Although commonly used as a diagnostic term, failed back syndrome (FBS) is a misnomer. The term “syndrome” should not be applied to patients with a “failed back” because it gives the perception that patients with FBS have a group of symptoms that commonly occur together. Taken as such, the danger exists that a clinician may disregard important signs and symptoms that will lead to the proper treatment of a patient. Fortunately, a set of diagnostic principles can be used to clarify these issues and, from a practical standpoint, be used methodically to achieve a more appropriate diagnosis.
I. FAILED BACK SYNDROME
The FBS is a clinical condition experienced by patients who undergo a surgical procedure, typically in the lumbosacral region, with unsatisfactory results. Back pain is the second most common reason, behind asthma, for patients to seek medical help. It has been estimated that 300,000 laminectomies were performed last year. With the advent of modern instrumentation systems, an increasing number of lumbar fusions are being performed each year. Unfortunately, not every operation is successful, the success rate ranges from 50% to nearly 100% depending on the indication. Consequently, the prevalence of FBS is quite high.
Categories of FBS include the following:
A. Failure to improve due to misdiagnosis. By definition, FBS implies previous surgery. Therefore the first priority in the evaluation of these patients is to understand the indication for the original operation. It is often helpful to ask the patient to compare current symptoms with those experienced prior to surgery in terms of location, frequency, and intensity. A patient who fails to improve at all following surgery is more likely to have been misdiagnosed than a patient who improves for a period of time. If the original indication for surgery is suspect, it is extremely unlikely that further surgical intervention will be helpful.
B. Failure to improve due to improper treatment. Patients who do not improve or worsen immediately after surgery may have suffered a technical error during surgery. These errors include inadequate decompression, wrong level surgery, or nerve root injury. Frank instability of the operated level may also worsen if an unstable spinal motion segment such as a mobile spondyllolisthesis is not stabilized or fused at the time of decompression.
C. Recurrent pathology. A patient who experiences identical recurrent symptoms after a postoperative period of significant improvement will likely harbor recurrent pathology. Disc herniations, for example, carry a lifelong risk of recurrence because the majority of the anatomical intervertebral disc remains after a discectomy and the annular tear that permits the herniation never completely heals. An infection may also be present, frequently becoming apparent within the first 4 weeks of surgery. Infections may cause recurrent symptoms as well as the new onset of significant back pain. These infections can be occult, and imaging studies should be performed to determine if endplate erosion or a fluid collection is present.
D. Progression of pathologic changes at unoperated sites. Surgical procedures of the lumbar spine are commonly performed for degenerative diseases. Spondylosis, or bony overgrowth of the facet joints and intervertebral endplates, in association with soft tissue ligamentous hypertrophy can cause significant stenosis. Spondylolisthesis, or malalignment of the spine, can progress after decompression alone, causing recurrent nerve root entrapment. Additionally, progression of degenerative changes at a level adjacent to a lumbar fusion may occur, resulting in stenosis and/or spondylolisthesis.
II. SIGNS AND SYMPTOMS THAT WILL AID IN THE ASSESSMENT OF FBS
A. Radiculopathy is a pain that shoots like a jolt of electricity and follows a particular dermatomal distribution. This is most often caused by a herniated disk, but not exclusively so. Many times there is associated sensory loss in the same dermatome. The associated myotome can manifest weakness in some cases. Abnormal reflexes can also help to localize the level of involvement in the spinal canal.
1. Imaging is helpful in this context to confirm the level implicated by the history and physical examination findings. However, it has been well-shown that healthy persons without back pain can harbor disks that would be concerning from a purely radiographic perspective. Therefore, imaging findings without a clinical correlate can typically be ignored.
2. The most common cause of the pathogenesis of radiculopathy is herniation of a disk followed closely by degenerative foraminal stenosis. Other entities, such as synovial cyst, are distinctly less common.
3. Whatever the cause, surgery for radiculopathy is focused on decompressing the affected nerve root. The prognosis is quite good; early good results are achieved in >95% of cases.
4. When this type of surgery is unsuccessful, strong consideration should be given to the possibility that the diagnosis was incorrect, the wrong level was operated on, or the patient has secondary issues that are preventing improvement.
5. Radiculopathy can be confused with hip disease in some cases. A positive Patrick’s test should be followed with an evaluation to rule out hip arthrosis.
B. Claudication is a cramping pain or sense of fatigue in the legs caused by exertion. Most patients report the onset of symptoms after walking a particular distance. The pain typically abates after several minutes of rest, such that the person can continue.
1. It is important not to confuse neurogenic and vascular claudication. Patients with neurogenic claudication exhibit a “shopping cart sign,” which is the ability to walk further when leaning forward. This flexed position slightly diminishes the ligamentous compression of the cauda equina, allowing the patient to walk further. For the same reason, a patient with neurogenic claudication will do much better on exercising bicycle than they would on walking. Patients with vascular claudication show no such improvement.
2. Neurogenic claudication is most commonly managed with lumbar laminectomy over the stenotic levels. The goal of surgery is to decompress the thecal sac by removing hypertrophied ligamentum flavum, the medial facet, and occasionally disc material. Foraminotomies are required to decompress the exiting nerve roots, and this may result in iatrogenic instability causing some patients to require fusion as well.
3. Imaging with either MRI or CT myelography shows a markedly compressed thecal sac with a characteristic trefoil configuration and amputation of the exiting nerve root sleeves.
C. Instability is another common indication for lumbar surgery. From both a theoretical and a practical standpoint, instability is distinct from stenosis and radiculopathy. Management of radiculopathy and stenosis is decompression; management of instability is fusion. The success of fusion operations is distinctly less than that of decompression. For this reason, many patients with FBS have experienced failed fusion.
1. Instability is defined as the ability of the bony components of the spine to withstand physiologic loads without mechanical pain or compromise of nerve root function.
2. Although instability often is thought of in a binomial way as either present or absent, in clinical practice there is a spectrum of instability ranging from gross instability, most often the result of trauma, to microinstability, which is found in the context of degenerative disease.
3. The underlying hypothesis in offering fusion to patients with degenerative spondylosis is that instability represents a painful dysfunctional motion segment. The pain is characteristically exacerbated by prolonged sitting or standing and often is relieved by recumbency. Because the pain does not radiate, it is not possible to localize the responsible spinal level by means of history or physical examination.
4. The pathogenesis of mechanical back pain is controversial and likely is multifactorial. There is evidence implicating the disk space as well as the facet joints. Many patients who improve after lumbar fusion fail to demonstrate overt instability on preoperative dynamic studies. Therefore, the specific pain generator is unknown, and the lumbar segment inclusive of the disc and facet joints is thought to be dysfunctional.
5. If flexion–extension radiographs (dynamic radiographs) show movement of >4 mm, the diagnosis is more certain. However, a large number of patients with movement in excess of 4 mm also do not have mechanical pain. Plain radiographs can provide indirect evidence of instability in the form of traction spurs that result from the tension placed on the bone from Sharpy’s fibers of the annulus or loss of disk height indicative of disk degeneration. MRI often shows Modic’s changes at the interspace thought to represent inflammatory reaction in the adjacent vertebral bodies secondary to disk disruption. Many of these findings are present in patients who are pain free, and therefore their utility is suspect.
6. In an attempt to better determine whether instability is present in a particular patient and whether it is responsible for the back pain being reported, several strategies have emerged. The trial use of a temporary external orthosisor percutaneous pedicle screws before surgical fusion has fallen out of favor.
a. Use of diagnostic facet blocks targeting a spinal level thought to be unstable can be helpful. Epidural steroids, although clinically beneficial, are of no diagnostic significance because they are not specific to an anatomic level.
b. Provocative diskography has been championed because it shows the disk disruption anatomically and functionally. Great care must be taken to inject both normal and diseased levels in a patient-blinded fashion in order to determine whether the targeted level(s) have pain concordant with the patient’s primary complaints of back pain.
7. Technical aspects of lumbar fusion have improved outcomes such as the use of supplemental interbody devices, less rigid implants, and bone morphogenic protein (BMP).However, it is still not possible to predict who will benefit from lumbar fusion and who will not with a high degree of certainty. This explains, in part, the relative lack of success with fusion operations compared with decompression operations for radiculopathy or stenosis. Most series have favorable outcome in 50% to 70% of cases when lumbar fusion is performed for degenerative disease.
8. If the indication for fusion was not present at the time of the first operation, revision surgery will be futile. Moreover, even when the original procedure is well conceived, revision surgery is effective only if a problem amenable to surgical correction is identified preoperatively. Examples consistent with a successful operation include pseudoarthrosis and degeneration at the level adjacent to the fusion. The plan should be well defined preoperatively.
III. SOMATIC PROBLEMS NOT RELATED TO THE SPINE
Many somatic problems not related to the spine can manifest as back pain. These must be excluded in a thorough review of systems.
A. Abdominal causes include aortic aneurysm, cholelithiasis, and pancreatitis. Pyelonephritis most often manifests as flank pain but can also lead to referred back pain.
B. In female patients, endometriosis can manifest as low back pain.
C. Sacroiliac joint (SIJ) pain is increasingly common due to abnormal shifting of the pelvis in patients with lumbar degenerative disease. Injections of the SIJ can be both diagnostic and therapeutic.
D. Osteoarthritis of the hip can be easily confused with back pain radiating into the buttock. Patrick’s test is useful to differentiate the two. Severe radiation of the pain to the groin is diagnostic of hip pathology.
E. Major depression has been shown to exacerbate the severity of back pain. It is also a poor prognostic sign for outcome after surgical intervention. Ongoing worker’s compensation litigation has also been shown to be an independent predictor of poor outcome.
IV. NONSURGICAL MODALITIES
After the rationale for the primary operation or previous operations is understood, emphasis should be given to nonsurgical modalities. In any cohort of patients with FBS, only a small number should ever come to revision surgery.
A. It should be well-understood that lumbar spondylosis is a degenerative disease. As such, surgery can ameliorate the most severe manifestations of the problem, but it can never address the underlying cause. For this reason, treatment such as weight loss, smoking cessation, and physical therapy offers the patient a better outcome, if successful, and can often make surgery unnecessary. Moreover, even when surgery is entertained, it should only be in the context of a complete treatment plan that embraces these other aspects of care. At the same time, the efficacy of surgery decreases with each subsequent operation, unless significant additional pathology has been uncovered.
B. Degenerative changes in the spine are often associated with deformities. These deformities can be focal such as spondylolisthesis, or more global, such as scoliosis. Recently, sagittal deformities have been demonstrated to have a significant impact on outcome. If these sagittal imbalances are fixed, that is, not amenable to postural changes by the patient, surgical correction of the deformity may be required. In the past, lumbar fusion surgeries focused less on sagittal issues and may have resulted in flat back syndrome, whereby the lumbar spine was fused in a hypolordotic position. Patients with FBS due to fixed sagittal imbalance require long cassette standing scoliosis studies to accurately assess the condition.
C. In most cases of FBS, the patient comes to medical attention with the chief symptom of pain. Pain is a subjective symptom. Considerable progress has been made in the development of outcome instruments (Oswestry, SF36) that attempt to quantify pain and functional level in an objective way. This has given spine specialists a means of comparing of the clinical effectiveness of various interventions.
D. The pain that patients report and the disability they experience have a great deal to do with their expectations. Pain is ordinarily an important protective phenomenon. When pain becomes chronic, as in FBS, the noxious percept that reaches consciousness serves no productive purpose. The assumption of many patients that the pain they are experiencing is evidence of ongoing damage is incorrect. When patients understand this, their perception of pain can become less noxious, and their functional abilities can improve. Therefore, patient education can play a therapeutic role.
E. As a clearer understanding emerges of the nature of pain in FBS, more effective, interdisciplinary treatments are being developed. It is increasingly recognized that depression not only exacerbates the symptoms of FBS but also is a consistent consequence of FBS. If relatively severe, depression should be addressed before surgical intervention is planned.
V. PAIN MANAGEMENT
The nonsurgical therapies discussed in IV address the causes of FBS in just as direct a manner as surgery does. In some cases, indirect measures can be considered purely with the intent of ameliorating a patient’s pain. Although these modalities are not directed at the underlying cause, they do enhance functional ability and improve quality of life.
A. There is increasing experience with narcotics in the management of chronic pain of a nonmalignant causation, such as FBS. This is an expansion of experience with cancer pain. Subsequently, some authors have expanded the indications to include patients with chronic pain of benign causation. Because the life expectancy of patients with FBS is much longer than that of patients with cancer, the duration of treatment is considerably longer. These patients must be followed for the development of toleration and habituation. Topical forms of opiates such as Fentanyl have been increasingly used for chronic pain because their pharmacokinetic profile lacks the peak and trough levels of oral opiates. Serum chemistries with liver function tests should be checked on a regular schedule. Moreover, it has been estimated that as many as 45% of patients with FBS are being medicated excessively, and this treatment remains controversial.
B. Intrathecal pumps for the administration of opioids have been used to minimize the side effects of systemic opiate therapy, such as sedation, lethargy, and decreased libido. Because the drug is delivered directly to the opiate receptors within the dorsal horn of the spinal cord, effective analgesia can be obtained at much lower doses. This leads to a much lower incidence of side effects. However, the use of intrathecal pumps has increasingly been reserved for patients with pain due to cancer, as oral and topical analgesics have become more effective.
C. The gate-control theory of pain proposed by Melzack and Wall was the inspiration for spinal cord stimulation. Modern implantable and rechargeable systems are much more effective and easier to insert than earlier versions. Most studies demonstrate favorable early and late success for these implants in treating FBS, although the cost effectiveness of these procedures is currently under debate. The complication rate is low, neurologic injury occurring in <1% of patients.
VI. SURGICAL MANAGEMENT
When conservative therapy is unsuccessful, operative intervention should be considered.
A. In cases of immediate failure, the patients never improve after surgery. This universally implies an error in diagnosis or a technical deficiency with the surgery. After the protocol outlined in IV and Vhas been implemented, these patients require MRI with and without contrast material. The prognosis for these patients is very good when an error is identified. If no deficiency is found, the outcome is considerably more discouraging. Surgery has no role in those cases.
B. The next group of failures manifests days to weeks after surgery. It is quite common, however, for patients who have initial improvement postoperatively in the hospital to experience a setback as they become more active on arriving home. This is a normal although not universal finding and is best managed expectantly. Patients who experience initial improvement and then experience clear deterioration need a more deliberate evaluation. These cases can represent recurrent disk herniation, iatrogenic instability, or sagittal imbalance. The physical examination and pain signature favor one diagnosis over the other. Recurrent nerve root compression is best evaluated with MRI with and without contrast material. Instability should be evaluated with CT to assess the bony removal, and dynamic plain radiographs. Standing long cassette studies in both the coronal and sagittal plane are required to rule out deformity. In this time frame, some patients have progressive causalgia-type or neuropathic pain. This may result from a prolonged preoperative duration of nerve compression or injury during surgery. This condition is very refractory, although recent success has been reported with Gabapentin and some antidepressants.
C. Another group of failures manifests weeks to months after surgery. The description of clear radiculopathy suggests recurrent herniation. Arachnoiditis manifests most often at this time. Patients often describe back and leg pain, which can be similar to the presenting problem. Classic cases manifest as symptoms of claudication or lower extremity causalgia. CT myelography is the study of choice and typically shows clumping of nerve roots and restricted flow of intrathecal contrast material. Surgery directed at the arachnoiditis is unsuccessful. There has been some success with spinal cord stimulators in these cases. Thankfully, the incidence has decreased dramatically with the advent of water-soluble contrast agents and the relatively infrequent use of myelography in the MRI era.
D. The last group of failures manifests after a pain-free interval of months to years.
1. Many of these cases have developed either iatrogenic lumbar instability because decompression was too wide or lumbar instability caused by the intrinsic disease. The incidence of postlaminectomy spondylolisthesis is somewhere between 2% and 10%. Even simple diskectomy has been associated with a 3% incidence of postoperative instability that necessitates subsequent fusion. It has been widely circulated that the medial half of the facet joint can be removed bilaterally without inducing instability. However, this admonishment is not consistent with the fact that the medial half of the joint comprises the descending facet almost exclusively and that removal of the medial half can leave the facet completely incompetent. There is consensus that complete laminectomy and bilateral facetectomy consistently produce instability. In cases in which this extent of resection is needed to accomplish decompression, fusion should be incorporated into the surgical plan.
2. Pseudoarthrosis after lumbar fusion manifests in a time frame similar to that of postlaminectomy spondylolisthesis. In part, the timing of presentation may represent the fact that most spine surgeons are not prepared to give up on a fusion for 9 to 12 months after surgery. There are also no commonly accepted criteria for diagnosis of pseudoarthrosis. Dynamic radiographs often appear normal, and bone scans are equivocal. The incidence of symptomatic pseudoarthrosis after a posterolateral lumbosacral fusion is between 5% and 15%. The cause can be either technical deficiency of the surgery or biologic deficiency of the patient. There is good evidence that smoking negatively affects the rate of fusion. The rate of pseudoarthrosis increases with the number of levels of arthrodesis. The use of BMP, external bone stimulators, and the use of allograft bone graft extenders have decreased the rate of pseudoarthrosis. Hardware implanted in the patient may also become symptomatic during this time frame. The modulus of elasticity of metal rods is 10 to 20 times stiffer than bone, so that even following a successful fusion, these implants may loosen at the bone-implant interface or break. Recent experience suggests alternate biomaterials, such as PEEK, may be a more suitable implant due to a its near identical stiffness to bone, and resistance to breakage.
VII. RADIOGRAPHIC EVALUATION OF FBS
Radiography plays an important role in the evaluation of FBS.
A. In the evaluation of recurrent disk herniation, contrast-enhanced CT or MRI is vital (Figs. 26.1 and 26.2). Postoperative scar becomes homogeneously enhanced. A herniated disk may have some peripheral enhancement, but because it is avascular, the disk does not become centrally enhanced. Although scar and disk can both cause compressive symptoms, surgery to remove scar tissue is rarely successful.
B. Radiographic evaluation of a fusion is more difficult. In short, there is no universally accepted way to assess successful fusion after arthrodesis.
1. Dynamic or flexion and extension radiographs are specific for instability if motion is detected. However, these studies are very insensitive. Fibrous nonunion or the instrumentation itself can prevent the flexion–extension radiographs from appearing abnormal.
2. Plain radiographs can show a robust fusion mass, but it is often impossible to know whether the bony mass is in continuity. Lucency or halos around pedicle screws suggest instability. CT with sagittal and coronal reconstruction can be helpful. Special techniques must be used to minimize artifact of the instrumentation.
3. Bone spectroscopy has been advocated, but this modality is unreliable for at least several years after surgery.
4. Three-dimensional CT has been advocated by some authors in the evaluation of FBS. It has the advantage of clearly imaging the bony resection and can clearly delineate the extent of bony fusion.
FIGURE 26.1 Preoperative lumbar MRI showing stenosis at L3–5 with an L4–5 spondylolisthesis.
FIGURE 26.2 Repeat lumbar MRI obtained 3 years following L3–5 laminectomies and fusion, and recurrent symptoms. Note junctional stenosis at the L2–3 level.
VIII. REFERRALS
A. Referral to a neurologist or spine surgeon is indicated for any patient with a new or progressive neurologic deficit. In the context of FBS and chronic pain, it is important not to miss this dramatic change in the patient’s course.
B. Imaging should be performed only in response to a significant change in the symptoms. When dramatic changes are found at imaging studies, patients with FBS should be reevaluated.
C. Often the job of weaning narcotics is left to an internist or general neurosurgeon. This can be appropriate; however, when reduction goals are not being met and the program becomes stalled, these patients should be referred to specialized centers. Long-term use of narcotic analgesics can be acceptable treatment in some cases, but it should be chosen explicitly, not as an ad hoc default.
D. FBS is a difficult management problem, and all these patients should be seen by a spine surgeon or pain center. In general, patients with FBS should be cared for by a multidisciplinary team. It is important to maintain vigilance in case new important symptoms arise which may alter the treatment regimen.
Recommended Readings
Burton AK, Waddell G, Tillotson KM, et al. Information and advice to patients with back pain can have a positive effect. Spine. 1999;24:2484–2491.
Chou R. Nonsurgical interventional therapies for low back pain: a review of the evidence for an American Pain Society clinical practice guideline. Spine. 2009;34:1078–1093.
Frey M, Manchikanti L, Benyamin R, et al. Spinal cord stimulation for patients with failed back surgery syndrome: a systematic review. Pain Physician. 2009;12:379–397.
Jee-Soo J, Sang-Ho L, Jun-Hong M, et al. Surgical treatment of failed back surgery syndrome due to sagittal imbalance. Spine. 2007;32:3081–3087.
Line J. Spinal cord stimulation versus conventional medical management for failed back surgery syndrome: long-term results from the PROCESS study. Neurosurgery. 2008;62:1426–1427.
Melzack R, Wall PD. The Challenge of Pain. New York, NY: Basic Books; 1983.
Van Buyten J, Linderoth B. “The failed back surgery syndrome”: definition and therapeutic algorithms—an update. Eur J Pain Suppl. 2010;4:273–286.
Weinstein J, Lurie JD, Tosteson TD, et al. Surgical versus nonoperative treatment for lumbar disc herniation. Spine. 2008;33:2789–2800.
Weinstein J, Lurie JD, Tosteson TD, et al. Surgical compared with nonoperative treatment for lumbar degenerative spondylolisthesis. J Bone Joint Surg. 2009;91:1295–1304.