonstrated that a herniated disc could cause nerve root encroachment, ultimately producing back pain.5 In 1951, Barr determined that a patient might have persistent low back pain, sciatica, or both, despite surgical intervention.6 In 1979, Finneson and Cooper made a statement that, “No matter how severe or intractable the pain, it can always be made worse by surgery.”7 The cause of FBSS has been recognized to be multifactorial over the past decade.

Population studies indicate that of all patients with acute back and leg pain, only 1% to 2% actually suffer from disc herniation and require surgery.8 Nearly 300,000 spinal surgeries are performed each year in the United States.9 Approximately 85% of these procedures involve laminectomy and discectomy and 15% are spinal fusions.10 The success rate in most surgical series ranged from 80% to 98%.11–17 Nachemson17 has pointed out that the success rate drops in a dramatic fashion after the first operation. The success rate drops to about 30% after the second operation, 15% after the third, and 5% after the fourth. The difference in patient populations and evaluation criteria make it difficult to compare the various series published to date. Principally, the results are variable, because of the different procedures being grouped as one.

In a retrospective study by Dvorak and colleagues, 371 post–disc surgery patients were interviewed 4 to 17 years later by neurologists.18 As many as 70% had residual low back pain; 23% had severe, permanent low back pain; 45% had residual nerve root pain; 35% patients were undertreated; 14% received disability benefits; and 17.2% patients received repeat surgery. Fritch and colleagues conducted a retrospective study on 182 revisions with FBSS from 1965 to 1990, to identify the cause of failure of primary discectomy, the outcome of revisions, and factors that influenced these outcomes.19 The rate of revision surgery ranged from 5% to 33%. In 80% of patients, the results were satisfactory in short-term evaluation, decreasing to 22% in long-term follow-up (2 to 27 years). Fritch also concluded that laminectomy performed as primary surgery was a major factor leading to a higher rate of revisions. Furthermore, it can be deduced from various studies that a significant percentage of patients who undergo spinal surgery do not achieve what they and their surgeons agree to be a satisfactory result. FBSS is, therefore, a common clinical entity that represents a significant treatment challenge to physicians.

Different authors have classified the various causes of FBSS in multiple ways. Goupille classified it according to the clinical presentation, as follows.20

• I. No improvement

• • a. Faulty technique

  • b. Creation of a new source of impingement

  • c. Poor patient selection

• II. Recurrence after a period of improvement

• • a. Recurrent herniation

  • b. Other source of impingement

  • c. Postoperative “fibrosis”

• III. Persistent low back pain

• • a. Destabilization

  • b. Disc degeneration

  • c. Facet joint syndrome

  • d. Complications (discitis)

Van Goethem and colleagues subdivided causes of FBSS according to mechanisms, diagnosis, and surgical approach.2

• I. Mechanical

• • a. Recurrent or residual disc herniation

  • b. Spinal canal stenosis

  • c. Foraminal stenosis

• II. Inflammation

• • a. Epidural fibrosis

  • b. Arachnoiditis

  • c. Spondylodiscitis

• III. Diagnostic

• • Failure to correctly identify the structural source of pain

  • a. Conjoined nerve roots

  • b. Disc herniation at different level

  • c. Peripheral nerve lesion

• IV. Surgically induced

• • a. Dorsal ramus nerve lesions

  • b. Facet joint strain

  • c. Sacroiliac joint dysfunction

  • d. Erector trunci tendinitis

Poor Patient Selection

Irrespective of the classification, the commonest cause of failed back surgery syndrome is poor patient selection.1,21 In fact, despite a scoring system for predicting the outcome of lumbar disc excision, the postoperative results have not been very promising.7,12,17,19 In a multi-center, nationwide, prospective study of the outcome of first-time back surgery, it was determined that over 90% of patients improved when chosen for surgery by expert spinal surgeons. A small group of patients considered to be inappropriate candidates for surgery, as determined by these same experts, had surgery elsewhere. Only 10% improved after an operation performed outside the study, with a resultant worsening of symptoms in a significant portion of these patients.22

Waddell, in an assessment of the outcome of low back surgery, stated that it is important to remember that failure of first surgery with residual pain is not an indication for second surgery.23 He noted that there was a tendency of some surgeons to carry out a decompression procedure with or without discectomy, and then to follow that with fusion when the first procedure failed to relieve pain symptoms. The problem is frequently that the first procedure was ill chosen. In addition, psychological or environmental factors have a major impact on the presentation of low back pain. Indeed, psychological evaluation seems more predictive of the outcome of surgery than the diagnostic studies or surgical findings.

Despite a surgeon’s refusal to perform surgery, many patients will “doctor shop”—to pursue a lawsuit, attract sympathy and attention, avoid an unpleasant job or employer, or acquire financial compensation.24 Thus, inappropriate behavior by the surgeon as well as the patient can lead to FBSS.

Work-related injuries or patients who are sick-listed seem not to benefit from any modalities of treatment.25 Hansson and Hansson conducted a prospective 2-year cohort study in six countries, including the United States, involving 2,080 subjects who had been sick-listed for more than 90 days. Their objective was to find out the impact of medical interventions, including surgery, in patients with low back pain who are incapacitated at work. The authors concluded that almost none of the frequently practiced medical interventions for patients who are sick-listed because of low back pain had any positive effects on either the recorded health measures or work resumption.

Lastly, some patients have an unrealistic expectation of a complete return to full premorbid function with total absence of pain postoperatively; this expectation can result in some disappointment. It is imperative to explain to the patient that the recovery of damaged nerves is very unpredictable. The expected recovery is variable and may take from months to a year. Clinically, a short duration of symptoms has a much more favorable outcome.

Failure to Correct the Initial Pathology

The failure to decompress the spinal canal, to remove a disc entirely, to decompress a nerve root or roots through foraminotomy, or to miss sequestrated free disc fragments can result in minimal symptomatic improvement. Also, operating at the wrong level or not stabilizing unstable segments can contribute to FBSS. The failures of interventions to correct the primary problem have been confirmed by repeat imaging studies.26

Epidural Fibrosis

Epidural fibrosis is one of the major causes of FBSS.30 The reported incidence of epidural fibrosis ranges from 10% to 75%.28 The exact pathogenesis of epidural fibrosis is not well established. One possible mechanism accounting for fibrosis following surgery may be related to persistent cotton debris from sponges used during surgery, serving as a fibrogenic stimulus.29 Metal dust arising from poor-quality surgical tools (e.g., periosteal elevators) has also been identified as one of the potential causative factors leading to the development of epidural fibrosis.26 The scar tissue that forms at the surgical site does not seem to correlate with the extensiveness of the surgical procedure.29

Alternatively, arachnoepiduritis may develop, causing ischemia and atrophy of nerve roots. It has also been shown that there is no correlation between the amount of scar tissue and the severity of symptoms. Although there is no clear evidence that epidural fibrosis can cause symptoms, adhesions may bind the dural sac and nerve root sheaths to the walls of the spinal canal, stretching nerve roots. It seems that postoperative fibrosis is an entity usually associated with a distinctive pattern of pain. The pain-free interval lasts weeks to several months (approximately 6 weeks to 6 months). The pain usually sets in gradually; its distribution is not confined to that of the initially compromised nerve root, but extends to multiple nerve roots. It is commonly described as a burning sensation with occasional lancinations.20

Some important nonsurgical causes of epidural fibrosis are annular tear, hematoma, infection, and intrathecal contrast media. Regardless of the cause, the pathology involves collagenous fibers that encapsulate nerve tissues, resulting in lateral spinal stenosis, which could impair arterial tissue perfusion and venous return. This anoxic phenomenon is one explanation for pain at rest. Also, epidural fibrosis causing adhesions may restrict nerve root mobility, leading to increased incidence of lateral herniated disk symptoms with increased tension, decreased blood flow, or additional trauma. Scar tissue is generally found in three components of the epidural space. Dorsal epidural scar tissue is formed by resorption of surgical hematomas and may be involved in pain generation.29 In the ventral epidural space, dense scar tissue is formed by ventral defects in the disc, which may persist despite surgical treatment and continue to produce chronic low back pain/radiculopathy past the surgical healing phase.30 The lateral epidural space includes structures like the nerve roots and dorsal root ganglia, called the sleeves that exit the root canals, which are susceptible to lateral disc defects, facet overgrowth, and neuroforaminal stenosis.31 On the other hand, Pawl investigated post-laminectomy surgical patients with epidural fibrosis and found no clear relationship to pain except in cases with excessive scarring in the first year of surgery.32

Recurrent Disc Herniation

Another important cause of repeat symptoms in patients previously operated on for lumbar disc herniation is recurrence of disc herniation at either the same or an adjacent level.33 In nearly 20% of postoperative patients, there is asymptomatic disc herniation even with impingement of the nerve root. Epidural fibrosis that may coexist with a disc herniation makes it difficult to determine the pain generator.

A herniated disc can cause pain through physical or chemical mechanisms. Pressure-induced changes in nerve fibers are dependent on the degree of force. For instance, an applied pressure of 10 mm Hg decreases the nutrient transport to the nerve roots by 20% to 30%.34 Compression may also produce edema and lead to formation of intraneural fibrosis. Pressures as high as 200 mm Hg can produce a direct mechanical effect by deformation of nerve fibers, displacement of the nodes of Ranvier, or invagination of the paranodal myelin sheaths.35 Such high pressure can result in persistent neural injury, especially when compression occurs near the neuronal cell body. Thus, it has been noted that delaying decompression of acute disc herniation for longer than 6 months after the onset is more likely to result in persistent postoperative pain. Moreover, premorbid conditions, including age and coexisting neuropathy, may also influence the recovery of neural function after surgical decompression.20

Chemical irritation is another factor responsible for back pain with or without radiculopathy. Degenerative disc disease and tears of the annulus fibrosus may result in the leakage of various chemical irritants. Surgical disc samples from patients with disc herniation have been shown to contain extremely high levels of phospholipase A2.36 This enzyme liberates arachidonic, the principal mediator of inflammation and swelling along a nerve root, from cell membranes. In addition it has been theorized that material from the nucleus pulposus might act as a foreign protein and trigger an autoimmune reaction.34 Obviously, chemically induced irritation can occur in the absence of compression by a disc herniation.

Other Perioperative Surgical Complications

Acute unremitting postoperative pain, tenderness, and neurologic deficits are secondary to hematoma formation. Evacuation of the hematoma produces dramatic pain relief on most occasions. There is an associated low risk of injuries to the spinal cord and nerve roots during surgery. However, more commonly, peripheral nerves, dura, joints, and muscles can be damaged and serve as a source of persistent pain postoperatively.

The failure of fusion can occasionally lead to spinal instability. Pseudoarthrosis in the adjacent joints can occur after a generous laminectomy, or pars fracture. Transition syndrome can result from accelerated degenerative changes at levels adjacent to a spinal fusion.37 Flat back syndrome follows loss of the normal lordosis. This abnormality is frequently found after extensive fusion or instrumentation. Abnormal posturing is often seen, and a common complication is accelerated degenerative changes causing new pain.

Nonsurgical Causes of Failed Back Surgery Syndrome

Arachnoiditis

Inflammatory changes in the nerve roots and the surrounding arachnoid mater aecondary to surgical trauma, infection, bleeding, disc disease, or myelography can result in arachnoiditis. Arachnoiditis progresses in three stages. The first stage is radiculitis, characterized by inflammation of the pia-arachnoid mater, with hyperemia and swelling of the nerve roots. There is minimal fibroblast activation, but collagen strands begin to deposit between the nerve roots and pia-arachnoid mater. The second stage is arachnoiditis, characterized by increased fibroblast proliferation and collagen deposition. Although there is less swelling, the nerve roots adhere to each other and to the arachnoid mater. The final stage, adhesive arachnoiditis, marks the end point of the inflammatory process. There is dense collagen deposition and complete encapsulation of the nerve roots, leading to progressive atrophy and stretching.38 Not all patients with histologic changes develop pain. Neither does a similar degree of trauma consistently give rise to arachnoiditis. The fibrous changes following arachnoiditis may vary from involvement of one nerve root to various parts of the cauda equina.39

Clinically, affected individuals present with unremitting pain in the low back and leg, which increases with movement or positions that stretch lumbar nerve roots. There may be patchy neurologic deficits, including bladder dysfunction with variable degrees of motor and sensory changes. Magnetic resonance imaging (MRI) results of patients with lumbar arachnoiditis can be divided into three groups. Group one is characterized by a conglomeration of adherent roots residing centrally within the thecal sac. Group two is distinguished by nerve roots, adherent peripherally to the meninges and giving rise to an “empty-sac” appearance. Group three demonstrates a soft-tissue mass replacing the subarachnoid space.40

Spondylodiscitis

This condition is seen in less than 0.5% of patients with FBSS. Wound infection immediately after surgery may lead to recurrent or persistent atypical lumbar pain, with or without pain radiating down to the legs, or back muscle spasms. It usually is associated with low-grade fever, leukocytosis, and high erythrocyte sedimentation rate. Aspiration of the disc in the early stages can confirm the diagnosis.41 MRI findings in the later stages of the disease show central disc enhancement, posterior annulus enhancement, and enhancing edematous bone marrow of both adjacent vertebrae.2

Spinal Stenosis

Spinal stenosis is one of the most common causes of FBSS.2 Both central and lateral canal stenosis can coexist. Lateral spinal stenosis is seen commonly as a result of a gradual loss in disc height.43 The symptoms of pain and neurologic deficit can be episodic and variable, as a result of the somewhat random ischemia of the vascular supply to the spinal nerves. The pain is usually described as deep and aching in the low back area, buttock, and thigh, with radicular pain down the leg, as well. Standing and walking are particularly painful. Most of the pain is dramatically relieved by forward flexion. Walking distance progressively diminishes over time. FBSS occurs if the predominant pain mechanisms are lateral or central canal stenosis.

Facet Joint Arthropathy

A previous discectomy, laminectomy, or spinal fusion may cause adjacent facet joints to undergo accelerated degenerative arthritic changes. The pain pattern is mostly axial, seldom radiating below the knee, aggravated by extension and prolonged sitting. Diagnostic block with local anesthetics is one of the easy ways to differentiate facet disease from other causes of low back pain.2,20

Pseudomeningocele

A pseudomeningocele is a cerebrospinal fluid leakage under pressure, leading to a cerebrospinal fluid–filled cavity. The patient may complain of low back pain. The diagnosis is usually confirmed by MRI, and surgical repair is the recommended treatment.34 This condition is a rare cause of FBSS.

Tumors

A missed diagnosis of spinal cord or cauda equine tumor (e.g., schwannoma or ependymoma) is, rarely, a cause of FBSS.

Central Sensitization

Intractable back pain and sciatica may persist despite appropriate surgery to relieve the symptoms caused by a compressed nerve root. Chronic compression of nerve fibers is a source of ectopic discharges. Wall and Devor have shown that normal resting spontaneous discharges in dorsal root ganglia are markedly increased as a result of peripheral neuroma formation.47,48 Peripheral nerve dysfunction can establish permanent alterations in cell organization and function of the dorsal horn, leading to hypersensitivity. The exaggerated persistence of pain stems from sensitization of the periphery as well as spinal cord mechanisms.

Early surgery is considered as too aggressive, whereas delayed intervention, after central sensitization has taken place, is too late. Hence, there is no clear understanding as to the optimal time for surgery in many cases.

Diagnosis of FBSS rests mostly on clinical evaluation and imaging studies. Ancillary modes include electromyography, nerve conduction studies, diagnostic blocks, provocative discography, and epiduroscopy. Because the leading cause of FBSS is inappropriate patient selection, a multidisciplinary diagnostic evaluation carried out in a tertiary care center is often warranted. Graham Smith emphasizes the importance of patient selection, particularly after surgery, by admitting patients for 5 days of in-depth assessment.46 This gives a chance for the different teams, including rehabilitation, to assess the patient’s physical, emotional, psychological, and social status in depth. It allows full investigation to make certain that no surgically remediable lesions remain and to determine rehabilitation potential. The role of the psychologist in exploring the patient’s attitude to disease and his or her ability to cope with disability is of grave importance in this respect. However, in the current environment of cost containment, a coordinated, multidisciplinary evaluation of patients with disabling, chronic pain is more commonly performed on an outpatient basis.

Clinical History

The medical and surgical history is of paramount importance for evaluating the patient’s chronic, disabling low back pain. A thorough review of the patient’s self-filed questionnaire and previous medical charts, including previous treatments and imaging studies, can be very helpful in narrowing the differential diagnosis of low back pain. Of foremost importance are the pain characteristics, neurologic symptoms, functional limitations, and comorbid medical and psychological conditions. The outcome of previous treatments will strongly influence the management plan. The identification of an acute neurologic deficit, maladaptive behavioral issues, secondary gain, pending lawsuits, or a history of substance abuse is pivotal in determining treatment options.

Carragee delineated some factors responsible for either favorable or poor outcome influencing surgery.47 Factors influencing poor outcome are (1) concurrent medical illness, (2) involvement with workers’ compensation, and (3) female gender. Factors resulting in favorable outcome are (1) duration of symptoms less than 6 months, (2) absence of litigation, and (3) younger age. Smoking is considered to be a major risk factor for failure of bony union; fusions may be postponed until tobacco abstinence is achieved. Pearce noted some clinical features distinguishing nonlitigants from litigants.48 Interestingly, when litigation is involved, leg radiation if present is often in the dermatomal distribution of L2 through L4, diffuse, or front of thigh not below the knee (nonanatomic), as opposed to radiation in patients not involved in litigation, which is in the L5-S1 distribution, back and side of buttock, and thigh to foot (anatomic distribution).

Detailed pain assessment is the key factor in the evaluation of patients with low back pain. The type of pain, location, radiation, duration, frequency, severity, association with movement and activity, as well as aggravating and relieving factors are integral to the assessment. Some of the easily distinguishable pain characteristics may help identify the anatomic location. For example pain resulting from spinal stenosis is relieved by forward flexion and aggravated by extension. Axial, deep, low back pain that is worse on extension could be a result of facet arthropathy, spondylolisthesis, or disc pathology. Return of pain within months to years is a result of epidural fibrosis, arachnoiditis, facet joint disease, segmental instability, and recurrent disc herniation.

Physical Examination

As with the clinical history of various pain complaints, review of previous examination findings is crucial in the process of diagnosis. The examination can be divided into components of musculoskeletal, neurologic, functional, and tension signs.

When examining the musculoskeletal system, attention must be given to the curvature of the spine and the limitations on range of motion. For example, pain on repeated lumbar flexion is probably a result of hypermobility problem.49

Provocative tests can be secondary to stretch of the spinal cord or sciatic nerve, or tests of increased intrathecal pressure. Tests to stretch the spinal cord or the sciatic nerve are otherwise known as tension tests. The straight-leg raise (SLR) is the most widely used. The pain must typically traverse below the knee to be called positive. The test is only considered provocative between 30 and 60 degrees. It cannot be considered as specific at less than 30 degrees because there has not been enough tension applied to cause movement of the lumbar plexus, which in turn creates movement of the neural structures through the foramina. At greater than 60 degrees, the nerve roots move ventrally and laterally to the pedicle. This causes the test to lose sensitivity.50 Cross-leg SLR test is considered to be more specific for radicular pain.

Reverse SLR or femoral stretch is specific for L2–3 radiculopathies. With respect to FBSS, the SLR should be interpreted in light of the slump test. The slump test should be performed with proximal and distal initiation. With proximal initiation, allow the patient to sit with legs dangling from the examination table. Then initiate vertical pretension with cervical flexion, thoracic flexion, and lumbar flexion. Follow this by dorsiflexion of the foot and extension of the knee to create slight distal motion and increase tension. A positive result is suggestive of a tension problem, such as a lumbar disc impinging on the nerve root. If the cervical spine is then extended, the tension decreases and there is decreased distal motion. If there is a tension problem, the pain decreases when the tension is diminished. On the other hand, a mobility problem, such as entrapment of the plexus with scar tissue results in increased pain on cervical extension. The results of the slump test should be compared with both distal and proximal initiation, and then with the results of the SLR.51

A thorough neurologic examination is compulsory. The “red flag” signs of cauda equina syndrome should be excluded. Any new-onset neurologic sign demands serious consideration. At times, residual neurologic effects, such as foot drop or diminished deep tendon reflexes, may linger despite surgical decompression.52

Functional tests, such as the flip-differential SLR test and the Hoover test, must also be performed to identify nonorganic signs. Waddell designated five cardinal signs of nonorganic low back pain53; namely, (1) superficial nonanatomic tenderness, (2) stimulated pain on axial loading, (3) distraction test (otherwise called the flip/differential SLR test), (4) regional or light skin pinch with a wide area of pain or weakness, or both, and sensory loss that is divergent from accepted neuroanatomy, and (5) overreaction with disproportionate verbalization and facial expression. On clinical examination any one of these signs is sufficient to establish an exaggeration of a physiologic state; thus, Waddell’s insistence on three factors is not generally necessary. However, one should be cautious in labeling illness behavior, because some functional findings can result from organic problems. For example, patchy, nonanatomic sensory loss is common with arachnoiditis. Allodynia and hyperalgesia secondary to central sensitization can present as “superficial nonanatomic tenderness.” Therefore, physical findings should be considered in the context of the whole clinical picture, not in isolation.

Diagnostic Studies

Imaging Studies

Interpretation of imaging studies is more difficult in patients with FBSS because they may have superimposed anatomic abnormalities produced by their primary surgery and may have had instrumentation that compromises the quality of certain imaging modalities.52 Recent advances in diagnostic capabilities have greatly facilitated the possibility of diagnosing the cause of surgical failure. Van Goethem and colleagues, in a review of imaging findings in patients with FBSS, stated that MRI should be the first imaging procedure because many of the most common and important imaging findings in these patients are seen best on MRI.2 They also commented that it is a misjudgment to first perform computerized tomography (CT). A repeat examination with MRI will usually follow, to differentiate intraspinal soft tissue masses. Furthermore, MRI allows a more accurate assessment of foraminal diameter and visualization of the foraminal contents. In fact, MRI with gadolinium is considered the gold standard modality for consistently diagnosing enhancing scar tissue and recurrent disc herniations or retained disc fragments. The enhancement remains constant regardless of the time since surgery. Scar tissue can have a mass effect and be contiguous with the disc space, demonstrating there remains room for error.49

However, Coskun and colleagues, like many others, have found no relationship between the severity of epidural fibrosis and pain following surgery.27 MRI also avoids the potential morbidity of subarachnoid contrast studies, which have been implicated in the etiology of arachnoiditis. MRI has limitations in defining bony anatomy, which often can be better visualized by plain x-ray and CT. These films are useful for a hypermobility problem. A plain CT may demonstrate the nature of a previous surgery or a pars fracture that was not detected on plain films. CT with myelography has approximately 67% to 100% accuracy in patients with multiple operations to allow the differentiation of scar tissue from disc disease.54 A detailed review of imaging studies is beyond the preview of this chapter.

Electromyography

This method has a reported success rate ranging from 20% to 90% for determining nerve root compression. The specificity is low in patients who have had multiple surgeries because of neurologic changes that are commonly observed postoperatively. At this late stage in patient evaluation, the original lumbar pathology, surgical trauma, or ongoing compression can all yield positive electromyographic results, thereby making a negative study more useful.55

Nerve Conduction Studies

Nerve conduction studies are useful in correlating clinical symptoms with noncompressed nerve roots (from MRI). Nygaard and colleagues used quantitative sensory testing perioperatively and 12 months after microdiscectomy.56 They found significant improvement in small, unmyelinated fibers, mostly in the ipsilateral neighboring nerve root, concluding that the observed recovery of function in the adjacent noncompressed nerve roots after successful surgical decompression in monoradiculopathy may result from reduced production of proinflammatory mediators when the disc herniation is removed. Scelsa and colleagues studied the electrophysiologic correlates of weakness in L5-S1 radiculopathy and concluded that low peroneal and tibial compound muscle action potential amplitudes may serve as surrogate measures for segmental weakness of functionally relevant muscles.55 Electrophysiologic measurements taken in patients with lumbosacral radiculopathy are useful in estimating the degree of physiologic and anatomic injury.

Provocative Diskography

Provocative diskography is performed to determine the effect of mechanical loading of individual discs in patients with unremitting, chronic, discogenic low back pain.57 Elicitation of the patient’s characteristic pain is considered a positive sign. Discography is mostly performed as a prelude to either (IDET) intradiscal electrothermal annuloplasty or fusion surgery. Only strongly positive and definite negative results are considered reliable. Manometry is of questionable benefit. It is recommended that an adjacent disc to the clinically correlated level of pain should also be tested as a control. Inappropriate responses most likely suggest illness behavior or a more diffuse pain syndrome.

Epiduroscopy

This diagnostic and therapeutic procedure is used for the visualization of fibrosis, neovascularization, granulation tissue, and entrapment of the nerve roots.58 The epiduroscope is passed through the sacral hiatus to visualize the lumbosacral epidural space. Then a stimulating catheter may be placed for neurophysiologic mapping. It not only is considered the diagnostic method of choice for epidural fibrosis, but also is considered effective in the treatment of radiculitis by mobilization of the adhesions around an entrapped and irritated nerve root (neuroplasty). Corticosteroids in a depot form may be injected, together with local anesthetic. It is not clear whether the therapeutic effects are a result of the accurate placement of medication or the irrigation of the inflammatory mediators.59

Diagnostic Blocks

Hildebrandt justifies the use of diagnostic nerve blocks, especially in nonspecific spinal pain with poor radiologic correlation.60 Examples of such blocks include facet joint, sacroiliac joint, and selective nerve roots. In a review, Hildebrandt states that 10% to 15% of axial pain is secondary to a facet joint problem, 15% to 20% is from the sacroiliac joint, and 40% is of discogenic origin. The advantages of blocks are as follows: (1) they allow a peripheral pain pathway to be localized, (2) selectively blocking the specific nerves teases out that faction of the pain generator, and (3) relief of pain after local anesthetic block is attributable solely to block of the target afferent neural pathway. Facet joint pain is commonly associated with other causes of FBSS. Depending on the response to a block, permanent ablative procedures can be considered. Selective nerve root blocks are most often considered when low back pain is associated with monoradiculopathy.

In addition, temporarily abolishing the patient’s pain by blocking neuronal activity does not necessarily equate to the primary site of pathology. For instance, a patient with pain and allodynia may experience relief of pain by nerve root block, which temporarily eliminates incoming messages from the painful area. Local anesthetic can spread centrally, as well, giving the impression that a peripheral block is helpful. But if the major pathophysiologic abnormality is within the dorsal horn of the spinal cord, an intervention directed at the primary afferent axons (e.g., foraminotomy or neurectomy) will have limited benefit.52

As in all disease states, prevention is the most important consideration in reducing the incidence of FBSS. Given the common refractory nature of FBSS, a therapeutic plan must be set up with more realistic patient expectations. It is also important to explain to patients that, as with any other tissues in the body, neural tissue recovery may take months to years to recover. Axonal regeneration can be associated with persistence or worsening of pain. Rarely, patients may have persistence of their radicular pain for months after surgery, with total or near total resolution of pain once reinnervation of the distal target is completed.52 This recovery-associated pain and paresthesia must be distinguished from FBSS pain. Men who are over 40 years of age, overweight, obese, and heavy smokers (i.e., more than 20 cigarettes/day for more than 20 years) are at increased risk of chronic low back pain.61,62

The impact of smoking on the microvascular circulation of peripheral nerves and the vascular network surrounding the intervertebral disks should not be underestimated.63 Smoking increases the serum proteolytic activity, which increases the intervertebral disk degeneration and also weakens the spinal ligaments, causing spinal instability.64 Central obesity with associated truncal and lower extremity muscle loss, seen in women over 45 years of age, is considered a risk factor for chronic low back pain.65 Biologic risk factors, such as osteoarthritis, rheumatoid arthritis, gout, and other collagen diseases, need to be identified. Their impact on surgery, healing, and long-term outcome should be explained to the patient, because it does not alter the progression of the underlying degenerative process. Repeat surgery is more commonly seen in patients with identified biologic risk factors.

Inappropriate or premature patient selection, in this age of advanced diagnostic imaging and other modalities of evaluation, is less likely. Patients with unrealistic expectations must be carefully evaluated prior to any intervention. Unless there is a clear surgical indication, or conservative management has failed, more time should be allowed for surgical decision making, especially with respect to re-exploration. Pearce stated that “there is now good evidence that prolonged, skilled conservative care will sometimes prevent the failed back that so often complicates ill-judged invasive procedures.”48

Because only 1% to 2% of patients with back pain undergo spinal surgery, the therapeutic approach should begin with conservative management and gradually proceed from less to more invasive options. Goupille emphasized that the decision to perform surgery should be made individually, based on factors such as: (1) clinical correlation of the patients symptomatology with the imaging findings, (2) presence of radicular pain more than axial pain, (3) absence of social or psychological deterrent factors, and (4) patient willingness for surgery.20 Moreover, it is also important to consider a one-time surgical approach using advanced surgical techniques (e.g., using pedicle screws, plates, interbody cages) to reduce failure of fusion. Pulsed electromagnetic field stimulation has been used for bony bridging in lumbar spinal fusion, to yield a better outcome.66 Preservation of ligamentum flavum, together with other epidural structures, and limited removal of lamina are important components in preventing epidural fibrosis.67

FBSS is a multifactorial, complex problem that offers a challenging management task for the clinician practicing pain medicine. The literature offers little evidence to substantiate either a conservative or an interventional approach. Many pain centers offer multiple permutations and combinations of pain management strategies to improve quality-of-life measures. As with many complex chronic pain disorders, the primary treatments are frequently combined with physical therapy, occupational therapy, psychotherapy, and mind-body therapy to achieve functional restoration. The minimum criteria defining treatment success is a 50% reduction in pain evaluations. Other important goals include the resumption of employment and activities of daily living, which serve to further assess the response to a therapeutic plan.

Role of Psychologist

Given that the foremost reason for FBSS is poor patient selection, a clinical psychologist can identify confounding behavioral problems that will continue to foster poor responses to interventional and invasive therapies. Psychologists identify how and to what degree depression, anxiety, phobias, maladaptive behaviors, personality disorders, and temperament affect an individual’s pain syndrome. The magnitude of any of these psychological conditions can influence the decision to proceed with any additional invasive options. Psychological evaluations are routinely obtained for procedures such as spinal cord stimulation, continuous drug administration systems, and intradiskal therapies. Patients are also assessed to determine appropriateness for psychotherapy, biofeedback, relaxation techniques, psychopharmacology, and mind-body chronic pain programs.

Another key role of the psychologist evaluating chronic pain syndromes is to investigate past or potential future inappropriate use of opioid medications. A past history of substance abuse is thoroughly explored during the evaluation. The psychologist then indicates how closely a patient must be monitored for abuse of opioid medication, if the decision is made to place that individual on chronic narcotic medication. Indeed many patients with FBSS who undergo evaluation have been taking opioids for years. Many of these individuals go on to develop tolerance and dependence but, interestingly, the majority do not develop behavioral signs of addiction.68

Role of Physiotherapy

The structural asymmetries, deconditioning, ongoing pain, and attitude of the patient with FBSS represent a formidable rehabilitative challenge. Diminution of the pain behavior that frequently accompanies FBSS is more of a focus for the therapist than actual pain relief.69 The goals of physical therapy are to improve range of motion, enhance strength, increase activities of daily living, improve ergonomics, and prevent or decrease musculoskeletal injury by increasing physical environment awareness. There is emphasis on daily exercise and stretching routines. In FBSS, there is loss of normal lumbar lordosis, chronic muscle spasm of paravertebral muscles, and tightening and shortening of gluteal, hamstring, and calf muscle groups.70,71 The rehabilitative process of FBSS is slow and challenging, requiring continuous education and reinforcement.

Pharmacologic Management

A pharmacologic approach remains a cornerstone in any pain management program. Patients who have responded favorably to one surgery in the past are more likely to derive benefit from medical management.52 But there exists no one medication or class of medications that has demonstrated universal efficacy.

Nonsteroidal Anti-Inflammatory Drugs (NSAIDs)

NSAIDs remain as the first line of medical treatment against the pain associated with musculoskeletal disorders. Long-term use is much more likely than short-term or sporadic employment to cause adverse effects such as gastrointestinal toxicity, renal dysfunction, platelet inhibition, water and sodium retention, and allergic reactions. Selective inhibition of the cyclooxygenase (COX-2) enzyme may reduce some of the gastrointestinal complications and eliminate platelet dysfunction.72

Recently, Mukherjee and colleagues determined that the risk of cardiovascular events is higher with COX-2 inhibitors compared with placebo or traditional NSAIDs, such as naproxen.73 Regardless of the agent chosen, careful monitoring is necessary. Chronic use of high-dose NSAIDs is common among FBSS patients. Individuals with FBSS may likely have elements of chronic inflammation generating a component of their overall pain and, thus, an NSAID may be of value. COX-2 inhibitors offer an option in patients who do not tolerate or have contraindications to nonspecific NSAIDs.

Systemic Opioids

Chronic administration of oral opioids has gained acceptance for the management of long-term nonmalignant pain. Many patients with FBSS are likely to be using pain medications, including narcotic analgesics. The use of chronic opioids does not exclude the possibility of future interventions. In fact, appropriate use of systemic opioids may prevent inappropriate future procedures that can lead to further debilitation. There are some patients who remain stable on chronic opioid therapy over the long term.

In a careful review of the literature, Brown and colleagues formulated recommendations for implementing chronic opioid analgesic therapy (COAT) in patients with chronic low back pain.74 An analysis of 566 case series reports showed that COAT appears to be safe and effective for many patients suffering from recalcitrant, chronic low back pain. Preexisting substance abuse, personality disorders, certain medical conditions, and occupational factors are identified as relative contraindications. Further work by Jamieson and colleagues in a randomized, prospective, long-term, repeated-dose study that compared either sustained-release morphine or short-acting oxycodone with naproxen found favorable outcomes for mean pain scores, anxiety, depression, and irritability in patients treated with either opioid analgesic in comparison with naproxen.75 Stable plasma levels of narcotic analgesics improve quality of life, and cause fewer side effects with no cognitive deterioration, when given on a time-rather than pain-contingent basis.76 Thus, there is now a consensus view that in carefully selected patients with nonmalignant pain, including FBSS, pain and function can be improved with COAT.77

There are, however, many patients with FBSS who are refractory to opioids or cease to respond. Incremental dose escalation may lead to copious amounts of narcotics without further benefit. Identification of tolerance, after excluding a new pain disturbance or progression of disease, can be managed by substituting an alternative narcotic or considering additional pain management strategies.78 The lack of response or intolerable side effects to one narcotic analgesic does not preclude trying several other opioids. Physical dependence is a normal phenomenon in any patient who uses chronic opioids, including patients with FBSS. Many patients choose to avoid using narcotics on a daily basis to avoid addiction or dependence.79 Actual addiction, or the inappropriate preoccupation with narcotic analgesics induced iatrogenically, has been known to be infrequent since the 1980s.80 Addiction rates currently estimated in North America indicate that between 3% and 16% of chronic pain patients suffer from an addictive disorder that parallels the life-time prevalence of addictive disease in the general population.81 Screening patients carefully and instituting vigilant monitoring further reduces the probability of opioid abuse or misuse.82 Nevertheless, published patient surveys and case reports support the efficacy and long-term safety of opioid analgesics in selected patients with chronic, nonmalignant, and nociceptive more than neuropathic pain.83

Adjuvant Analgesic Agents

FBSS often has elements of nociceptive pain as well as neuropathic pain, which is commonly refractory to NSAIDs and opioids. Anticonvulsant and tricyclic antidepressant medications have been demonstrated to have efficacy against neuropathic pain syndromes, such as painful diabetic neuropathy and postherpetic neuralgia. These classes of medications have not been formally studied in the context of FBSS; however, they have an accepted role in the treatment of neuropathic pain, and, therefore, enhance analgesia in FBSS.

Other categories of medications include muscle relaxants and α2 agonists. Muscle relaxants such as carisoprodol, cyclobenzaprine, and metaxalone have been used in patients with persistent muscle spasm with limited success. Two unique agents in the muscle relaxant class are tizanidine (α2 agonist properties) and baclofen (γ-aminobutyric acid agonist, type B). Baclofen is used as a muscle relaxant, but more appropriately it is an antispastic agent that has shown benefit in patients suffering from neuropathic pain or spasticity. Tizanidine exerts a central α2 adrenergic agonist effect that has successfully relieved muscle spasm and neuropathic pain.84 The efficacy of any of these agents in FBSS is uncertain, and clinical experience suggests that as monotherapy, it would be modest.

Epidural Corticosteroid Injections

Despite the lack of randomized, well-controlled studies, the placement of corticosteroids into the epidural space as a treatment for FBSS remains the most common therapeutic modality. The rationale is to reduce any residual inflammation via phospholipase A2 inhibition. Reduced inflammation and direct C-fiber inhibition by corticosteroids reduces the sensitivity of nociceptors to mechanical stimuli and tensile force.85 Factors associated with a favorable response to epidural corticosteroids include advanced educational background, lumbar radiculopathy, and pain duration of less than 6 months.86

On the other hand, factors that serve to limit the success of epidural corticosteroids are alteration of lumbar anatomy following surgery, limitation in epidural needle placement as a direct result of surgical scarring and altered anatomy, and epidural fibrosis, which interferes with epidural diffusion and penetration to the appropriate targets. A review of 12 randomized controlled trials by Cousins and colleagues found benefit in 6 of the 12 studies.87 The negative studies essentially found short-term relief. The use of fluoroscopy to improve needle placement and therapeutic response is debated in the literature. But in clinical practice, it has proven invaluable in patients following complex fusions and laminectomy.88 The incidence of wet tap in patients after laminectomy or fusion can approach 20%, and this is likely reduced through the use of fluoroscopy.3 Moreover, the injection of radiocontrast into the epidural space confirms placement and spread to the sites of probable pathology.

The use of transforaminal blocks (or nerve root blocks) allows for placement of epidural corticosteroids on the affected nerve root without the necessity of identifying the epidural space. This approach avoids the risk of postdural puncture headache and improves spread of corticosteroid to the affected nerve root by circumventing epidural fibrosis. Devulder and colleagues performed transforaminal blocks for mononeuropathy in 60 patients with FBSS in a randomized, single-blind study.89 The aim was to compare short- and long-term relief with bupivacaine plus hyaluronidase, and with or without corticosteroid. At 6 months, only 27% of patients had greater than 50% reduction of their initial pain scores, and no benefit was seen from adding corticosteroid to hyaluronidase and bupivacaine. A series of risk factors for failure of this procedure were identified, which include lower level of education, smoking, lack of employment during the initiation of treatment, ongoing litigation, constant nonradicular pain patterns, duration of pain greater than 1 year, pain that is not influenced by activity, pain behavior, and neuropathic features indicative of allodynia and central sensitization. In conclusion, epidural injection of corticosteroids and hyaluronidase appears to be of benefit in selected patient with FBSS. Injection of local anesthetic into the epidural space to silence sympathetic nervous system involvement of nociceptors is without any meaningful long-term efficacy.

Epiduroscopy

The epiduroscope can be used for visualization of epidural fibrosis, neovascularization, granulation tissue, and fibrotic entrapment of nerve roots.58 Using complete aseptic measures, the scope is introduced at the level of S3. The catheter is fed through the ventral surface until it reaches the L4–5 interspace, which is the usual level of pathology. The catheter can then be maneuvered closer to the sites of epidural scarring. An epidurogram is then performed to detect a so-called filling defect, indicating an area of scarring and possible nerve entrapment. Alternatively, a stimulating catheter may be placed for neurophysiologic mapping. Thereafter, hyaluronidase is injected to hydrolyze excess ground substance in epidural scar tissue, thus allowing diffusion of local anesthetic and corticosteroid to target areas. Racz and colleagues found this technique of epidural neuroplasty effective in 49% of patients with FBSS.59

Facet Joint Injections

Degenerative disease of the facet (zygapophyseal) joints is an underappreciated cause of low back pain with radiation to the knee level in FBSS. The clinical examination is, however, unreliable and diagnostic or therapeutic facet injections can offer more accurate insight. A subjective report of greater than 50% pain relief is regarded as a positive response, in that the pain is derived from the facet joints.

As would be expected, many FBSS patients do not sustain benefit from intra-articular facet injections. The response to medial branch blocks is then evaluated prior to pursuing radiofrequency lesioning of the facet joints. After diagnostic blocks and radiofrequency lesioning, investigators found 50% pain relief in 50% of the patients 3-years follow-up.90,91

Facet denervation often alleviates a component of the FBSS, and, therefore, other generators of pain should be considered. A facet problem in FBSS is often accompanied by compensatory muscle spasm. Trigger point injections and physical therapy may be added to offer a more complete rehabilitative approach. Botulinum toxin A has been applied in cases of spasm of the quadratus lumborum, psoas major, piriformis, and paravertebral muscles.92,93 Chemodenervation is often limited to refractory cases, and its efficacy in FBSS is unknown.

Intradiskal Electrothermal Therapy (IDET)

Patients suffering from FBSS may have internal disruption of a disk as a generator of their pain. The disk may have formerly been operated on, or it may represent new pathology at adjacent levels. The outcomes for IDET procedures on patients with FBSS are unknown. Studies evaluating IDET for patients with diskogenic pain as suggested by history, physical examination, and diskography have generated mixed results.94–96 Improved patient selection, psychologic testing, and better technique will probably be combined to offer a select group of patients a nonoperative solution to diskogenic pain.

Spinal Cord Stimulation

Neuromodulation has gained increase popularity given the low success rate of ablative procedures. Dorsal root ganglionectomy, whether performed percutaneously with radiofrequency or surgically, has produced only a 15% success rate at 5 years in patients with FBSS.52 There has been increased recognition of the importance of avoiding deafferentation in a chronic pain syndrome.

Spinal cord stimulation has been used for FBSS since 1967. The pioneering work of Shealy and colleagues has created the basis for interventional pain management for a host of neuropathic pain syndromes.97 The gate theory of Melzack and Wall and central inhibitory mechanisms evoked by Aβ stimulation are commonly accepted mechanisms of analgesia brought on by spinal cord stimulation.98

In the United States, FBSS is the most common indication for spinal cord stimulation. Long-term follow-up studies have shown a success rate of 50% to 75% for FBSS at 5 years.99 North has demonstrated that spinal cord stimulation is more effective than either dorsal root ganglionectomy or reoperation.100 Dual lead placement in patients with FBSS allows for more programming options to cover some axial pain, as well as persistent coverage of the extremities even with some lead migration.100 Indeed, system complications such as lead migration, lead fracture, infection, and battery failure are the most common problems.101

Spinal cord stimulation can lower the medical costs of FBSS over time, by reducing office and emergency department visits.102 Spinal cord stimulation pays for itself within 2.1 years. Medication requirements tend to stabilize or decrease, activities of daily living increase, and some patients are able to return to meaningful employment.103 Overall, spinal cord stimulation is a viable option for patients who fail to obtain sufficient pain control with less-invasive modalities. Future options may include deep brain stimulation, especially for the nociceptive component of FBSS. Nearly 62% of patients with FBSS derived meaningful analgesia from deep brain stimulation, but only half of these may go on to have sustained benefit over the long term.104

Central Neuraxial Infusions

Over the past decade, there has been a significant surge in the use of implantable pump technology for nonmalignant pain, particularly intrathecal opiates for FBSS.105 Drug administration systems delivering intrathecal narcotics over time can result in tolerance, hyperalgesia, and hormonal disturbances.106 Despite these problems, long-term intrathecal infusion of opiates can reduce pain scores by 25% to 50% after 2 years of therapy, with improved coping skills in FBSS.107 Axial pain often respond less favorably than radicular or neuropathic pain to intrathecal infusions. Like spinal cord stimulation, implantable pumps appear to be cost-effective over a period of at least 12 to 22 months.108 Pump refills and the cost of the technology are far exceeded by frequent emergency department and office visits, as are repeat diagnostic tests, and multiple medication trials.

Reoperation

The rate of success after reoperation varies from 12% to 100%.1 Reoperation should only be considered if there is a clear anatomic reason for worsening pain and conservative measures have failed. The dogma to reoperate only because alternative methods have failed will simply lead to more morbidity. Epidural fibrosis is always a risk, even in carefully conducted surgery. Reoperation often requires careful evaluation to avoid the enhancement of a syndrome that is already difficult to treat.

In all studies of back pain, 10% to 15% of patients account for 80% to 90% of the total health care consumption for spine disorders, and the 2% of patients who undergo spinal surgery are the most expensive group.109 At least 15% of patients who have spinal surgery develop FBSS. Interestingly, spinal decompression for other reasons, such as tumor, often yields impressive clinical results. The lack of such satisfactory results in nonmalignant pain represents the magnitude of the complexity of back pain in a significant number of patients. In addition to structural abnormalities, psychological factors, social comorbidities, and complex peripheral and central processing of nociceptive information may account for low back pain. The multidimensional nature of this complex clinical problem and the lack of prospective randomized, controlled data have obscured outcome assessment of pharmacologic, surgical, and other interventions in FBSS.