Introduction

Failed back surgery syndrome (FBSS) is the name given to the chronic pain syndrome that occurs when a surgical procedure involving the lumbosacral spine culminates in persistent or recurrent pain.

By 1985, investigators recognized that FBSS is an important public health problem affecting 25,000 to 50,000 new patients each year.¹ Many causes have been implicated for the development of FBSS, including inappropriate patient selection criteria for lumbosacral surgical procedures, shortfalls in the surgeon’s diagnostic or technical skills (operation not indicated, wrong site, incomplete decompression, and/or fusion),² and inadequacies in available surgical techniques.

In the United States, FBSS is the primary indication for spinal cord stimulation (SCS), even though SCS is not the primary therapy for FBSS. SCS is a modern application of the ancient use of electricity to treat pain (in ancient times healers used current generated by electrical fish).

Electrical stimulation with implanted devices followed the 1965 publication of Melzack and Wall’s gate control theory of pain³ and the development of cardiac pacemaker technology. Today SCS is delivered with sophisticated techniques that take advantage of multichannel pulse generators powered by rechargeable batteries.

By reducing the level of pain associated with FBSS, SCS can allow patients to decrease or eliminate use of pain medication, improve their physical functioning and ability to engage in the activities of daily life, enhance their quality of life, and return to work. The success of SCS, however, depends in large part on the physician’s ability to (1) select appropriate candidates, (2) use the right equipment to treat a specific pain condition, and (3) implant and adjust the equipment in an optimum manner.

Establishing Diagnosis

Diagnosing FBSS seems fairly straightforward in patients suffering from persistent or recurrent pain following a surgical procedure on the lumbosacral spine. However, because FBSS has many possible etiologies, identification of the syndrome is merely the first step in meeting its treatment challenges.

The assessment of an FBSS patient should be multifaceted and should follow the same procedure used for any chronic pain syndrome. Review of the patient’s history and operative record (which should be sought routinely) helps to establish the underlying diagnosis. The presence of any issues of secondary gain, psychological or behavioral problems, or co-morbid pain conditions is of interest. A thorough pertinent physical examination helps to corroborate the diagnosis. A validated numeric rating or visual analog scale (VAS) can help to determine the intensity of the pain and follow the patient’s progress.⁴ Imaging studies provide valuable information that will guide treatment. Abnormalities revealed by imaging studies and the physical examination should be consistent with the patient’s pain. The patient might demonstrate nonorganic responses (Waddell signs) during the physical examination,⁵ but organic findings should predominate.

Specific prognostic factors have been identified for patients with FBSS concerning the likelihood that they might benefit from SCS. For example, the extent to which the patient’s pain is radicular is important (relieving axial low back pain with SCS is technically more difficult than relieving radicular pain).^6,7 Technological advances, however, continue to permit clinicians to improve outcomes and extend the circumstances in which SCS is indicated for FBBS.

Anatomy

FBSS often occurs because a surgeon has assumed that a patient’s pain was caused by an anatomical abnormality that could be corrected with a surgical procedure. The same or similar anatomical abnormalities, however, might occur in asymptomatic individuals.⁸ Indeed, in consecutive patients with FBSS, Long and associates² reported that most did not meet standard indications for their first surgical procedure. Treatment of FBSS with a repeat surgical procedure remains indicated, however, if a patient has a large disc fragment or severe stenosis compressing a nerve or nerves and causing a significant neurologic deficit or if there is gross spinal instability.

Computer modeling of the electrical fields produced by SCS in the spinal cord^9,10 revealed current and voltage distributions consistent with those found in studies of cadavers and primate spinal cords. The modeling studies have predicted that bipolar stimulation with closely spaced electrical contacts separated by 6 to 8 mm or less would be the best way to target longitudinal midline fibers and that the electrical field between two cathodes bracketing the physiological midline does not sum constructively in the midline.¹¹ Modeling has also predicted advantages for three or more columns of contacts with lateral anodes.¹² Clinical experience has confirmed that correct positioning and spacing of SCS electrodes is essential for pain relief.¹³ The longitudinal position of an electrode largely determines the segmental effects of stimulation; and rather than being beneficial, positioning electrodes more cephalad than the target area commonly elicits unwanted local segmental effects.¹⁴

Electrical impulses are more easily conducted through cerebrospinal fluid (CSF) than through any other tissue in the spinal canal. Because the depth of cerebrospinal fluid differs at various locations in the spine, thresholds and recruitment patterns vary; because CSF depth changes as an individual shifts between supine and prone positions, patients usually report that the perceived intensity of stimulation-induced sensation (paresthesia) changes when they change their posture.

Basic Science

Neuropathic pain, which often produces a radiating, burning sensation, results from nerve or nervous system damage. Activation of peripheral nerve fibers in pathological conditions can more easily activate wide dynamic range (WDR) neurons in the superficial laminae of the corresponding dorsal horn, resulting in hyperalgesia (extreme sensitivity to pain) and/or allodynia (pain from normally nonpainful stimuli).

In experimental studies SCS suppressed long-term potentiation of WDR neurons by reducing the C-fiber response¹⁵ and also changed the concentration of several neurotransmitters and their metabolites in CSF, including serotonin and substance P,¹⁶ glycine, adenosine, and noradrenaline.^17,18 Supraspinal microdialysis in conscious rats revealed that SCS causes γ-aminobutyric acid (GABA) release in periaqueductal grey matter.¹⁹ SCS also induces GABA release in the dorsal horn²⁰ (activation of the GABA-B receptor might be responsible for the therapeutic effect of SCS) and decreases the release of glutamate and aspartate.²¹ It is likely that SCS has additional, complicated effects on as-yet-unidentified neural transmitters and modulators. SCS is not thought to affect opioid receptor-mediated analgesia because naloxone does not inhibit SCS efficacy.²²

In patients undergoing successful SCS treatment to reduce otherwise intractable neuropathic leg pain, positron emission tomography (PET) studies suggest that SCS also modulates supraspinal neurons.²³ In these patients SCS increased cerebral blood flow significantly in the thalamus contralateral to the painful leg and in the associated bilateral parietal area, and this was associated with changes in pain threshold. SCS also activated the anterior cingulate cortex and prefrontal areas, which control the emotional response to pain.

Imaging

Imaging studies undertaken to establish the diagnosis of FBSS provide information about the patient’s postsurgical anatomy and whether or not the anatomic goals of the surgical procedure were met. This might explain the failure of the surgical procedure to relieve pain; however, established nerve injury can lead to persistent pain, even after technically successful surgery (Table 8-1). Radiographic imaging studies should reveal abnormalities concordant with the patient’s current pain complaints (see Table 8-1).

Table 8-1 Uses of Imaging Technology in Spinal Cord Stimulation Treatment

CT, Computed tomography; FBSS, failed back surgery syndrome; MRI, magnetic resonance imaging; SCS, spinal cord stimulation.

Imaging is also used to guide SCS treatment (see Table 8-1). For example, imaging the thoracic spine provides valuable information about the placement of thoracic electrodes. Imaging should take place before the procedure to rule out any pathological condition that might contribute to the patient’s pain or confound (or increase the risk of) electrode placement (e.g., stenosis). Fluoroscopic imaging during the procedure helps guide placement of the electrode and documents the final electrode position. Imaging is also used to diagnose the cause of a complication such as suspected electrode migration or fracture.

Guidelines

Many guidelines are published for medical therapies, invoking principles of evidence-based medicine (EBM). Ironically, to date little evidence exists that EBM or guidelines have improved patient care. We have published a set of practice parameters as a reference for referring physicians, clinicians offering SCS treatment, and patients.²⁴

Indications/Contraindications

To be eligible for SCS, FBSS patients must have pain that is refractory to more conservative care. The definition of “more conservative” is not precise; for example, it is a matter of opinion whether opioid therapy is more or less “conservative” than SCS, and some patients are referred for SCS to avoid opioids. Neuropathic pain is generally more responsive to SCS than is nociceptive pain; distinguishing these is not always straightforward (e.g., FBSS), and a therapeutic trial of SCS might be the most practical approach to determining eligibility. Likewise, radicular pain is generally more responsive than axial low back pain; again, individual cases might be most practically approached by simply offering an SCS trial.

Relative contraindications to SCS include unresolved issues of secondary gain (e.g., an outstanding lawsuit or compensation claim), a major untreated psychiatric co-morbidity, and/or inappropriate medication use. The presence of a demand cardiac pacemaker requires electrocardiogram (ECG) monitoring and/or changing the pacemaker mode to a fixed rate.²⁵

Absolute contraindications include uncorrected coagulopathy, untreated sepsis, a patient’s inability to cooperate or to control the device, and/or a projected need for the patient to undergo magnetic resonance imaging (MRI).

SCS is problematic if the patient has a separate, co-morbid chronic pain syndrome. As noted previously, before receiving SCS treatment some FBSS patients require a repeat surgical procedure to correct a serious anatomical defect.

Equipment

Equipment needed for the SCS screening trial (see following paragraphs) includes an electrode that will be connected to an external pulse generator and external programming equipment. A complete SCS system for chronic use requires at least one electrode (Fig. 8-1) with an extension cable and an implantable pulse generator (IPG) (Fig. 8-2).

Fig. 8-1 Spinal cord stimulation electrodes are arrays with multiple contacts. Some require laminectomy; others are inserted percutaneously through a modified Tuohy needle.

Fig. 8-2 Implanted pulse generators used for spinal cord stimulation support multiple contacts, which may be programmed noninvasively as anodes and cathodes. Some are powered by implanted batteries (visible here); older versions accept or require power from an external device.

Two types of SCS electrodes are available: percutaneous catheter electrodes and plate/paddle electrodes (also known as laminectomy, surgical, or insulated electrodes). Percutaneous electrodes can be inserted with a minimally invasive procedure using a Tuohy needle. When performed under fluoroscopy, percutaneous placement facilitates longitudinal mapping of multiple levels for optimal positioning of the electrode.

Placement of plate/paddle electrodes requires surgical exposure of the epidural space. Plate/paddle electrodes have dorsal insulation to protect against excess posterior stimulation, and they offer better performance than do percutaneous electrodes in FBSS patients.^26,27 They are available in one-, two-, three-, and five-column configurations. Compared with percutaneous electrodes, plate/paddle electrodes require only half the battery power.²⁸ They require open (albeit minimal) exposure; this limits longitudinal mapping. They are more difficult to revise, remove, or replace, once encapsulated in scar tissue; however, this makes them inherently more resistant to migration.

Each type of electrode has multiple electrical contacts that can be configured in a multitude of ways (various combinations of anode/cathode/off/on). The SCS programming options are so numerous that it is impossible to test every combination (e.g., a four-contact electrode has 50 functional bipolar combinations of anodes and cathodes, an eight-contact electrode has 6050). Computerized methods are useful in finding and recording options for an individual patient.²⁹ Typical stimulation parameters are set at 60 Hz frequency (pulse repetition rate) with 0.2- to 1-msec pulse width. Amplitude should be adjusted to the minimum level, on a scale from perceptual to discomfort (or motor) threshold that elicits adequate coverage of the area(s) of pain by paresthesia.

The longitudinal position of the electrode determines which segment of the body will experience paresthesia, and bipolar (or tripolar) stimulation has the greatest selectivity for longitudinal midline fibers.¹³ FBSS patients with associated axial low back pain require low thoracic electrode placement and sometimes need complex electrode arrays.

As shown in Fig. 8-2, the stimulator energy sources in use are: (1) radiofrequency-coupled passive implants that have a long life but require an external antenna, which can cause skin irritation and fluctuations in stimulation amplitude; (2) primary cell IPGs that require replacement at the end of battery life; and (3) IPGs with rechargeable batteries. Patients can turn IPGs on and off and use either an external magnet to make limited adjustments in amplitude or a remote transmitter capable of complicated adjustments.

Technique

This following discussion of the placement of an SCS system should not be confused with installation instructions. This information is meant to give the reader an idea of what is involved in the procedure rather than to bestow permission to undertake SCS implantation. Any specialist who wishes to offer SCS must have appropriate training, which must be supervised by an experienced implanter. The SCS procedure is straightforward, but positive results require facility and complete understanding of sophisticated details. Therefore, the following should be considered a general description.

The technical goal of SCS for FBSS is to cover the area of pain with a tingling sensation known as paresthesia (Table 8-2). Pain/paresthesia overlap is necessary (but not sufficient) to achieve pain relief. The paresthesia must be comfortable, and the stimulation must not cause a motor reaction. If pain/paresthesia overlap only occurs with uncomfortable stimulation (i.e., outside of the [often narrow] “usage range” between perception of paresthesia and discomfort or motor effects), treatment is compromised. In addition, the perception of extraneous stimulation paresthesia outside the area(s) of pain should be minimized.

Table 8-2 Positive Outcomes of Spinal Cord Stimulation Treatment

Paresthesia can be directed to one location or another by changing stimulation parameters (pulse amplitude, width, and repetition rate) and specific contact combinations (anode/cathode/off).

Patient Management/Evaluation

After a routine postoperative examination and SCS programming adjustment, the patient should return for suture or staple removal and any needed additional adjustments on postoperative day 7 to 14. Routine follow-up visits should taper from monthly to annually, with additional visits scheduled as needed to ensure safe and effective operation of the stimulator.

SCS patients should disable the system before entering an electromagnetic field produced by antitheft devices, a metal detector, or any other security-scanning device. They should also avoid scuba diving more than 10 m deep or entering hyperbaric chambers with an absolute pressure above 2.0 atm. They should not engage in any activity that might place excessive stress on the implanted system.

Special precautions are required for an SCS patient to undergo certain medical tests such as cardiac monitoring, radiation therapy with the IPG in the active field, radiofrequency ablation, and electrocautery. Ultrasound over the device and diathermy anywhere are contraindicated, as is MRI, although the latter is under study.

If SCS pain relief disappears, it is important to determine first if the system is operating correctly. Sometimes fibrosis around implanted electrodes increases impedance and interferes with treatment; this generally can be overcome through reprogramming. A minority of SCS patients experience clinical failure, which is the unexplained loss of pain relief despite a functioning system that continues to provide appropriate pain/paresthesia overlap. When this occurs, SCS can sometimes be potentiated with adjuvant medication.³¹

Successful treatment of infection usually requires removal of the entire SCS system. A new system can be placed after the infection clears. Therefore, not only is infection costly, but the patient loses the SCS pain control during the treatment period. Clinicians and patients should take appropriate measures to avoid infection.

Outcomes Evidence

Patient-rated pain relief is the usual primary outcome criterion for SCS and for pain treatments in general, with success commonly defined as a minimum 50% relief (see Table 8-2). Secondary outcome measures include ability to conduct activities of daily living, work status, medication requirements, neurological function, and patient satisfaction with the procedure. To reduce bias in SCS studies, collection of follow-up data by a disinterested third party is desirable. Since SCS elicits paresthesia, blinding is not feasible; thus any randomized controlled trial (RCT) of SCS loses points on rating scales commonly used in EBM.

Two decades ago we compared retrospective data in FBSS patients who underwent reoperation³² with those who received SCS and found that SCS patients enjoyed reduced morbidity and pain and improved neurological function, quality of life, and ability to engage in activities of daily living.^6,33

In the first RCT of SCS vs. reoperation in FBSS patients,³⁴ 45 subjects (90% of those who received insurance authorization for study participation) were available for a mean follow-up of 3 years. SCS success was 9 of 19, whereas reoperation success was 3 of 26. Only 5 of the 24 subjects randomized to SCS crossed to reoperation, whereas 14 of the 26 randomized to reoperation crossed to SCS. No patient who crossed from SCS to reoperation achieved success with reoperation, but 6 of the 14 who crossed from reoperation to SCS achieved success with SCS. Success was defined as at least 50% pain relief and patient satisfaction with treatment.

An international multicenter RCT (the PROCESS study)³⁵ randomized 100 subjects with FBSS to conventional medical management (CMM) or SCS plus CMM. By 6 months the subjects randomized to SCS achieved significantly greater pain relief and improved functional capacity and health-related quality of life than did those randomized to CMM. The investigators followed the 42 subjects randomized to SCS who actually received SCS for 24 months and found significantly improved leg pain relief, functional capacity, and quality of life compared with baseline scores.³⁶ At the time the randomized group reached 24 months’ follow-up, 72 patients had received SCS as a final treatment, through either randomization or cross over. Of these, 34 (47%) achieved the primary outcome (more than or equal to 50% pain relief) vs. 1 of 15 patients who received only CMM.

The initial cost of an SCS system is high. Nevertheless, several cost-effectiveness analyses have demonstrated that SCS treatment lowers the total cost of health care for patients with neuropathic pain compared with alternative treatments. In a cost study based on data from the first 40/42 (of 50) patients enrolled in the RCT of SCS vs. reoperation,³⁴ every analysis (intention-to-treat, treated-as-intended with cross over counted as failure of randomized treatment, and final treatment) showed that SCS achieved economic dominance by being more effective and less expensive than reoperation.³⁷

In 2008 the National Institute for Health and Clinical Excellence (NICE) in the United Kingdom conducted a systematic review and technology assessment of the use of SCS.³⁸ The model, which compared the cost of treating FBSS with SCS vs. CMM and reoperation and assumed an IPG battery life of 4 years, predicted that SCS would produce additional quality-adjusted life years at a cost the United Kingdom health service would be willing to pay.

Despite the fact that the rechargeable IPG systems cost more initially than the primary cell systems, the rechargeable systems should improve the cost-effectiveness of SCS by reducing the cost and potential morbidity associated with replacing IPGs because of battery depletion.

Technical Aspects of Spinal Cord Stimulation

Beginning in 1986 we developed and used a patient-interactive computer program to allow patients to find the parameter settings that would optimize pain/paresthesia overlap and battery life and as a means of conducting blinded RCTs comparing electrode designs.³⁹ In one study, we compared results with a four-contact percutaneous electrode vs. a four-contact surgical plate/paddle electrode and found that the plate/paddle electrode provided the best pain/paresthesia coverage, low back coverage, and battery longevity.²⁸ At mean follow-up of 1.9 years, compared with patients with percutaneous electrodes, twice as many patients with plate/paddle electrodes reported a successful outcome (at least 50% sustained relief of pain and patient satisfaction) and a reduction or elimination of pain medication.²⁶ A statistically significant advantage for the plate/paddle electrode disappeared, however, at longer follow-up in our small sample of 24 patients.

To determine if use of two percutaneous electrodes bracketing the physiological midline would enhance paresthesia coverage of the low back, we tested parallel percutaneous electrodes bracketing the midline (for chronic use) vs. a single percutaneous electrode placed on the midline for temporary use during the screening trial.⁴⁰ The single electrode provided the best pain/paresthesia overlap at the lowest amplitude requirement. Nevertheless, 53% of the patients reported success at 2.3-year mean follow-up with the parallel percutaneous electrodes.

Next we compared results of treating axial low back pain using a four-contact percutaneous electrode with those obtained using a surgical plate/paddle electrode with two parallel rows of eight contacts (16 total).⁴¹ The percutaneous electrode provided marginally better pain/paresthesia overlap with significantly improved symmetry using significantly lower voltage. Compared with the surgical plate/paddle electrode, however, the percutaneous electrode required a slightly higher scaled amplitude to cover the low back and produced significantly increased extraneous coverage.

Our findings indicate that an electrode array comprising dual columns of contacts bracketing the midline presents disadvantages that might be overcome if a third column is placed on the midline. Indeed, a computer model developed at the University of Twente predicts the effect of such electrode configurations on the stimulation of dorsal column and root fibers.⁴² Both a longitudinal tripole electrode and a transverse tripole (each with a central cathode) reportedly have advantages in selectively recruiting the presumed stimulation target neurons.¹²

Reducing Risks and Avoiding Complications

The risks and complications associated with SCS can be characterized as biological, procedural, and equipment-related. Sometimes routine system maintenance (e.g., replacing a depleted battery or adjusting stimulation parameters) is mistakenly referred to as a complication.

SCS implantation can lead to spinal cord or nerve injury, dural puncture causing CSF leakage, hematoma, or infection (Table 8-3). To avoid spinal cord or nerve injury, it is helpful to obtain an MRI or computed tomography (CT) myelogram of the target area before placing an electrode. To avoid dural puncture, the patient should be conscious during the procedure; and, if possible, electrodes should not be placed in scarred areas. A standard preoperative coagulation profile should be undertaken to help avoid hematoma, and the patient should be monitored overnight. Standard precautions should be taken to avoid infection.

Table 8-3 Reducing the Risk of Spinal Cord Stimulation Complications

CSF, Cerebrospinal fluid; CT, computed tomography; IPG, implantable pulse generator; MRI, magnetic resonance imaging.

Equipment-related complications include generator failure, electrode fatigue fracture, electrode migration, and disturbance from exposure to an electromagnetic field (see Table 8-3). Generators can fail, but the incidence of battery failure can be reduced by helping the patient learn to use the system properly and by implanting rechargeable batteries. Lead fatigue failure can be reduced by minimizing the use of connectors, positioning service loops to relieve strain, avoiding crossing mobile body segments, and placing the generator in the patient’s flank or lateral abdomen. Fixing the electrode in place properly virtually eliminates migration.³⁰

Conclusion

The multiplicity of the factors that affect chronic pain, even in syndromes less complex than FBSS, makes it difficult to construct and conduct rigorous studies to determine appropriate therapies and patient selection criteria. The cross-over randomized trial comparing SCS with reoperation is likely as rigorous as possible and shows clear benefits for SCS in its carefully selected patient population. The available literature on cost-effectiveness provides evidence that, despite its initial cost, SCS pays for itself within a few years.

FBSS patients will likely be best served when clinicians and investigators begin to identify, study, and treat subcategories of FBSS (i.e., to redefine the syndrome). This is likely the necessary first step toward identifying the most appropriate medical, surgical, and neuromodulation therapy for the various subtypes of FBSS. In the meantime, the paucity of available literature despite the large number of people who suffer from FBSS indicates that, until we learn to prevent FBSS, more research on its treatment needs to be conducted and published. In the present state of the art, SCS has proved to be superior to alternative treatments for FBSS in properly selected patients, and SCS technology continues to improve.