CHAPTER 107 Spinal Cord Stimulation in Chronic Pain

Robert E. Windsor, Markus Niederwanger, Steven Lobel

INTRODUCTION

Since the first published paper on spinal cord stimulation (SCS) by Shealy in 1967 there have been a total of over 2000 articles, presentations, symposia, and abstracts on the topic of neuroaugmentation.^88,⁹² The long-term results of SCS published in the 1970s were disappointing yet promising.^28,^30,^32,⁹³ Most of the studies published in the 1970s and early 1980s demonstrated success rates of approximately 40%.¹² As with many new devices, problems included poorly designed hardware and software, inadequate patient selection criteria, and suboptimal surgical technique. Early on, the hardware typically consisted of a single or dual electrode system that was implanted epidurally. They provided a small electrical field and thus were unable to consistently stimulate the spinal cord. In addition, these systems were implanted via laminectomy or laminotomy with the patient under general anesthesia, thus eliminating the possibility of surgeon–patient interaction. The electrodes were commonly implanted in the high thoracic or lower cervical region for lumbar pain syndromes and patients were not consistently screened for psychological dysfunction, drug habituation, secondary gain issues, pain topography, and quality of pain. All of these factors have considerable impact on the overall efficacy of SCS.

Significant advances in SCS have been made in recent years. The hardware is more durable and the electrode size and interelectrode distance have been improved. Improved software allows for more complex programming and multiple, simultaneously operating programs which generally allows for improved coverage and pain relief of difficult pain patterns. The devices may be implanted percutaneously under fluoroscopic guidance, which allows operator–patient interaction and may lead to more accurate positioning of SCS leads. Paddle-style leads implanted via laminectomy may have certain advantages relating to energy consumption and long-term maintenance of back pain. Three decades of experience have provided improved patient selection criteria. The net result is an improved capacity to control chronic pain.¹² This chapter will discuss the clinical indications, outcomes, complications, and other methods of implantation which may positively impact future outcomes and areas of study.

PAIN ANATOMY AND PHYSIOLOGY

Pain is an uncomfortable sensation associated with an emotional response.^35,¹⁰⁶ The International Association for the Study of Pain (IASP) defined pain as ‘an unpleasant sensory and emotional experience associated with actual and potential tissue damage, or described in terms of such damage.’³⁵ It may originate from stimulation of chemical, mechanical, or thermal receptors found in free nerve endings within injured tissue. This is known as afferent pain, and can occur in ligamentous or muscular injuries of the spine.^11,^47,^54,⁹⁰ Pain can also occur from direct injury to the peripheral nerve, which results in burning or shooting pain in the distribution of the affected nerve. This is called peripheral deafferentation (neuropathic) pain and is demonstrated in conditions such as complex regional pain syndrome, peripheral neuropathy, or radiculopathy.^30,^64,⁶⁵ Central deafferent pain appears after injury to certain structures within the central nervous system, such as the thalamus, that are responsible for the transmission of pain.

Peripheral pain signals are transmitted by either thinly myelinated A-delta or unmyelinated C fibers. The A-delta fibers convey discrete, sharp, fast pain at approximately 15 m/sec, whereas the C fibers transmit vague, chronic, burning, slow pain at less than 1 m/sec.^25,¹⁰⁸

Pain fibers typically enter the spinal cord through the dorsal root and then ascend or descend two to six segments within the dorsolateral fasciculus (Lissauer’s tract)^25,¹⁰⁸ The A-delta fibers synapse with the dorsal gray horn neurons located in laminae 1, 2, 5, and 10, whereas the C fibers synapse with dorsal gray horn neurons located in laminae 1, 2, and 5. The majority of fibers then cross to the opposite ventrolateral portion of the spinal cord before ascending in the spinothalamic, spinoreticulothalamic, and spinomesencephalic tracts. The lateral spinothalamic fibers terminate in the thalamic ventralis posterolaterales and posteromedialis nuclei, from which fibers are projected into other areas of the thalamus and to the somatic sensory cortex. The medial spinothalamic, spinoreticulothalamic, and spinomesencephalic tracts end in the reticular activating system within the medulla, pons, midbrain, periaqueductal gray, hypothalamus, and thalamic medial and intralaminar nuclei (Fig. 107.1).

Fig. 107.1 Neuroanatomic pathway for nociceptive pain transmission. (A) Lateral tracts and (B) medial tracts.

In: Bonica JJ, ed. The Management of Pain. 2nd edn. Philadelphia: Lea & Febiger; 1990:89.

The thalamus plays the primary role for conscious pain perception, and the cortex is involved in interpreting pain quality and locality. The A-delta fibers convey a distinctive, sharp pain, and C fibers conduct a characteristic diffuse, burning, or aching pain. This is likely a reflection of the A-delta fibers terminating at the cortical level versus C fibers, which end diffusely in the brain stem and diencephalon.

In 1965, Melzack and Wall published their ‘gate control’ theory in which they hypothesized that a ‘gate’ system existed for pain modulation located in the dorsal gray horn within the substantia gelatinosa (laminae 2 and 3).⁶⁰ They proposed that excess tactile signals traveling along the large myelinated A-delta fibers closed the gate, which then inhibited the propagation of pain impulses along the unmyelinated C fibers (Fig. 107.2).

Fig. 107.2 Melzack and Wall gate control theory of pain: large-diameter fibers (L), small-diameter fibers (S), substantia gelatinosa (SG), first central transmission cells (T). excitation (+), inhibition (–).

From Melzack R, Wall PD. Pain mechanisms: A new theory. Science 1965; 150(699):971–979.

Although the pain pathway is still not completely understood, researchers have uncovered important parts of the neuronal system. This includes descending inhibitory influences from the brain, which have been shown to suppress transmission of pain.^10,^80,⁸⁶ There is also evidence of an endogenous system of opioids that modulate sensory input.^41,^89,⁹⁵

Today, there is a better awareness that the pain experience is not just physiologic but is also influenced by culture, religion, and psychological makeup.^29,^37,^61,⁶² In order to provide appropriate treatment, all of these factors must be taken into consideration when evaluating patients.

MECHANISM OF ACTION

Although the exact mechanism for pain control from SCS is not entirely understood, it is believed to result from direct or facilitated inhibition of pain transmission.^28,^30,^32,^40,^60,⁹² There exist five mechanistic theories for SCS which should be noted: (1) gate control theory – segmental, antidromic activation of A-beta efferents; (2) SCS blocks transmission in the spinothalamic tract; (3) SCS produces supraspinal pain inhibition; (4) SCS produces activation of central inhibitory mechanisms influencing sympathetic efferent neurons; and (5) SCS activates putative neurotransmitters or neuromodulators.⁴⁰

The gate control theory motivated Shealy et al. in 1967 to apply SCS as a means to antidromically activate the tactile A-beta fibers through dorsal column stimulation.⁹² Shealy et al. reasoned that sustained stimulation of the dorsal columns would keep the gate closed and provide continuous pain relief. While the theoretical model put forth by Melzack and Wall has been shown not to be precisely accurate, pain gating or pain control has been shown to exist.^28,^30,^32,⁶⁰

Others believe that pain relief from SCS results from direct inhibition of pain pathways in the spinothalamic tracts and not secondary to selective large fiber stimulation.²¹ This theory has been supported by Hoppenstein, who showed that the posterolateral stimulation of the spinal cord provided effective contralateral pain relief with substantially less current than posterior stimulation.³⁴

Some investigators speculate that the changes in blood flow and skin temperature from spinal cord stimulation may affect nociception at the peripheral level.^11,^17,^27,^54,⁵⁵ This postulate is further supported in part by data from Marchand et al. who investigated the effects of SCS on chronic pain using noxious thermal stimuli.^22,^31,^36,^58,^70,⁸⁷ Since it was discovered that SCS causes vasodilation in animal studies, clinicians have used this modality for the treatment of chronic pain due to peripheral vascular disease and this is one of the leading indications for SCS in Europe today.^31,^36,^40,^87,¹⁰⁰ The precise methods of action of pain modulation by SCS are still being elucidated. A better understanding of the pain system may lead to continued improvement in SCS hardware and software, implantation techniques, and even improved clinical outcomes.

SPINAL CORD STIMULATION LEADS

Only three companies manufacture SCS systems in the United States: Medtronic, Inc. (Minneapolis, MN), Advanced Neuromodulation Systems (Dallas, TX), and Advanced Bionics (Sylmar, CA).

Advanced Neuromodulation Systems (ANS) produces several leads for percutaneous placement with four or eight electrodes (Fig. 107.3A). The four-electrode lead has four 3 mm electrodes separated by an interelectrode distance of either 4 mm or 6 mm and spans a distance of either 24 mm or 30 mm. The eight-electrode lead has eight 3 mm electrodes with an interelectrode distance of 4 mm. The notion of spanning a larger distance is to help thwart the effects of possible migratory pain patterns which is thought to be a quality of complex regional pain syndrome type 1 or to provide programming redundancy in case of lead migration. The four-electrode lead spans one average thoracic vertebral body while the eight-electrode lead spans two average vertebral bodies or three average cervical vertebral bodies.

Fig. 107.3 (A) Advanced Neuromodulation Systems percutaneous implantable lead types. This picture demonstrates an eight-electrode Octrode lead and two four-electrode leads. The electrodes on all leads are 3 mm long. The eight-electrode and one of the four-electrode leads have an interelectrode distance of 4 mm. The other four-electrode lead has an interelectrode distance of 6 mm. (B) Advanced Neuromodulation Systems laminectomy implantable lead types. This picture demonstrates four paddle-style leads implanted via laminectomy. There are two wide leads and two narrow leads. One of the narrow leads has eight electrodes and the other one has sixteen electrodes. There are two wide leads; one has eight electrodes and the other one has sixteen electrodes.

ANS also produces several different paddle electrodes for surgical implantation via laminotomy or laminectomy. They have a four-electrode Lamitrode Four, an eight-electrode Lamitrode 44 which has four electrodes side by side, and eight-electrode Lamitrode 8, and a 16-electrode Lamitrode 88 with eight electrodes side by side. ANS recently released curved contour laminotomy leads in dual quad, and dual Octrode configurations. The new curved contour laminotomy leads conform more accurately to the shape of the epidual space in the dorsal spinal canal and potentially provide more consistent service than previous laminotomy leads (Fig. 107.3B).

Medtronic also manufactures leads designed for percutaneous or laminotomy implantation. The percutaneously implanted leads have either four or eight electrodes. They have a tough, polyurethane outer covering and a helicoil substrate making the leads very resilient with columnar strength and flexibility. There are three different four-electrode leads and one eight-electrode lead. They each have variable electrode lengths and interelectrode distances. The Pisces Quad Plus has four 6 mm electrodes with an interelectrode distance of 12 mm; the Pisces Quad lead has four 3 mm electrodes with an interelectrode distance of 6 mm; the Pisces Quad Compact lead has four 3 mm electrodes with an interelectrode distance of 4 mm; and the Octad has eight 3 mm electrodes with an interelectrode distance of 6 mm. In addition to these lead configurations, Medtronic has the capability to individually produce a wide variety of four- or eight-electrode leads to accommodate an individual physician’s specifications or to treat complex or difficult pain patterns. In general, the smaller the interelectrode distance, the less risk for rootlet stimulation.

Medtronic also produces five electrodes for implantation via laminotomy: the Symmix, the dual-paddle Symmix, the Resume TL, the Resume, and the Specify. The Symmix, Resume TL, and the Resume leads have four electrodes each, the Specify has eight, and the dual-paddle Symmix has two paddles with two electrodes each. The Symmix has four electrodes arranged in a diamond pattern and the Specify has a total of eight electrodes with four electrodes arranged side by side. Both of these electrode arrangements facilitate bilateral extremity stimulation. The dual-paddle Symmix has two separate paddles designed to allow implantation over two contact sites on the spinal cord to cover a more complex pain pattern. The Resume lead is the most commonly implanted paddle electrode.

Both Medtronics and ANS have internally implanted batteries with pulse generators (internal pulse generator or IPG) and an externally powered radiofrequency (RF) controlled implanted receiver. Both IPG powered systems have the advantage of having the device completely internalized, which makes the system more flexible in that the patient can swim with it on if desired. However, the IPG powered system is limited in that at this time it can only power a maximum of eight electrodes at a limited frequency and will require surgical replacement at some point in time. The RF receiver systems have the advantage of being externally powered, thus obviating the requirement for surgical replacement. The ANS receiver also has the capacity to run at a frequency of up to 1500 Hz, thus potentially allowing for a local anesthetic effect on the stimulated nerves.^2,⁵⁰ Both the ANS IPG and receiver have the capacity for complex programming and running several different programs simultaneously, which may have benefits when dealing with complex pain patterns, and has yet to be evaluated in a controlled study (Fig. 107.4).

Fig. 107.4 (A) Advanced Neuromodulation Systems dual lead bipole. This picture demonstrates the electrical field experienced by the posterior spinal cord with a single active electrode on each percutaneously placed Octrode. Note that the bulk of the electrical field is biased to the left side, the cathode. (B) Advanced Neuromodulation Systems dual lead bipole. This picture demonstrates the electrical field experienced by the posterior spinal cord with a single active electrode on each percutaneously placed Octrode. Note that the bulk of the electrical field is biased to the right side, the side of the cathode. (C) Advanced Neuromodulation Systems single lead guarded array. This picture demonstrates the electrical field experienced by the posterior spinal cord with a single-lead guarded array. This is a cathode with an anode on either side of it. Note that the bulk of the electrical field is biased in the center adjacent to the cathode. (D) Advanced Neuromodulation Systems transverse guarded array. This picture demonstrates the electrical field experienced by the posterior spinal cord with a transverse guarded array. This is a cathode on one lead with two anodes on the other lead. Note that the bulk of the electrical field is biased to the side of the cathode.

PATIENT SELECTION CRITERIA – INDICATIONS

Proper patient selection is essential to the long-term success of a SCS system. Improper selection criteria were one of the main reasons for suboptimal results reported in the 1970s. During the 1970s and early 1980s, most studies evaluating the long-term efficacy of dorsal column stimulation quoted success rates of approximately 40%. Technical advances leading to improved hardware coupled with improved patient selection, have improved the rate of long-term efficacy to approximately 70%.^12,^92,⁹³

An SCS system should be considered for patients who have failed all reasonable medical rehabilitation and inteventional spine techniques.¹² An ideal patient should be motivated, compliant, and free of drug dependence.⁵⁶ Psychological screening is recommended but not mandatory to exclude conditions that predispose to failure of the procedure. It is extremely important that the patient and the physician have a discussion regarding probable outcome prior to implantation and that the patient has realistic expectations. If a patient has unrealistic expectations of the intervention, the treating physician should not consider that patient for implantation.

Diagnoses that are typical indications for this procedure in the United States include chronic radiculopathy, FBSS, neuropathic pain, and complex regional pain syndrome.^8,^16,^25,^43,⁵⁹ In Europe, SCS is also used for peripheral vascular disease and angina that are not amenable to medical therapy with reportedly excellent results.^6,^31,^36,^87,^97,¹⁰⁰ Other indications for SCS include transformed migraines, perineal pain, and interstitial cystitis.^4,⁸³

When considering pain topography, extremity pain responds better than axial pain, and the more distal the extremity pain the greater the clinical response.^73,¹⁰³ Middle and upper lumbar pain as well as thoracic, cervical, and chest wall pain are difficult to adequately control and maintain long term. Pain, due to severe nerve damage superimposed upon cutaneous numbness (i.e. anesthesia dolorosa) is also difficult to treat with SCS. Central pain syndromes do not respond to SCS and are best treated by other modalities.

The use of 3–7 day outpatient trials with an SCS system has proved helpful in determining which patients will respond well enough to warrant a permanent implantation.^18,^19,^73,¹⁰³ Absolute criteria that must be present for a patient to have a positive trial include tolerance of paresthesia, greater than 50% pain relief, and overall patient satisfaction. Relative requirements for a positive trial include improved functional level and reduced usage of pain medication. Long-term goals include reduced use of the medical system.

CLINICAL OUTCOMES

Original long-term results of pain control from spinal cord stimulation in the late 1960s and 1970s were disappointing.^28,^30,^32,^93,⁹⁶ This led to widespread disenchantment with SCS. Poor patient selection, inadequate equipment, and failure to perform implantations with the patient awake accounted for the disappointing results. The advent of new technology, careful patient selection, trial implantation, percutaneous placement or strict attention to positioning a paddle electrode in the same place as a percutaneously placed lead during a trial, and active physician–patient interaction during the procedure have all contributed to the success of spinal cord stimulation over the past 15 years.

In 1998, Kumar et al. published a case series of 235 patients who had undergone an SCS implantation.⁴⁴ The follow-up ranged from 6 months to 15 years after implantation with a mean of 5.6 years. One hundred and eighty-nine patients (80%) experienced satisfactory initial pain relief during a trial period of SCS and underwent a permanent SCS implant. The study population included 150 men and 85 women with a mean age of 51.4 years. Indications for SCS included FBSS (114 patients), peripheral vascular disease (39 patients), peripheral neuropathy (30 patients), MS (13 patients), RSD (13 patients) and other chronic pain etiology (26 patients). One hundred and eleven patients (59%) received satisfactory pain relief and 47 were gainfully employed (as compared with only 22 prior to implantation). Improvements in daily living as well as a decrease in analgesic usage were reported in successful outcomes. The shorter the duration of time to implantation after FBSS, the greater the rate of success. In the group with FBSS there were 101 successful stimulation trials and 13 unsuccessful stimulation trials. In this group, fifty-two patients received long-term pain relief while 49 patients who underwent permanent implantation were classified as failure. Out of the 235 patients who underwent permanent implantation, 189 had a successful trial SCS and 111 (58.7%) were successful long term.

In 2000, Villavicencio et al. published a retrospective nonrandomized case series comparing the long-term effectiveness of SCS using laminectomy-style electrodes versus percutaneously implanted electrodes.¹⁰⁵ The etiology of pain was FBSS (n=24), CRPS (n=7), neuropathic pain (n=4) and other (n=6). Mean follow-up after implantation of SCS was 8.6 months for the laminectomy-type electrode and 10.3 months for the percutaneously placed electrodes. Twenty-seven of 41 patients (66%) participating in the trial SCS implantation went on to receive a permanent SCS implantation. The laminectomy-style electrodes were placed under general anesthesia, while the percutaneous ones were placed under local anesthesia. Outcome was measured per VAS scale via phone interview. Visual analogue scale scores decreased by an average of 4.6 points after laminectomy-style electrode stimulation and by 3.1 after percutaneously implanted electrode stimulation. Electrodes placed through laminectomy provided significantly greater long-term pain relief than percutaneously placed ones (p=0.02). Overall, 89% of patients had greater than 50% pain relief. Five of 12 patients in the laminectomy group and seven of 15 patients in the percutaneous group underwent revision due to loss of pain relief. All 3 patients under the age of 65 in the laminectomy group were able to go back to work. Of the four patients working in the percutaneous group, two were able to go back to work, one retired, and one remained disabled due to pain.

In 2001, Leveque at al. published the results of a retrospective study involving 30 patients with FBSS who underwent permanent SCS.⁵³ Each patient had had multiple spinal surgeries and was deemed to no longer be a candidate for traditional spine surgery. Each patient had back and leg pain and all patients underwent a screening trial of SCS prior to permanent implantation. Sixteen patients had greater than 50% pain relief during the trial SCS and underwent permanent implantation. Of the 16 patients who underwent permanent implantation, 9 were implanted with percutaneously implanted catheter-style leads and the remaining 7 patients underwent permanent implantation with a paddle-style lead via a laminectomy. Mean follow-up was 34 months with a range of 6–66 months. At last follow-up, 12 of the 16 patients continued to have at least 50% pain relief. All 6 patients who underwent implantation via laminectomy continued to have greater than 50% pain relief with or without narcotics while only 6 out of the 9 percutaneously implanted patients had greater than 50% pain relief with or without narcotics. The authors concluded that SCS is an effective means of treating intractable back pain in the FBSS population. They also suggested that paddle-style SCS leads placed via laminectomy may be superior for long-term efficacy to catheter-style leads placed percutaneously.

In 2001, Ohnmeiss and Rashbaum published a retrospective study which evaluated patient satisfaction of SCS for predominant axial low back pain.⁷⁹ The literature prior to this study does not support the use of SCS for chronic spinal pain that is primarily limited to the low back. The patient sample included 41 patients who underwent SCS for axial low back pain of an average duration of 83 months. Inclusion criteria included failure of aggressive nonoperative care and patients were not considered surgical candidates. Thirty-eight out of forty-one patients had FBSS with back pain greater than leg pain. In all 41 patients, a SCS trial was performed for an average of 5.9 days. Five patients did not receive significant benefit from SCS trial and were not implanted. The remaining 36 patients were implanted with a Medtronics Mattrix system. Of the 36 patients, 4 later had the system removed due to lack of efficacy. The study included a mean follow-up of 10.5 months.

Seventy percent of patients implanted reported being satisfied, 78.8% of patients would recommend SCS to others, and 75.8% would have the implantation again. Sixty percent of patients reported improvement, 33% of patients reported no change in their pain. And 6.1% of patients reported being worse off after the implant than before.

The authors comment that much of the literature pertains to older devices no longer in use. They report a nonstatistically significant trend towards lower patient satisfaction with greater chronicity of symptoms pre-implantation. An interesting finding revealed in the questionnaire was that some patients dissatisfied with the results were still recommending the procedure to somebody else and would have the procedure done again. This may be due to overly inflated expectations of pain relief after SCS implantation.

In 2001, Barolat et al. published a retrospective review of 44 patients who had undergone SCS implantation with a paddle-style electrode and a radiofrequency receiver via a laminectomy for the treatment of intractable low back pain.⁹ Only patients in whom further lumbar surgery was not indicated or those in whom medical conditions contraindicated surgery underwent implantation. All patients had leg pain in addition to their back pain. The study was a multicenter, prospective study with follow-up at 6, 12, and 24 months. Follow-up data were available in 41 patients.

At 6 months, 91.6% of the patients reported fair to excellent pain relief in the lower extremities and 82.7% pain relief in the back. At 12 months, 88.2% of the patients reported fair to excellent pain relief in the lower extremities and 68.8% pain relief in the back. Significant improvement in function and quality of life was reported at both 6 and 12 months using the Oswestry and SIP, respectively. The majority of the patients report that the procedure was worthwhile and no patients considered the procedure not to be worthwhile. The authors concluded the SCS was beneficial in the treatment of intractable low back pain. The authors further opined that paddle-style electrodes may have an advantage over percutaneous catheter-style electrodes due to patient-reported broader coverage of stimulation and lower energy consumption.

In 2001, Dario et al. published a five year retrospective analysis (1992–1997) of 49 patients treated for FBSS at the authors’ center.²³ Twenty-one patients had predominantly leg pain, 22 had leg pain only, and six had back pain only. Outcome measures include change in VAS, Oswestry Scale, and Pain Disability Index as measured at entry and every three months for a period of 24–84 months with a mean duration of 42 months. The medical therapy was structured and all patients received the same protocol. After 6 months of medical therapy without significant benefit, patients underwent implantation of a Medtronic Itrel 2 or 3 SCS systems. Twenty-two patients had percutaneous leads implanted and 2 patients had laminectomy paddle-type leads implanted. The following measures were evaluated for SCS implanted patients: efficacy of therapy for leg and back pain, return to work status if previously employed, ADL activity if retired, and the concomitant use of medications and dosages after implant. The results revealed medical therapy improved both back and leg pain with mean VAS decreasing from 76 to 25. Oswestry scale results dropped from 23 to 6, and PDI scores went from 42 to 4. Of the 22 patients who underwent SCS implant, the VAS results were broken down into leg pain only and back pain only. In the leg pain only group, VAS went from 85 to 22, whereas in the back pain only group the VAS went from 45 to 40. The overall SCS groups mean PDI went from 51 to 7 and the Oswestry scoring went from 12 to 9. The authors concluded that neurostimulation only partially resolved the need for chronic medical therapy.

In 2002, North and Wetzel selected 26 patients with FBSS to undergo trial SCS.⁷⁶ Selection criteria included FBSS with leg pain greater than back pain and one or more of the following: recent abnormal diagnostic imaging results, neurological deficit consistent with the patient’s complaints and history, and/or a well-documented history of surgery for appropriate indications. All patients underwent a percutaneous trial with a four-electrode SCS lead. Of the 26 patients trialed, 24 were chosen to undergo a permanent implantation at the same spinal level as the trial lead was placed. Twelve patients were implanted with a percutaneous four-electrode lead similar to the one temporarily implanted for the trial stimulation and the other twelve patients were implanted with an insulated four-electrode lead via laminectomy. The study demonstrated that the laminectomy-placed leads outperformed percutaneously placed leads with better patient-reported pain coverage and lower amplitude requirements which should, according to calculations, yield a doubling of battery life.

In 2002, Kumar et al. evaluated the overall cost-effectiveness of SCS therapy versus conventional pain therapy for a consecutive series of 104 patients with FBSS.⁴⁵ Sixty patients with FBSS were treated with SCS versus 44 patients with conventional pain therapies (CPT). Outcome measures utilized included the Oswestry disability questionnaire at enrollment, yearly, and at 5-year follow-up. Cost of treatment was determined based on year 2000 Canadian dollars and expenditures were regulated by the Canadian Healthcare System. Treatment costs included primary care physician and specialist evaluations, imaging studies, surgery and devices for the interventional arm of the study and medication, physiotherapy, chiropractic, massage, and acupuncture for the control arm. The cumulative totals for the SCS arm were $29 123/patient compared to the control arm figure of $38 029/patient. The cost savings of SCS over the CPT group were realized at an average of 2.5 years due to the high initial cost of surgery and the device. After 2.5 years the savings of SCS versus CPT were realized through reduced drug use, reduced use of physical therapy services, and reduced use of the healthcare system in general. It is possible that in the US healthcare system the cost savings would be even greater. A prior study by Bell et al. puts the breakpoint at 2 years based on projected dollar values.¹³

In 2004, Turner et al. published a systematic review of the literature including the MEDLINE, EMBASE, Scientific Citation Index, Cochrane Controlled Trials Register, and Current Contents bibliographic databases back to their starting dates for articles published on the effectiveness of spinal cord stimulation (SCS) in treating failed back surgery syndrome (FBSS) and complex regional pain syndrome (CRPS) up to May 16, 2003.¹⁰⁴ Seven of 583 articles met the inclusion criteria for review of SCS effectiveness (two CRPS, three FBSS, two mixed diagnoses) and 15 others met the criteria for the review of SCS complications. The average number of previous spine surgeries (in the three studies that reported it) ranged from 2 to 3.3. Mean follow-up time was 33.6 months (range 6–60 months) and an average of 72% (range 58–96%) of patients received the permanent implantation after trial implantation. The studies included different SCS implantation methods (percutaneous trial followed by percutaneous implantation, percutaneous trial followed by implantation by laminectomy, and direct implantation by laminectomy without trial). Both FBSS studies (Dario ²³ and Ohnmeiss and Rashbaum⁷⁹) that reported leg and back pain separately found greater improvement in the leg pain than the back pain with SCS. Due to the study design of the seven publications, no conclusions could be drawn regarding effectiveness of SCS in returning patients to work. There was a suggestion that pain relief after SCS implantation decreases over time; however, no prospective study with a control group was available. Among 40 FBSS patients in an Ohnmeiss et al. study done in 1996, benefit from SCS decreased from 83% after 12 month to 70% after 24 months.⁷⁸ One large prospective multicenter study published in 1996 by Burchiel et al. with 182 FBSS patients who received a permanent SCS was excluded due to the fact that only 70 completed the follow-up.¹⁹ In this study, a one-year follow-up, showed statistically significant pain decrease and improvement in quality of life measures, but not in medication use or work status. However, due to the large number of patients lost in the follow-up, it is unclear if the patients at follow-up were representative of the original sample.

Also, in 2004, Mailis-Gagnon et al. performed a systematic review of the medical literature from 1980 to September 2003 regarding the efficacy of SCS in relieving certain types of chronic pain, its complications, and its adverse effects.⁵⁷ Studies that were only descriptive or observational were excluded. Primary outcome measure was pain relief. Secondary outcomes were indirect statements about pain relief, e.g. well-being, happiness, functional status, etc. A total of 1629 titles and abstracts were screened, 18 papers were retrieved, 16 of those were excluded, and two were included: one by Kemler et al.³⁸ and one by North et al.⁷⁴ Both studies were randomized, controlled trials but the Kemler et al. study was a truly prospective study while the North et al. study used a cross-over design. The Kemler et al. study evaluated CRPS type I and had 12-month follow-up while the North et al. study evaluated FBSS and had a 6-month cross-over between SCS and operation. Kemler et al. concluded that SCS is more effective and less expensive than standard treatments protocols for chronic CRPS and North et al. concluded that SCS should be considered prior to operation of the spine on a repeat disc herniation at a previously operated segment. The conclusion of Mailis-Gagnon et al. was that there is limited evidence in favor of SCS for FBSS and CRPS type I. Further, they felt that more trials were needed to confirm whether SCS is an effective treatment for certain types of chronic pain.

The most common SCS application in North America today is in the treatment of chronic low back and lower extremity pain due to chronic radiculopathy despite surgery.^19,^46,^78,^91,^95,¹⁰³ The largest SCS study incorporates 320 consecutive patients who underwent either temporary or permanent implants at the Johns Hopkins Hospital between 1971 and 1990.⁴⁰ This series included follow-up on 205 patients, the majority of whom had the diagnosis of failed back surgery syndrome. Permanent SCS implants were placed in 171 of these patients. At follow-up (mean interval 7.1±4.5 yrs), 52% of patients had at least 50% continued pain relief, and 58% had a reduction or elimination of analgesic intake. About 54% of patients younger than 65 were working at the time of follow-up; 41% had been working preoperatively.

The percentage of patients having long-term pain relief is similar in the majority of large published SCS series of implants for FBSS. The success rate in most of these studies, which is generally reported as 50% or more pain relief, is approximately 50–60%.^9,^20,^23,^43,^52,^53,^68,^79,^81,⁸⁵ Some studies report success rates as high as 88% and others as low as 37%.^39,⁹⁴ Although these latter studies differ in implantation technique and screening protocols, the success rate for pain reduction generally remains the same.

More recently published reviews have specifically looked at the efficacy of SCS in FBSS for pain control, reduction in narcotic consumption, function, and work status.^24,⁵⁰ ^69,^70,^84,¹⁰⁵ According to these studies, long-term pain reduction (at least 2 years after implantation) can be expected to range 50–70% in approximately 60% of SCS patients. In 50–90% of individuals, there will be an elimination or reduction in the use of opioids. The return to full employment rate after SCS reported by two studies is 25–59%, which is significant when comparing it to the usual return to work rate in this population of 1–5%.^50,⁷⁰ Reasons for the disparity between pain reduction and return-to-work rates appear to reflect the high percentage of unskilled laborers among this population, the prolonged periods of disability and the attendant socio-behavioral changes tend to take place in chronic pain patients. Despite this disparity, there is a general increase in function and activities of daily living.