Interventions Based on Direct Spinal Cord Stimulation

Initial investigations of spinal cord stimulation were conducted in the early 1970s and were directed at individuals with multiple sclerosis (Cook & Weinstein 1973). Later studies have examined the effect of this approach in people with SCI to enhance bladder or bowel function and for the relief of pain and spasticity (Richardson & McLone 1978; Illis et al. 1983; Dimitrijevic et al. 1986a; Barolat et al. 1988). Typically, these studies employ a surgically implanted electrode under either general or local anaesthesia placed over the dorsal columns of the spinal cord that supplies ongoing electrical stimulation. Pinter et al. (2000) noted a declining interest with this approach in the 1990s because of technical concerns and “the realization that spinal cord stimulation was less effective in patients with severe spasms of the lower limbs” (Dimitrijevic et al. 1986b; Barolat et al. 1995).

Author Year

Research Design

Total Sample Size

Methods Outcome
Hofstoetter et al. 2014




Population: Mean age: 32.7 yr; Gender: males=2, females=1; Injury etiology: motor-incomplete SCI=3.

Intervention: Examine the effects of transcutaneous spinal cord stimulation (tSCS) on lower-limb spasticity.

Outcome Measures: Wartenberg pendulum test (WPT), Neurological recordings of surface electromyography (EMG), Non-functional co-activation during volitional movement, Timed 10 metre walk test (10MWT).

1.     Average Index of spasticity from pendulum test changed from 0.8±0.4 to 0.9±0.3, improvement in subject with lowest pre-stim index, and no changes other two subjects.

2.      All subjects decreased EMG activities during WPT.

3.      All subjects showed decreased exaggerated reflex responsiveness after tSCS, most effect on passive lower-limb movement (pre- to post-tSCS EMG ratio: 0.2:0.1).

4.      Gait speed for two subjects increased by 39%.

5.      Subject reports: lightness feeling in limbs, increased sensation, especially of foot sole during ground contact.

6.      Anti-spastic effects felt for 2-6hr post tSCS.

Pinter et al. 2000




Population: Age range: 18-34 yr; Gender: males=4, females=4; Level of injury: C5-T6; Level of severity: AIS: A–C; Time since injury range: 19–94 mo.

Intervention: Epidural spinal cord stimulation over the upper lumbar cord. Final internal placement for surgical implantation determined following an 8 wk trial period during which the stimulator was external.

Outcome Measures: Ashworth scale (AS), Clinical Rating Scale, Pendulum test, EMG activity in response to passive stretch.

1.     Spasticity was reduced as indicated by reduced Ashworth scale scores (p=0.0117).

2.     Pendulum test in 4 of 8 subjects showed reduced spasticity when stimulator was on for at least 1hr versus off for >12 hr.

3.     6 subjects showed marked reductions and 2 subjects showed moderate reductions with the clinical rating scale. It was not described what this entailed.

4.     EMG responses to stretch in the presence of stimulation were significantly reduced for all muscles combined of the left (p=0.0040) and the right (p=0.0035).

5.     Antispastic medication discontinued in all but 1 individual when continuous stimulation started. This individual had baclofen dose reduced and tizanidine discontinued.

Barolat et al. 1995




Population: Age range: 17-66yr; Gender: males=43, females=5; Level of injury: cervical=32, thoracic=16; Severity of injury: complete=25, incomplete=23; Time since injury range: 6–545 mo.

Intervention: Spinal cord stimulation following surgical implantation of the Medtronic Resume® electrode in the dorsal epidural space. Stimulus parameters determined in a training period 1-2 days after implantation typically resulted in a therapeutic window of stimulation between the motor and sensory threshold.

Outcome Measures: Average number of spasms, intensity of spasms and frequency of spasms. Severity score including both the intensity and frequency of the spasms. All were collected just prior and 3, 6, 12 and 24 mo after implantation.

1.     Of 48 initial subjects, 40 provided data at 3 mo, 33 at 6 mo, 31 at 1 yr and 18 at 2 yr. The remainder were discontinued due to lack of efficacy or lost to follow up.

2.     Average # of spasms/hr improved=19.9 initially, 11.3 at 3mo, 9.2 at 6 mo, 8.8 at 1 yr and 12.9 at 2 yr.

3.     A significantly greater proportion of subjects indicated reduced severity scores over time with significant differences at 6mo (p=0.0424), 1yr (p=0.0001) and 2 yr (p=0.0012) relative to baseline.

4.     Spasm intensity showed improvement over time with the proportion of individuals experiencing severe spasms being 83% initially, 33% at 3 mo, 45% at 6 mo, 32% at 1 yr and 28% at 2 yr.

5.     Subjective rating of spasm relief also decreased with 68% of individuals experiencing good or excellent relief at 3 mo, 69% at 6 mo, 70% at 1 yr and 79% at 2 yr.

Dekopov et al. 2015


Case Series


Population: Mean age: 35.7 yr; Gender: males=40, females=31; Injury etiology: cerebral palsy=52, spinal trauma=13, multiple sclerosis=2, spinal cord tumour=2, spinal stroke=1, pyogenic epiduritis=1; Level of Injury: paraparesis=41, tetraparesis=11.

Intervention: Assess the effectiveness of spinal cord stimulation (SCS) in different groups of individuals with spasticity of different origin.

Outcome Measures: Ashworth Scale (AS), Range of passive and active motion; Group 1: Gross Motor Function Measure (GMFM-88); Group 2: Arens Scale.

1.     Decrease in muscle tone in all cases for groups with spinal spasticity (3.71±0.61 to 2.26±0.56, p>0.001).

2.     Therapy was discontinued in 8(11%) cases due to improvement (Cerebral palsy=7 cases, spinal injury=1 case).

3.     In spinal injury spasticity group, Arens Scale showed significance for functional improvement (2.68±1.2 to 3.31±1.6; p=0.004).

4.     1 case in spinal injury group had SCS system removed after 6mo due to worsening spasticity and function, fixed contractures before operation.

Midha & Schmitt, 1998


Case Series


Population: Age range: 29-63 yr; Level of injury: C4-T10; Level of severity: complete, incomplete; Time since injury range: 6 mo-30 yr.

Intervention: Retrospective analysis of those having undergone implantation of an epidural stimulator between 1986 and 1988.

Outcome Measures: Telephone follow-up (asked to quantify symptom relief on a scale from 0-10, 10=total symptom relief) and how long they had had the symptom relief since the time of the implantation.

1.     At the time of the retrospective study, 1 of 17 individuals reported that the epidural stimulator was producing symptomatic relief.

2.     The average length of time that all units produced symptomatic relief was 6mo (range 0-96mo).

3.     Fourteen units were removed within 3.4 yr (5 days-7 yr); 9 implantations failed from the day of implantation.

4.     Total cost of initial implantation (not including follow-up) is $23,600 per unit.


Hofstoetter et al. (2014) conducted a pre-post study (n=3) on subjects with motor-incomplete SCI who could walk ≥10 metres to examine the effects of transcutaneous spinal cord stimulation (tSCS) on lower-limb spasticity. Two interconnecting stimulating skin electrodes were placed paraspinally at the T11-12 vertebral levels. Biphasic two millisecond width pulses were delivered at 50 Hz for 30 minutes at intensities producing paraesthesias but no motor responses in the lower limbs. The Wartenberg pendulum test (WPT) and neurological recordings of surface-electromyography (EMG) were used to assess the effect of exaggerated reflex excitability, non-functional co-activation during volitional movement, and for clinical function assessment the timed 10-m walk test (10MWT) was used. The study demonstrated that the average Index of spasticity from pendulum test changed from 0.8±0.4 to 0.9±0.3 with improvement in the subject with the lowest pre-stimulation index, and no changes in the other two subjects. All subjects showed decreased EMG activities during WPT and on exaggerated reflex responsiveness after tSCS, with the most effect seen on passive lower-limb movement (pre-to post-tSCS EMG ratio: 0.2±0.1). Gait speed for two subjects increased by 39% and all subjects reported a lightness of feeling in the limbs, increased sensation especially of the sole of the foot during ground contact and anti-spastic effects of two-six hr post tSCS. The study suggests that tSCS may be used for spasticity control without negatively affecting residual control in iSCI.

Pinter et al. (2000) showed improvements following implantation of an epidural spinal cord stimulator with a variety of clinical measures including significant decreases in Ashworth scale scores (p=0.0117), the pendulum test and muscle activity as indicated by reduced summed EMG activity collected during passive movements in both the left (p=0.0040) and the right (p=0.0035) lower limb. In addition, it was possible to discontinue anti-spastic medication in seven of eight subjects and reduce the dose in the remaining subject. These positive findings were achieved in a rather small population (N=8) and additional studies from independent groups are required to further demonstrate the feasibility and efficacy of this approach. In particular, the long-term effectiveness of spinal cord stimulation is uncertain, as this study did not specify the specific time points when measures were collected. However, they did state that spinal cord stimulation had been conducted for a mean of 14.4 months (Pinter et al. 2000). These authors asserted that better results were obtained with their approach as they were more careful in optimising location and other methodological aspects and outcomes could be further enhanced by improved stimulator design.

Barolat et al. (1995) also reported beneficial reductions in spasticity with epidural spinal cord stimulation as assessed by subjective scales of spasm frequency and intensity. Spasm intensity and spasm frequency were reduced significantly over the follow-up period of two years and a significantly greater proportion of subjects indicated reduced spasticity severity scores over time with significant differences at six months (p=0.0424), one year (p=0.0001) and two years (p=0.0012) relative to baseline. It should be noted that the positive nature of the long-term findings are somewhat muted as subjects were increasingly dropped from the analysis over time when they were lost to follow-up or discontinued due to lack of efficacy. Of 48 initial subjects, 40 provided data at three months, 33 at 6 months, 31 at one year and 18 at two years.

In contrast to these findings, Midha and Schmitt (1998) conducted a telephone or in-person follow-up of individuals having epidural stimulators implanted between 1986 and 1988 to determine their long-term status (N=17). In only one of these individuals was the stimulator continuing to provide symptomatic relief although most felt it was initially effective with an average time of effectiveness of six months. The rate of stimulator failures was high with several removals and re-implantations of devices. At the time of follow-up only 10 individuals reported having an implanted stimulator.


There is level 4 evidence (from three pre-post studies and one case series: Hofstoetter et al. 2014; Pinter et al. 2000; Barolat et al. 1995; Dekopov et al. 2015) that ongoing spinal cord stimulation may provide some relief from otherwise intractable spasticity.

There is level 4 evidence (from one pre-post study and one case series: Hofstoetter et al. 2014; Midha & Schmitt 1998) that the beneficial effects of spinal cord stimulation may subside for most initial users over a short period of time. This, combined with the potential for equipment failure and adverse events, suggests that spinal cord stimulation may not be a feasible approach for ongoing management of spasticity.