Spinal Cord Stimulation


This chapter explores spinal cord stimulation (SCS) with thorough discussion of mechanisms, patient selection, indications, contraindications, perioperative considerations, equipment, programming, procedure, complications, and cost-effectiveness.

Three types of spinal cord stimulation are discussed: traditional or tonic stimulation, burst stimulation, and high frequency stimulation.

Conventional stimulation allows therapeutic paresthesias to overlap the distribution of the patient’s pain and thereby overwhelm the painful sensation. Burst stimulation utilizes unique waveform characteristics and provides minimal paresthesias to the patient. High frequency SCS (HF-10) is a paresthesia-free modality which utilizes high frequency (10 kHz), low amplitude pulses (1-5 mA) of short duration (30 μs) in a charged balanced, biphasic waveform. Based on animal models, it is hypothesized that HF-10 reduces wide dynamic range dorsal horn neuron activity, attenuates nociceptive pain signal transmission, and mitigates overall excitability. Burst stimulation and HF-10 both are claimed to provide superior pain control compared to conventional stimulation as well as relief of axial back pain which is traditionally hard to cover. The available clinical data and outcomes are reviewed for each type.

A set of recommendations are included for performing SCS trials and implants. The authors propose these based on the available literature, clinical experience, and recommendations from the Neuromodulation Appropriateness Consensus Committee (NACC).


burst stimulation, high-frequency stimulation, neuromodulation, neurostimulation clinical outcomes, spinal cord stimulation, tonic stimulation, trial and implant procedures



Chronic pain poses a serious burden in American healthcare that ranks among cancer, heart disease, and diabetes. An estimated 100 million adults suffer from persistent pain at any point in time, leading to striking estimates ranging from $100 billion to $300 billion in annual economic losses. As opioids forfeit their place in the evidence-based treatment algorithms for neuropathic pain, pharmacotherapies available to chronic pain providers and patients become increasingly limited, and interventional approaches are more important than ever. Neuromodulation represents an essential, evolving approach to treating the chronic pain patient.

Interestingly, the concept of alleviating pain with electrical stimulation is not new and dates back to 2500 bc , when the ancient Egyptians attempted to treat pain by applying electrogenic fish to painful sites. However, the first written medical reports of this practice came from the ancient Romans, who credited the black torpedo fish with relieving pains through electrical discharges.

In 1965 Melzack and Wall proposed the gate control theory of pain, which links the central and peripheral nervous systems in the complex perception and processing of pain. It hypothesizes that Aδ fibers, C fibers, and Aβ fibers all synapse at the dorsal horn, where they must wrestle to pass through a single-lane physiologic “gate.”

Aδ fibers and C fibers transmit pain signals from mechanical, thermal, or chemical nociceptive inputs. Aδ fibers are myelinated and medium diameter, whereas C fibers are unmyelinated and small diameter. Thus Aδ fibers transmit pain more quickly than C fibers (40 mph vs. 3 mph), creating two stages in the pain perception separated by a brief interval. The immediate, sharp pain felt by an individual after touching a hot stove is the result of Aδ-fiber firing, whereas the deeper throbbing pain that persists after the hand is withdrawn is attributed to C-fiber conductance. Meanwhile, Aβ fibers are involved in transmission of touch, pressure, and proprioception and do not conduct pain.

Gate theory explains why rubbing over the site of a soft tissue injury or an insect bite may alleviate pain. If the “gate” is overwhelmed with nonpainful input from the Aβ fibers, it should fail to transmit input from the nociceptive Aδ and C fibers. Although the aforementioned description may be oversimplified, Melzack and Wall’s theory laid the foundation for spinal cord stimulation (SCS).

In 1967, 2 years after their theory was published, Norman Shealy successfully implanted a monopolar SCS lead in the intrathecal space near the dorsal column. The patient’s impressive neuropathic pain improvement brought neuromodulation into the limelight, and the field has evolved significantly ever since.

Mechanism of Action

The mechanism of SCS is only partially understood at this time, and more research is needed. SCS is used for a variety of indications, ranging from neuropathic pain to angina, and it has the ability to impact multiple organ systems by inducing changes in either the local autonomic nervous system or viscerosomatic reflexes. It is likely that the mechanism of pain relief varies according to the pathology being treated. For instance, SCS may work differently to relieve neuropathic pain than it does for ischemic pain. In neuropathic pain the suppression of central excitability is likely the key. However, for ischemic pain, vasodilation and inhibition of sympathetic outflow are more likely factors.

Animal studies of SCS have allowed for some hypotheses on pathophysiology. Because neuropathic pain is the most common current indication for SCS, animal model findings regarding neuropathic pain and SCS are detailed next.

  • Hyperexcitability of dorsal horn wide dynamic range (WDR) neurons after nerve injury has been found to be due to dysfunction of the local spinal gamma-aminobutyric acid (GABA) system and increased release of excitatory neurotransmitters in the dorsal horn (i.e., glutamate).

  • SCS inhibits dorsal horn WDR neuron excitability, increases release of GABA, and decreases release of glutamate.

  • SCS increases concentrations of serotonin, norepinephrine, and adenosine in the dorsal horn.

  • Segmental hypothesis: SCS requires the participation of only a few spinal segments, with minimal additive inhibitory signals from the brain.

  • Supraspinal hypothesis: A supraspinal pathway is activated. Dorsal horn neurons are inhibited by descending antinociceptive pathways.

Patient Selection and Preoperative Considerations

Appropriate patients for neurostimulation implant must meet the following criterion: (1) the patient has a diagnosis amenable to this therapy, (2) the patient has failed conservative therapy for at least 6 months, (3) all significant psychological issues have been ruled out, (4) a history of illicit drug abuse is not present, and (5) a trial has demonstrated pain relief. However, pure neuropathic pain syndromes are relatively less common than the mixed nociceptive/neuropathic disorders, including failed back surgery syndrome (FBSS) ( Fig. 73.1 ). In addition, many patients with chronic pain will have some depressive symptomatology, and psychological screening can be extremely helpful to avoid implanting SCS in patients with major psychological disorders. An interesting study by Olson and colleagues revealed a high correlation between many items on a complex psychological testing battery and favorable response to a trial of SCS. It is recommended that extensive risk/benefit discussion and support for patients are ensured prior to implementing an invasive procedure.

FIG. 73.1

(A and B) Ideal candidates: failed back surgery syndrome/complex regional pain syndrome. Note the radicular versus axial pain pattern.

Courtesy Medtronic Inc.

To maximize the probability of improvements in functional outcomes as well as pain, part of the preoperative evaluation is establishment of functional goals. Doing so will reinforce to the patient that improvement in pain is not the primary endpoint, but rather it is the return of functional activity. Patients are often fearful that pain is indicative of damage, and this belief may limit their rehabilitation even after pain improves. Counseling on functional goals has the potential to question or even break a detrimental belief system.

Preoperatively, images should be reviewed carefully for optimal procedure planning. Consideration should be given to anatomic changes in the spine and variations. Epidural scarring, previous laminectomy, age-related changes, degenerated discs, and tight interlaminar spaces have the potential to vastly complicate the placement of percutaneous leads—and in some cases may even make it impossible. Severe spinal stenosis or the presence of surgical hardware may result in an obliterated epidural space, increasing the risk of complications from stimulator implant, including cord compression from an implanted lead.


SCS is a technically challenging interventional/surgical pain management technique. It merits extensive training and understanding of neuroanatomy, surgical technique, and perioperative patient care. Collaboration between the pain physician and spine surgeon is advocated for optimal success with neurostimulation. SCS involves the placement of platinum alloy electrodes into the posterior epidural space to electrically stimulate the dorsal columns of the spinal cord. It is important for the interventional pain physician to understand the equipment options available.

The SCS electrodes are of two types: percutaneous leads versus paddle electrodes ( Fig. 73.2 ). Percutaneous leads are flexible, cylindrical polyurethane catheters characterized at the distal tip by platinum alloy electrode contacts of variable numbers, lengths, and spacing patterns (depending on model and make). These leads are placed into the epidural space via needle insertion under fluoroscopy and are popularly placed by interventional pain physicians because they are minimally invasive when compared with surgical paddle leads. Percutaneous leads generally range from 8 to 16 contacts. Due to their cylindrical design, they generate circumferential current flow. This may cause stimulation of dorsal structures, such as the ligamentum flavum, and thereby lead to uncomfortable sensations.

FIG. 73.2

Neurostimulator leads: (left to right) percutaneous type to paddle type.

Courtesy SJM Inc.

The paddle leads are flat, wide, and rectangular at the distal end. They have insulation on one side and electrical pads on the other, with up to 32 electrode contacts. Current flow is unidirectional towards the cord. Therefore the electric field generated by paddle leads is more efficient than that of percutaneous leads, and paddle leads may avoid dorsal structure stimulation. However, these leads require a more invasive approach and are placed under direct visualization into the epidural space after a laminotomy or laminectomy by the spine surgeon. They have been linked in a large retrospective survey by Rosenow and colleagues to increased complications, fractures, and infections when compared with percutaneous leads. However, no adequate prospective studies comparing the percutaneous and paddle leads have been performed to date. Paddle leads may be chosen when anatomy, postsurgical changes, or epidural scarring make percutaneous lead placement difficult. In addition, paddle leads should be considered if lead revision is needed because of lead migration, if the patient experiences undesirable positional stimulation, or if more lead stability is required.

There are three types of power source options: primary cell implantable pulse generators (IPGs), rechargeable IPGs, and radiofrequency (RF) units ( Fig. 73.3 ). RF units, the first of implantable SCS devices, are not limited by battery life but require an external power source. The patient must tape an antenna to the skin on top of the device receiver. This is inconvenient and may result in skin irritation. Thus RF units are rarely used nowadays. Primary cell IPGs were introduced in 1980, based on pacemaker technology. Although they tend to be larger and have a shorter life span of 3–4 years, they are low maintenance because they do not require charging or interfere with daily activities. Rechargeable IPGs have been introduced and contain lithium ion cells with a life span of 9 years.

FIG. 73.3

(A) Schematic view of an implanted pulse generator system.

Courtesy Medtronic Inc.) (B) Schematic view of an implanted radiofrequency spinal cord stimulation system. (Courtesy SJM Inc.) (C) Representative implanted pulse generator neurostimulation units with leads. (Courtesy SJM Inc.


An SCS trial is conducted using fluoroscopy, under sterile conditions, and typically one or two leads are placed percutaneously. Prophylactic antibiotics are advocated for infection prevention, and the current recommendation is a cephalosporin, such as cefazolin. Clindamycin is recommended when the patient is allergic to β-lactams, and vancomycin is the agent of choice if the patient is methicillin-resistant Staphylococcus aureus (MRSA) positive. After a 14-gauge 9-inch Tuohy introducer needle, supplied with the kit, is percutaneously placed in the dorsal epidural space, using the standard loss of resistance technique, the SCS lead is introduced into the epidural space through the epidural needle ( Fig. 73.4 ). Once in the epidural space, the SCS lead is steered under fluoroscopic guidance into the posterior paramedian epidural space up to the desired anatomic location. Although two SCS leads, juxtaposed in the posterior epidural space, are most common and give superior stimulation options, single lead placement may suffice if adequate coverage of the painful area is achieved. For traditional SCS (excluding high-frequency SCS), short-acting sedatives are preferred, so that the patient can be readily awakened after the lead placement for evaluation of paresthesia coverage over the area of pain. After the successful epidural lead placement, the epidural needle is carefully withdrawn avoiding any lead migration. After the final lead position is confirmed by fluoroscopy, the lead is firmly anchored to the skin using nonabsorbable sutures and/or proprietary anchoring device and a sterile dressing is applied.

FIG. 73.4

(A) Percutaneous lead placement: marking the interspinous level. (B) Percutaneous lead insertion. (C) Dual lead trial.

Courtesy Medtronic Inc.

When the patient returns after a trial of percutaneously placed leads, the dressing is removed, the sutures are clipped, and the leads are removed and discarded. If the trial is successful, new leads are placed and connected to an implanted IPG when the patient returns for the SCS implantation.

Alternatively, SCS trial leads can be placed through an open wound, over the dorsal spine, using the technique similar to percutaneous lead placement. However, while the original SCS leads are buried subcutaneously, only the tunneled extensions exit the skin. In this circumstance, if the trial is successful, during permanent implantation, only the extensions are discarded and the original trial leads are connected to the implanted generator. This method has the advantage of retaining the same lead position in a successful trial, but it adds an incision at the time of the initial trial, which increases postoperative pain confounding the trial interpretation. Furthermore, trial leads implanted through an open incision may have greater risk for infection than the straight percutaneous method.

A careful trial period of 5–7 days is advocated to decrease the risk of a failed implant. Patients are encouraged to pursue normal activities, with the exception of aggressive bending or twisting to prevent lead migration. Despite careful patient selection and introduction of improved and redundant multilead systems, which gives a host of varied stimulation options, clinical failures of SCS trials remain all too common. Prudent pain practitioners therefore must critically evaluate the trial outcomes and adhere to strict selection criteria. Even though most practitioners consider 50% or more of pain relief as indicative of a successful trial, other factors, such as functional improvement, increased activity level, and reduced analgesic intake should also be taken into account.

If the trial succeeds, the patient returns for permanent lead and generator placement. The technical challenges of permanent lead placement depend on (1) proper fixation and (2) lead redundancy. Consistent and reliable stimulation depends on applying an electric field over a small area of the spinal cord. SCS leads have a limited capacity for stretch, and certain body movements can stretch leads significantly and prompt lead migration. Although sclerotic changes in the epidural tissue surrounding the implanted leads stabilize them in the long run, during the acute phase, proper anchoring is a major factor in successful lead placement ( Fig. 73.5 ). In the event of minor lead migration, electrode redundancy is used to accommodate for minor shifts by using alternative electrodes to accommodate the desired electric field.

FIG. 73.5

Anchoring the lead.

Courtesy Medtronic Inc.

The generator unit is generally implanted in the lower abdominal area or in the posterior superior gluteal area ( Fig. 73.6 ). For cervical or occipital leads, generators are often placed in between the scapula. In general, generators should be in a location the patient can access with his or her dominant hand for adjustment of settings or charging.

FIG. 73.6

(A and B) Permanent implant: pulse generator internalization.

Courtesy Medtronic Inc.

Placement of leads in the cervical spine should account for the fact that the spinal cord in the mid-to-lower cervical spine has high mobility, making it difficult to maintain constant paresthesias with tonic stimulation as the neck changes position. This increased mobility also increases lead migration risk in the cervical region. In general, cervical lead placement involves entry into the epidural space between T1 and T4, 11 with percutaneous leads advanced to the appropriate level in the cervical spine.

In the thoracic region the spine is more fixed and immobile. This allows less vulnerability to lead migration. At the T5 level the CSF diameter is largest dorsally, and the spinal cord is smallest. Thus stimulation thresholds are higher and postural changes are problematic when leads are placed at this level.

Leads may be placed in the lumbar spine, and electrical currents directed to conus medullaris or cauda equina, when attempting to provide relief for pelvic, sacral, or foot pains. With spinal nerves free-floating at these vertebral levels, stimulation is vulnerable to patient movement and can be suboptimal.


Complications with SCS range from minor problems, such as lack of appropriate paresthesia coverage, to devastating complications, such as paralysis, nerve injury, and death. Overall complication rates from SCS range from 28% to 42%. In a systemic review by Cameron the most common complication was found to be lead migration or breakage, which occurred in 22% of implanted cases. Studies by Barolat and May reported lead revision rates due to lead migration of 4.5% and 13.6% and breakage of 0% and 13.6%, respectively. The generator can also be a source of revision if changes in body habitus affect the source position.

Studies have demonstrated superficial infectious rates ranging from 2.5% to 7.5%, but fortunately only rarely these progressed to more serious infections (<0.1%). To avoid infectious complications, proper sterility should be observed in the operating environment. In addition, the patient should be instructed on wound care and on recognition of signs and symptoms of infection. Many superficial infections can be treated with oral antibiotics, but more serious infection may require surgical exploration and removal of the device. Although less common, abscess of the epidural space can lead to paralysis and death if not identified quickly, so a high index of suspicion is warranted. Staphylococcus aureus is the predominant pathogen for percutaneous epidural catheter–related infections.

Programming and Technical Overview

A full understanding of SCS requires appreciation of Ohm’s law: I = V / R, current (I) equals voltage (V) divided by resistance (R). In SCS circuit, voltage is supplied by the IPG and resistance is the impedance between the lead electrodes (determined by the tissues surrounding the electrodes including the dorsal columns), with these two variables determining the flow of current between the SCS electrodes. Factors such as epidural fibrosis, a large epidural space or small spinal cord diameter at any particular level, and a thick CSF layer may increase resistance or impedance to current flow between the electrodes, requiring higher voltages by the IPG to maintain adequate stimulation. An understanding of the anatomy of the spinal cord and particularly of the size/shape of the epidural space at varying levels is therefore important. Studies confirm that a lead placed in the cervical spine, where the spinal cord is larger and the epidural space narrow, has lower impedance than a lead placed in the lower thoracic spine, where the epidural space is large and the spinal cord diameter is small. Increased resistance will require higher voltage to maintain optimal current flow. As power is directly proportional to voltage and current, appropriately manipulating these variables may allow for better power performance. To date, the following factors have not been correlated with changes in impedance: age, spine surgery, number of leads, or model. However, males were found to have 21% higher impedance than women in the thoracic spine, without clear physiologic explanation.

There are four basic parameters in traditional tonic neurostimulation, which may be adjusted to create stimulation paresthesia in the painful areas, thereby mitigating the patient’s pain ( Fig. 73.7 ). They are amplitude, pulse width, rate, and electrode selection.

FIG. 73.7

Typical patterns of coverage using different anodal and cathode combinations.

Courtesy Medtronic Inc.

Amplitude is the intensity or strength of each individual pulse and can be controlled by voltage or current. There is no evidence of superiority for voltage or current control. However, current-control systems are theoretically more immune to changes in electrical resistance in the tissue due to sclerosis and patient positional changes. As such, mathematical modeling predicts more even paresthesia. Amplitudes are variably optimized for each individual patient, but typical initial settings are 60%–90% of motor threshold. Pulse width is the duration of a pulse measured in microseconds (μs). It is usually set between 100 and 400 msec. Similar to increasing the amplitude, a larger pulse width delivers more energy per pulse and typically provides broader coverage. Common initial settings are 0.2 msec. Rate is measured in hertz (Hz) or cycles per second, between 20 and 150 Hz. At lower rates, the patient may feel myoclonic vibrations, whereas at higher frequencies, the feeling is more of a buzzing sensation. Higher frequencies (>500 Hz) are suggested to increase blood flow and decrease vascular resistance.

Targets for lead placement is a complex topic. Barolat and colleagues mapped data of sensory response and coverage patterns based on lead location in the epidural space in 106 patients. Most patients’ stimulators are programmed with electrode selection such that the patient obtains anatomic coverage, and then the pulse width and rate are adjusted for maximal comfort.

Placement of the electrodes for a few important and frequently used targets includes the following:

  • C2: lower half of face

  • C2-C4: neck, shoulder, hand

  • T5-T6: abdomen

  • T7-T9: back

  • T10-T10: leg

  • T12-L1: foot

  • L1: pelvis

The lowest acceptable settings on all parameters are generally used to conserve battery life. Other programming modes that save battery life include a cycling mode during which the stimulator cycles full on/off at patient-determined intervals (minutes, seconds, or hours). The patients’ programming may change over time and reprogramming needs are common.

Scs Types: Traditional, High-Frequency, and Burst Scs

Traditional Scs

Thus far we have primarily addressed traditional SCS. Conventional SCS, in line with the gate theory, allows agreeable, therapeutic paresthesia to overlap the distribution of the patient’s pain and thereby overwhelms the painful sensation. The IPG delivers stimuli by activating a constant voltage or constant current system and generally uses frequencies between 50 and 150 Hz. The effective amplitude, which impacts the intensity of the paresthesia, is adjusted for the individual patient. This method mimics tonic firing patterns in the brain and is referred to as tonic stimulation. Source-localized electroencephalogram (EEG) data indicate that tonic SCS stimulation mainly activates the lateral pain pathway, which accounts for the sensory-discriminatory components of pain.

Burst Scs

Burst SCS uses unique waveform characteristics and offers minimal paresthesia to the patient. Burst stimulation delivers a train of five spikes at 500 Hz per spike, 40 times per second, with a pulse width of 1 millisecond and constant current. This form of SCS mimics burst firing in the thalamic cells of the nervous system, and based on functional magnetic resonance imaging (fMRI) and source-localized EEG data, it is hypothesized to stimulate both the medial and lateral pain pathways. The lateral pain pathway (affected in tonic and burst modes) is thought to be responsible for the discriminative component of pain, whereas the medial pathway (affected in burst mode) is thought to be responsible for the affective-motivational component of pain. Despite the lack of paresthesia, burst SCS still necessitates the placement of the SCS leads in the appropriate location in the epidural space based on Borolat’s mapping of sensory responses to electrical stimulation at varying spinal cord levels. Burst SCS has been claimed to provide superior pain control, as well as relief of axial back pain, which is traditionally hard to cover. It has been proposed to offer a rescue option to patients who fail tonic stimulation.

High-Frequency Scs

High-frequency SCS (HF-10) is a paresthesia-free SCS modality used for the relief of axial and radicular pain. HF-10 uses high-frequency (10 kHz), low-amplitude pulses (1–5 mA) of short duration (30 msec) in a charged balanced, biphasic waveform. These pulses are applied via lead contacts positioned in the T8-T11 epidural space, most commonly with two stacked leads. The exact mechanism of pain relief with HF-10 is not entirely clear; however, based on rat and goat models, it is hypothesized that HF-10 reduces WDR dorsal horn neuron activity, attenuates nociceptive pain signal transmission, and mitigates overall excitability. A study by Shechter and colleagues in a rat model supported frequency- and intensity-dependent analgesia and showed that high-frequency SCS allowed for greater inhibition of neuropathic mechanical hypersensitivity than traditional SCS. HF-10 does not require the overlapping of paresthesia with painful areas; thus it eliminates the need for intraoperative paresthesia mapping and allows for deep sedation or even general anesthesia for the procedure. It is hypothesized that HF-10 is more forgiving to lead migration when compared with tonic stimulation because precision lead placement in the epidural space is not necessary. The change in the intensity of paresthesia with change in body position and the subsequent need for adjustment of stimulation parameters by the patient, characteristic of tonic stimulation, is also eliminated with HF-10. This may prove key to patient satisfaction and adequate sleep. High-frequency SCS has been claimed to provide superior relief of axial pain, and it provides an alternative for patients who experience painful or uncomfortable stimulation with conventional SCS.

Dorsal Root Ganglion Stimulation

Dorsal root ganglion stimulation is a new avenue for more selective stimulation. Because another chapter is dedicated to this topic ( Chapter 75 ), this will not be discussed further here.

Outcomes: Clinical Studies of Traditional Scs

Outcomes research for SCS is evolving. Greater research is needed to refine the technology, pursue increased efficacy, and limit complications. Much of the initial evidence is observational in nature or are comparative effectiveness studies. This is due to the invasive nature of the modality and due to the fact that blinded or placebo treatment, especially in the case of traditional tonic stimulation with the necessity for paresthesia, is difficult. Important studies on outcomes of traditional SCS are detailed later and in Table 73.1 .

TABLE 73.1

Clinical Outcomes Studies on Tonic Spinal Cord Stimulation

Study Design Details Significant Results Analysis
Kemler and colleagues (2004)

  • n = 54

    • Cohort: CRPS

    • Randomized 2:1 to SCS + PT or PT only

  • RCT

  • Prospective

  • 2-year follow-up

In SCS + PT group:

  • VAS ↓ by 2.1 (vs. no ↓ in PT only)

  • Patient satisfaction higher

  • SCS for CRPS may allow for better pain reduction when combined with PT

  • 38% complications

May reflect older

  • Not blinded

North and colleagues (2005)

  • n = 50

    • Cohort: FBSS

    • Radicular pain

    • n = 45 follow-up

    • SCS vs. reoperation

  • RCT, optional crossover

  • Prospective

  • 3-year mean follow-up

  • Pain relief > 50%:

    • 9/19 in SCS group

    • 3/26 in reoperation group

    • SCS better than reoperation ( P < .01)

  • More opioid use in reoperation group vs. SCS group ( P < .025)

  • Patients less likely to crossover from SCS group ( P < .02)

  • SCS had better outcomes than reoperation for FBSS


  • not ITT analysis (crossovers counted as outcome measures).

  • some reoperation patients switched to SCS despite meeting success criteria

  • not blinded

Kumar and colleagues (2008)

  • n = 100

    • Cohort: FBSS

    • Neuropathic pain

  • SCS + CMM vs. CMM only

  • RCT

  • Multicenter

  • Prospective

  • 2-year follow-up

  • 1-year results published separately (detailed in text)

  • At 2 years ≥50% pain:

    • Modified ITT: 37% SCS group vs. 2% CMM ( P = .003)

    • As treated analysis: 47% SCS group vs. 7% CMM ( P = .02)

    • SCS group: Better secondary outcomes based on ODI, QOL, patient satisfaction

    • No difference in analgesic intake

  • SCS + CMM was more effective than CMM alone for FBSS


  • Not blinded

  • Comparison group not truly placebo controlled

CMM, Conventional medical management; CRPS, complex regional pain syndrome; FBSS, failed back surgery syndrome; ITT, intention to treat; ODI, Oswestry Disability Index; PT, physical therapy; QOL, quality of life; RCT, randomized controlled trial; SCS, spinal cord stimulation; VAS, visual analog scale.

Failed Back Surgery Syndrome

There are two notable randomized controlled trials (RCTs) on SCS for FBSS. North and colleagues selected 50 FBSS patients who were randomized to either repeat laminectomy or SCS, and crossover between the groups was permitted after 6 months. Of the 26 patients who had undergone reoperation, 54% (14 patients) crossed over to SCS, and the 24 patients who had undergone SCS, 21% (5 patients) opted for crossover to reoperation. At average of 3 years, 90% of the patients were available for long-term follow-up and showed that, using standard measures, SCS was significantly more effective than reoperation.

The second RCT was a multicenter international study that randomized 100 FBSS patients with neuropathic radicular pain to SCS plus conventional medical management (CMM) (SCS group) or CMM only for 6 months. Primary outcome was 50% or greater reduction in pain, and secondary outcomes included quality of life indicators, functional capacity, pain medication use, satisfaction, and complications. Crossover was permitted at the 6-month interval, with an intention-to-treat model, and patients were followed for an entire year. The results showed a statistically significant advantage of SCS and CMM over CMM only for the primary ( P < .001) and secondary ( P ≤ .05%) outcomes, except for decrease in pain medication use, which trended downward but was not statistically significant. After the study midpoint, 5 of 50 SCS patients crossed over to CMM versus 32 of 50 CMM to SCS. However, at the study conclusion, 32% of SCS patients experienced device-related complications. The 2-year follow-up study results were published separately in 2008 and provided similar results. Forty-six patients randomized to SCS and 41 patients randomized to CMM were available for follow-up.

There are three valuable systematic reviews on neurostimulation for chronic pain of spinal origin. Turner and colleagues completed a meta-analysis of articles on treatment of FBSS by SCS published from 1966 to 1994. Even though pain relief exceeding 50% was experienced by 59% of the patients, the authors concluded that there was insufficient evidence of effectiveness of SCS relative to no treatment or other treatments. The review by North and Wetzel consisted of case-control studies and two prospective control studies. It concluded that patients reporting a reduction in pain of at least 50% during a trial and demonstrating improved or stable analgesic requirements and activity levels had significant benefit from permanent SCS.

Complex Regional Pain Syndrome

High-quality research regarding SCS and complex regional pain syndrome (CRPS) is limited, but existing data are overwhelmingly positive in terms of pain reduction, quality of life, analgesic use, and function.

The International Association for the Study of Pain (IASP) has recommended that SCS be implemented within 12–16 weeks when conservative therapy does not provide successful outcomes. The CRPS-I Task Force of the Dutch Society of Rehabilitation Specialists and the Dutch Society of Anesthesiologists was established in 2010 to provide evidence-based guidelines on CRPS-I treatment. After examining the available literature, the group concluded that SCS for CRPS-I led to long-term pain relief and improved quality of life but no improvement in function. They indicated that SCS was a reasonable treatment option for carefully selected CRPS-I patients who were refractory to other treatment modalities.

Kemler and colleagues published a randomized comparative trial of SCS versus conservative therapy for CRPS. Patients with a 6-month history of CRPS of the upper extremities were randomized to either SCS plus physiotherapy versus physiotherapy alone. At a 2-year follow-up, the patients in the SCS group had a significantly greater reduction in pain and had higher satisfaction. The authors concluded that in the short term, SCS reduces pain and improves the quality of life for patients with CRPS involving the upper extremities. However, the 5-year follow-up data showed that the effect of SCS diminished over time.

Several case series have been published on the use of neurostimulation in the treatment of CRPS. Calvillo reported on a series of patients with advanced CRPS treated with either SCS, peripheral nerve stimulator (PNS), or both. After a 3-year follow-up, patients with SCS had a statistically significant reduction in pain scores and improvement in return to work. The authors concluded that in late stages of CRPS, neurostimulation (with SCS or PNS) is a reasonable option when alternative therapies have failed.

Another case series reported by Oakley evaluated treatment response to SCS in CRPS patients. The study followed 19 patients and analyzed the results using a sophisticated battery of outcomes tools (McGill Pain Rating Index, the Sickness Impact Profile, Oswestry Disability Profile, Beck Depression Inventory, and visual analog scale [VAS]). After an average follow-up of 8 months, all scales showed statistical benefits of SCS and all patients received at least partial pain relief, with 30% receiving significant pain relief. A prospective, observational case series by Geurts and colleagues published in 2013 followed 79 CRPS-I patients who had SCS placed during the 11-year time period from 1997 to 2008. The authors found that 41% of patients experienced 30% or greater pain relief.

Two long-term follow-up studies are available. In 2011 Kumar and colleagues published results (mean follow-up of 7 years) for 25 CRPS-I patients treated with cervical SCS. They noted decreased need for analgesics and significant improvements in pain score, functional status, and quality of life. They found the highest success with patients under 40 years of age and when implant was within the first year of onset of disease. Also in 2011 Sears and colleagues published results of 18 patients with CRPS who had SCS using implanted paddle electrodes. They reported significant pain reduction and high patient satisfaction at 5-year follow-up. However, the efficacy of the SCS did slightly decrease over time.

A literature review by Stanton-Hicks of SCS for CRPS consisted of seven case series. These studies ranged in size from 6 to 24 patients. Results were noted as “good to excellent” in greater than 72% of patients over a time period of 8–40 months. The review concluded that SCS was a powerful tool in the management of patients with CRPS.

Peripheral Ischemia and Angina

Cook reported in 1976 that SCS successfully relieved pain associated with peripheral ischemia. The results have been confirmed and noted to have particular efficacy in conditions associated with vasospasm, such as Raynaud disease. Studies have also shown efficacy of SCS in treating intractable angina with success rates greater than 80% in some of the studies. The use of SCS for these indications is widespread outside of the United States. This active area of research has an expanding body of literature, and interested readers are encouraged to evaluate the literature themselves because it is beyond the scope of this chapter.

Clinical Studies of Burst Scs

The literature for burst SCS is evolving because the burst mode of SCS allows for placebo-controlled, double-blinded trials more readily than traditional SCS due to the lack of paresthesia. Important studies are detailed next and in Table 73.2 .

TABLE 73.2

Clinical Outcome Studies on Burst Spinal Cord Stimulation

Study Design Details Findings Analysis
De Ridder and colleagues (2014)

  • n = 102.

    • Cohort: FBSS/PDN

    • tSCS × 6 months before study

    • 2 groups: responders vs. nonresponders to tSCS

    • All switched to burst

  • Retrospective

  • Multicenter

  • tSCS vs. burst tested

  • 94.8% responders to tSCS had improved pain relief with burst

    • Axial: burst 29% better than tSCS

    • Limb: burst 31% better than tSCS

  • 62% nonresponders to tSCS “rescued” with burst

  • Burst may salvage nonresponders to tSCS

  • Burst provided greater relief than tSCS in both tSCS responders and nonresponders


  • No control group

  • Follow-up only 2 weeks

Schu and colleagues (2014)

  • n = 20

    • Cohort: FBSS

    • All used tSCS × 3 m

  • Burst vs. tonic vs. placebo

  • Patients received each mode × 1 week each

  • Prospective, randomized, double-blind, placebo-controlled

  • 80% preferred burst

  • Burst ↓ NRS pain intensity score ( P < .05) compared to other modes NRS:

    • Burst: 4.7 ± 2.5

    • Tonic: 5.6 ± 1.7

    • Placebo: 8.3 ± 1.1

  • Burst pain ↓ SFMPQ scores ( P < .05) indicating improved pain quality

    • Burst: 19.5 ± 10.5

    • Tonic: 25.0 ± 7.1

    • Placebo: 33.5 ± 11.8


  • Burst improved pain intensity and pain quality scores compared to tonic and placebo in FBSS

  • Limitations:

  • Each mode only evaluated for 1 week

  • Possible carry-over effects

  • Small scale study

de Vos and colleagues (2014)

  • n = 48

    • Cohort: Neuropathic pain (FBSS, PDN)

    • All used tSCS for ≥6 months

    • 3 groups: FBSS with good response to tSCS = 24, FBSS poor response to tSCS (PR) = 12, PDN = 12

  • Burst for 2 weeks

  • Prospective

  • Burst VAS ↓ from baseline:

    • 77% in PDN, 57% in FBSS, 23% in PR (all p values < 0.05)

  • Burst vs. tSCS:

    • Burst: 44% more relief in PDN patients ( P < .001) and 28% more in FBSS ( P < .01)

    • No significant difference in poor responders

60% of tSCS patients experienced better pain relief with burst SCS

  • Only 2 weeks follow-up

  • Possible carry-over effects

  • Not randomized, placebo-controlled, double-blind study

De Ridder and colleagues (2013)

  • n = 15

    • Cohort: Limb and axial back pain

  • Burst vs. tSCS vs. placebo

  • Patients had each mode × 1 week each

  • VAS score measurement

  • Prospective, randomized, double-blind, placebo controlled

  • All patients preferred burst

  • Pain relief:

  • Burst vs. placebo, VAS ↓ by

    • Axial: 51.3% vs. 18.9%

    • Limb: 52.7% vs. 11.7%

    • General: 55.0% vs. 10.9%

    • All P < .05

  • Burst vs. tSCS, VAS ↓ by

  • Axial: 51.3% vs. 30.3%

  • General: 55.0% vs. 30.9%

  • Both P < .01

  • Paresthesia:

  • tSCS—more paresthesia than burst and placebo ( P < .05).

  • No difference in paresthesia between placebo and burst

  • Burst SCS provided better pain relief for back and general pain than placebo and tonic, with reduced paresthesia


  • Each mode only evaluated for 1 week

  • Possible carry-over effect

  • Small sample size

  • Sample with skewed gender demographic

De Ridder and colleagues (2010)

  • n = 12.

    • Cohort: Neuropathic pain (11/12 FBSS, 1/12 PDN).

  • Prospective.

  • Paddle leads.

  • tSCS vs. burst tested with VAS during trial

  • Implanted with mode of choice and followed 1 year

  • All patients preferred burst

  • Pain relief at trial sessions

    • Axial pain: VAS ↓ by

    • tSCS: 1.83 pts ( P = .05)

    • Burst: 5.25 pts ( P < .001)

    • Limb pain: VAS ↓ by

    • tSCS: 4.41 pts ( P < .001)

    • Burst: 7.29 pts ( P < .001)

  • At 1 year with burst: significant pain relief (axial 58.7%, limb 70.2%)

  • Paresthesia: 17% burst, 91% tonic

  • Burst had better pain suppression and less paresthesia than tonic


  • Not a placebo controlled, double-blind study

  • Small sample size

FBSS, Failed back surgery syndrome; NRS, numeric rating scale; PDN, painful diabetic neuropathy; PR, poor responders; SFMPQ, short form McGill pain questionnaire; tSCS, tonic spinal cord stimulation; VAS, visual analog scale.

In 2010 De Ridder and colleagues published prospective data on 12 patients with neuropathic pain, the majority having FBSS. The patients were trialed in random order with two, 1-hour sessions in tonic mode and an equal number/length of sessions in burst mode on a separate day. All patients preferred the burst mode. The patients reported significantly better suppression of axial and limb pain with burst mode than tonic mode based on VAS. Additionally, only 17% of the patients experienced paresthesia with burst mode of SCS compared to 91% of the patients with tonic SCS. At 1-year follow-up, the patients still experienced significant axial and limb pain reduction with the burst mode. While this study had a small sample size and was not a placebo controlled, double-blinded trial, it highlighted the potential for paresthesia-free stimulation with burst SCS as well as the potential for axial pain relief.

In 2013, De Ridder and colleagues performed a prospective, randomized, double-blinded, placebo controlled study in 15 patients with mixed limb and axial back pain. Burst SCS was studied against tonic SCS and placebo. Patients received each mode for 1 week, in a randomized sequence and VAS scores were measured. The authors showed that burst SCS was superior to placebo in decreasing VAS scores for axial, limb, and general pain. It was also superior to tonic SCS in suppressing axial and general pain, but there was no significant difference in limb pain. Given that each mode was evaluated for only 1 week, the conclusions are restricted to short-term evaluation. The study was also limited by the small sample size and a skewed gender demographic because the sample consisted of 4 men and 11 women.

In 2014 de Vos and colleagues prospectively studied 48 patients with neuropathic pain secondary to FBSS or painful diabetic neuropathy (PDN). All had used tonic SCS for at least 6 months before the study. Three groups were created: (1) FBSS with good response to tonic SCS, 24 patients, (2) FBSS with poor response to tonic SCS, 12 patients, and (3) PDN, 12 patients. All groups were switched to burst mode for 2 weeks. Overall, 60% of the patients experienced better pain relief with burst SCS as compared with tonic SCS. Subgroup analysis showed that burst SCS allowed 44% more pain relief in PDN patients and 28% better relief in FBSS. The study was not a randomized, placebo-controlled, double-blind study. In addition, it was limited by only 2 weeks of follow-up, failure of the cohort to isolate to one neuropathic pain state, and questionable washout interval between modes, potentially allowing for carry-over effects.

Also in 2014 Schu and colleagues performed a study that was similar in design to De Ridder’s 2013 study but looked at FBSS patients who already had received tonic SCS stimulation for 3 months. This was a prospective, randomized, double-blind, placebo-controlled study and tested burst SCS against tonic SCS and placebo. The patients received each mode for 1 week, in a randomized sequence. Numeric Rating Scale (NRS) pain intensity scores were measured, as were short-form McGill pain questionnaire (SF-MPQ) pain quality scores. Burst SCS significantly improved NRS and SF-MPQ scores compared with tonic SCS and placebo in FBSS patients who were already using tonic SCS. However, it should be noted that each mode was only evaluated for 1 week, possible carry-over effects existed from the baseline tonic stimulation mode, and the study was small in scale.

Finally, De Ridder and colleagues published a retrospective, multicenter study on 102 patients with FBSS or PDN who had all used tonic SCS for at least 6 months prior to the study. The patients were categorized into responders to tonic SCS group and nonresponders to tonic SCS group. Both groups were switched to burst mode SCS. Interestingly, 94.8% of responders to tonic SCS had improved pain suppression with burst stimulation. In addition, 62% of the nonresponders to tonic SCS were “rescued” with burst mode SCS. Thus it was concluded that (1) burst SCS may salvage nonresponders to tonic SCS and that (2) burst SCS may provide greater relief than tonic SCS in both tonic responders and nonresponders. This study was limited due to its retrospective nature, the lack of control group, and short-term 2-week follow-up.

Clinical Studies of High-Frequency Scs

Notable studies for high-frequency SCS are discussed next and summarized in Table 73.3 . Most recently, in 2016, Kapural and colleagues published a multicenter, prospective, randomized, controlled trial of 198 patients with back and leg pain who were naïve to SCS. The patients were randomly assigned to receive either tonic SCS or HF-10 in a 1:1 ratio, and success was defined as ≥50% axial pain reduction. At 2 years, the HF-10 SCS suppressed back and leg pain in more subjects than tonic SCS did. Limitations of the study included the lack of a placebo group and its nonblinded nature. This would have been difficult to accomplish with only one mode of SCS generating paresthesia.

TABLE 73.3

Clinical Outcome Studies on High-Frequency Stimulation

Study Design Details Findings Analysis
Kapural and colleagues (2016)

  • n = 198

    • Cohort: back ± leg pain

    • n = 171 successful trial and implant

    • Naive to SCS

  • Multicenter, prospective, randomized, controlled trial

  • Follow-up: 24 months

  • Responders—HF vs. tSCS:

    • leg pain: 76.5% vs. 49.3%

    • back pain: 72.9% vs. 49.3%

    • P < .001

  • Degree of response—HF vs. tSCS:

    • back pain: 66.9% ± 31.8% vs. 41.1% ± 36.8%

    • leg pain: 65.1% ± 36% vs. 46% ± 40.4%

    • P < .001

  • HF at 10 kHz better suppressed back and leg pain for more subjects than tonic SCS over long term.


  • Not blinded

  • No placebo group

Tiede and colleagues (2013)

  • n = 24

    • Cohort: back pain > leg pain

  • Multicenter case series

  • Prospective, open-label

  • Trialed with tSCS (4 days) and then with HF (4 days)

  • 88% preferred HF to tSCS

  • 77% reduction of VAS for back pain and overall pain from baseline with HF ( P < .001)

  • HF at 10 kHz may be preferred by patients and effective for axial pain


  • Not randomized or controlled

  • No placebo arm

  • Short duration

  • Small sample

Van Buyten and colleagues (2013)

  • n = 83

    • Cohort: low back pain (81% FBSS)

    • n = 72 for implant

  • Prospective, multicenter, open-label, observational

  • Follow-up: 6 months

  • VAS ↓ from baseline at 6 months:

    • Back pain: 8.4–2.7

    • Limb pain: 5.4–1.4

    • P < .001 (both)

  • ODI ↓ by 17 points from baseline at 6 months ( P < .001)

  • At 6 months, 75% with >50% pain relief

  • 11/14 who failed tSCS were rescued with HF

  • Similar safety to conventional SCS

  • HF at 10 kHz improved pain control for back/limb pain from baseline, and improved disability scored. May also salvage tSCS nonresponders.


  • Not randomized or controlled

  • No placebo arm

Perruchoud and colleagues (2013)

  • n = 40

    • Cohort: stable relief with tSCS

    • Results for n = 33

  • Randomized, controlled, double blind

  • HF at 5 kHz (instead of 10 kHz) vs. placebo

  • 2-week duration

  • Mean benefit of HF at 5 kHz not statistically significant over placebo

  • Strong “period effect” ( P < .006)

Possible that HF at 10 kHz is efficacious, but HF at 5 kHz is not

  • Therapy applied subthreshold kHz

  • Short duration

  • Does not address patients naïve to SCS

Only gold members can continue reading. Log In or Register to continue

Sep 21, 2019 | Posted by in PAIN MEDICINE | Comments Off on Spinal Cord Stimulation
Premium Wordpress Themes by UFO Themes