Chronic neuropathic pain in spinal cord injury: Efficiency of deep brain and motor cortex stimulation therapies for neuropathic pain in spinal cord injury patients



In spite of all the scientific advances in pharmacological research, a great number of patients cannot efficiently manage their chronic pain with conventional pharmacological treatments. Brain stimulation techniques have considerably improved these last 10 years. These techniques could be an interesting option after a rigorous selection of patients. We aim to evaluate the efficacy of brain stimulation (deep brain stimulation [DBS] and motor cortex stimulation [MCS]) within the framework of neuropathic pain management in spinal cord injury (SCI) patients and elaborate some recommendations.

Material and method

The methodology used, proposed by the French Society of Physical Medicine and Rehabilitation (SOFMER), includes a systematic review of the literature, the gathering of information regarding current clinical practices and a validation by a multidisciplinary panel of experts.


DBS is more effective on nociceptive pain than deafferentation pain. For the central pain of SCI patients, the long-term efficacy of DBS is quite low (three patients out of 19, amounting to 16%). MCS seems to have an interesting potential with a long-term efficacy of 57% (four patients out of seven), with less complications than DBS.


For central pain in SCI patients, there is no sufficient level of evidence to validate the use of DBS. There is however a low level of evidence for MCS. These results must be validated by larger comparative or controlled versus placebo clinical studies.



Dans le cadre des douleurs neuropathiques, le contrôle pharmacologique des douleurs s’avère souvent insuffisant. Les techniques de neuromodulation se sont développées ces dernières années, semblant offrir une alternative intéressante après sélection minutieuse des patients. Nous proposons d’évaluer l’efficacité des stimulations cérébrales (profondes et du cortex moteur) dans la prise en charge de la douleur neuropathique des patients blessés médullaires et d’élaborer des recommandations.

Matériel et méthode

La méthodologie utilisée, proposée par la Société française de médecine physique et de réadaptation (Sofmer), associe une revue systématique de la littérature, un recueil des pratiques professionnelles, une validation par un panel pluridisciplinaire d’experts.


La stimulation cérébrale profonde (SCP) se montre plus efficace pour les douleurs par excès de nociception que par déafférentation. Pour les douleurs centrales des blessés médullaires, l’efficacité à long terme de la SCP est très faible (trois patients sur 19, soit 16 %). La stimulation du cortex moteur (SCM) suscite actuellement un intérêt indiscutable, avec une efficacité à long terme de 57 % (quatre patients sur sept) et moins de complications que la SCP.


Pour les douleurs centrales des blessés médullaires, il n’y a pas de niveau de preuve scientifique suffisant pour l’utilisation de la SCP. Il y a en revanche un faible niveau de preuve en faveur de la SCM. Ces résultats doivent être confortés dans des études cliniques comparatives plus larges ou contrôlées contre placebo.

English version


In spite of all the scientific advances in pharmacological research, a great number of patients cannot efficiently manage their chronic pain with conventional pharmaceutical drugs. Physical and psychological therapies often do not bring any significant improvement to severe chronic pain. Stimulation techniques are being proposed more and more often for specific pathologies such as Parkinson’s disease, obsessive compulsive disorders, dystonia or chronic pain as a substitute to ablation surgery or in addition to pharmaceutical treatments .

Neurostimulation of the brain represents the most invasive technique. It should only be proposed when faced with a severe debilitating chronic pain and after the complete failure of other non-invasive treatments, especially pharmacological ones. This concerns two areas of the brain: thalamic stimulation, done since the 50s, and more recently cortical stimulation. These techniques require the surgical implantation of electrodes on the nervous tissues and electrical stimulation of the subjacent structures.

Material and method

The methodology proposed by the French Society of Physical Medicine and Rehabilitation (SOFMER) was used . The keywords used by the scientific committee were in the English language: chronic neuropathic pain in spinal cord injury; human; adult (and) motor cortex stimulation; deep brain stimulation. In the French language: douleur neuropathique, blessé médullaire, stimulation cérébrale profonde, stimulation du cortex moteur .

Only the clinical cases and series published in evidence-based medicine articles were included in this review, congress reports or book extracts were not selected. This review is based on and completes the “evidence-based report” published in the United States in September 2001 .


Deep brain stimulation (DBS)

A total number of 36 spinal cord injury (SCI) patients benefited from DBS . After implantation, the tests reported an improvement for 18 patients, 19 went on to a permanent implantation. On the long term, only three patients reported an improvement ( Table 1 ).

Table 1

Studies results on deep brain stimulation.

Authors Total of patients Spinal cord injury tests Targets Implanted spinal cord injury patients Long-term success Follow-up (months)
Young et al., 1985 48 6 VPL ± PVG ± PAG 6 1 20
Levy et al., 1987 141 11 VPL ± PVG/PAG 4 0 80
Kumar et al., 1997 68 3 VPL 1 0 78
Rasche et al., 2006 56 12 VPL ± PVG/PAG 5 1 6
Hamani et al., 2006 21 4 VPL ± PVG/PAG 3 1 60
Total 334 36 19 3

VPL: ventroposterolateral nucleus of the thalamus; PVG: periventricular gray matter; PAG: periaqueductal gray matter.

The 36 SCI patients come from five published case series, for a total of 334 implanted patients ( Table 1 ). The general inclusion criteria were chronic pain refractory to medical and/or surgical treatments. The etiology of the spinal cord injury was reported in two studies , the level of injury in one study , and the nature of the injury (complete/incomplete) suggested in one study . The patients’ characteristics were generally the same. . Two studies gave the age and sex of 16 SCI patients: nine men and three women with a mean age of 57 and four men with a mean age of 43 .

No specific description for the nature or characteristics of the neuropathic pain in these SCI patients was found. The pain evolution before DBS was reported in two studies, respectively eight years and 13 years . The mean follow-up duration for the entire series of patients varied from 27 to 80 months; for the SCI patients it was respectively six months and five years .

SCI patients with deafferentation pain were treated by stimulation of thalamus sensory nucleus – ventroposterolateral nucleus (VPL) – , of the periaqueductal or periventricular gray matter , or the internal capsule . The stimulation parameters were established for all patients, except in the Hamani et al. study . The test duration time lasted generally from five to seven days , sometimes up to 14 days . The criteria for a positive response to the test were: an improvement on the visual analog scale (VAS) of more than 50% , a “substantial improvement of chronic pain” , or the long-term use of this stimulation technique .

The functional result after implantation was evaluated using different tools: binary scale for the short-time improvement of pain , percentage of pain decrease , McGill Pain Questionnaire , assessment of pre-and post-surgery VAS , improved daily life activities , decrease in drugs consumption , long-term stimulation .

Adverse events seem to affect 19 to 22% of patients . A total of 80 major adverse events out of 266 implantations were reported in four studies and required 63 new surgical procedures . Infections were the most frequent complications ( N = 31), followed by scalp erosion ( N = 11), intracranial hemorrhages ( N = 7), seizures ( N = 3). These complications led to 17 definite surgical removal of the stimulation device. The frequency of dysfunctions related to the stimulation device (26 cases out of 189 reported in the 80s) has decreased . Among the minor complications we find headaches in 22% of cases or migraines in 51% of cases .

Stimulation of the motor cortex (SMC)

Seven SCI patients benefited from SMC. Four of them reported long-term improvements ( Table 2 ). These seven SCI patients are issued from the two published series and one unique case , for a total of 52 patients ( Table 2 ). The inclusion criterion was chronic neuropathic pain refractory to conventional treatments. The etiology for the spinal injury was traumatic in four patients , and caused by a herniated disc in three others . Age and sex of patients was available for two studies, two men and two women, mean age 47 years . Four patients were reported to have an incomplete SCI .

Table 2

Studies results on motor cortex stimulation.

Authors Total of patients Targets Number of electrodes Implanted spinal cord injury patients Long-term success Follow-up (months)
Nguyen et al., 1999 32 Unilateral 1 3 2 27.3
Tani et al., 2004 1 Bilateral 2 1 1 6.0
Nuti et al., 2005 19 Unilateral 1 or 2 3 1 49.0
Total 52 7 4

The pain description was reported in three studies , the pain evolution duration before implantation was five years for three patients . The mean follow-up for all patients varied from 27.3 months to 49 months , and follow-up period of at least six months for the unique patient .

A quadripolar stimulation electrode was implanted in the epidural space in light of the motor cortex contra lateral to the painful area . If needed a second electrode was implanted in some patients to cover the somatotopic representation for pain or to perform bilateral stimulation . The stimulation parameters were the same for all patients.

The MCS results were assessed using the VAS , pain decrease in VAS scores, or reported by the patient , improvement in daily life activities , medications consumption , or based on patients’ interviews . Among the four patients reporting long-term efficacy, three had seen a 40% or more decrease in the intensity of their pain .

Only the Nguyen et al. study reported adverse events, i.e. infection requiring the removal of the implanted device ( N = 1), headaches ( N = 3) and spinal hematoma ( N = 1) all had a spontaneous positive evolution, and a dehiscence of the stimulator’s pocket requiring another surgical intervention . An infection occurring on the implanted area required a unilateral removal of the device at six months .


Neurostimulation techniques have considerably improved these last 10 years, mainly due to the scientific advances in stimulation techniques and neuroimaging development, thus rendering these techniques more efficient while limiting the risks . These techniques are totally conservative, adaptable and reversible. They can only be envisioned after a careful assessment of the various factors and after having tried specific pain-relieving pharmaceutical treatments .

DBS and MCS action mechanisms remain quite unclear and hypothetical. The stimulation of the periaqueductal or periventricular gray matter seems to interact with the system producing opioid substances. The stimulation of specific sensory structures would lead to an inhibition of the extralemniscal nociceptive pathways as well as a modulation of the descending inhibitory pain pathways, leading to a control of the nociception . To explain the impact of MCS, we think that certain fibers establish some connections with subcortical structures involved in the transmission and control of the nociceptive stimuli, either in its sensory-discriminative area (thalamus, brain stem, spinal dorsal horn convergent neurons…) or in its emotional and behavioral area, involving the anterior cingulate gyrus, the insula, even the limbic system .

Generally, DBS is more efficient for pain triggered by excessive nociceptive stimuli than for deafferentation pain (63% vs 47% for long-term success). For neuropathic pain, the highest success rates are found for peripheral lesions (phantom limbs, neuropathies, plexopathies) . For the central neuropathic pain of SCI patients, 18 patients out of 36 (50%) reported an improvement during the trial period, but the long-term efficacy is very weak: three patients out of 19 (16%) .

For central neuropathic pain, the mean range of success is around 50% . For SCI patients, the three articles reported a 57% long-term efficacy (four patients out of seven) .

Most physicians surveyed at the SOFMER think that cortical stimulations have their place in the array of therapeutic options for chronic refractory SCI pain and they would refer their patients to a competent neurosurgery department specialized for this indication (Appendix 1).

Conclusion and recommendations

The level of scientific evidence of these studies is low (level 4), for DBS as well as MCS. For central neuropathic pain in SCI patients, there is no scientific argument in favor of DBS (grade C). With a success rate above 50% in the long-term, MCS seems more promising (grade C), but would require further comparative studies, with larger cohorts or controlled versus placebo.

Appendix 1. Results from the 116 attendees at the French Society of Physical Medecine and Rehabilitation (SOFMER) conference and the 50 other physicians who answered via the SOFMER website

Do you think that cortical stimulations have their place within the array of therapeutic options for treating chronic refractory pain in SCI patients?

  • Dnk:


  • Yes:


  • No:


Would you refer your patients to a competent neurosurgery department for this indication?

  • Dnk:


  • Yes:


  • No:


Dnk: does not know

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

Stay updated, free articles. Join our Telegram channel

Apr 23, 2017 | Posted by in PHYSICAL MEDICINE & REHABILITATION | Comments Off on Chronic neuropathic pain in spinal cord injury: Efficiency of deep brain and motor cortex stimulation therapies for neuropathic pain in spinal cord injury patients

Full access? Get Clinical Tree

Get Clinical Tree app for offline access