Added value of multiple versus single sessions of repetitive transcranial magnetic stimulation in predicting motor cortex stimulation efficacy for refractory neuropathic pain

Restricted access


Selection criteria for offering patients motor cortex stimulation (MCS) for refractory neuropathic pain are a critical topic of research. A single session of repetitive transcranial magnetic stimulation (rTMS) has been advocated for selecting MCS candidates, but it has a low negative predictive value. Here the authors investigated whether multiple rTMS sessions would more accurately predict MCS efficacy.


Patients included in this longitudinal study could access MCS after at least four rTMS sessions performed 3–4 weeks apart. The positive (PPV) and negative (NPV) predictive values of the four rTMS sessions and the correlation between the analgesic effects of the two treatments were assessed.


Twelve MCS patients underwent an average of 15.9 rTMS sessions prior to surgery; nine of the patients were rTMS responders. Postoperative follow-up was 57.8 ± 15.6 months (mean ± standard deviation). Mean percentage of pain relief (%R) was 21% and 40% after the first and fourth rTMS sessions, respectively. The corresponding mean durations of pain relief were respectively 2.4 and 12.9 days. A cumulative effect of the rTMS sessions was observed on both %R and duration of pain relief (p < 0.01). The %R value obtained with MCS was 35% after 6 months and 43% at the last follow-up. Both the PPV and NPV of rTMS were 100% after the fourth rTMS session (p = 0.0045). A significant correlation was found between %R or duration of pain relief after the fourth rTMS session and %R at the last MCS follow-up (R2 = 0.83, p = 0.0003).


Four rTMS sessions predicted MCS efficacy better than a single session in neuropathic pain patients. Taking into account the cumulative effects of rTMS, the authors found a high-level correlation between the analgesic effects of rTMS and MCS.

ABBREVIATIONS LEP = laser evoked potential; MCS = motor cortex stimulation; MEP = motor evoked potential; NPV = negative predictive value; NRS = Numerical Rating Scale; PPV = positive predictive value; ROC = receiver operating characteristic; rTMS = repetitive transcranial magnetic stimulation; SSEP = somatosensory evoked potential; %R MCS = percentage of pain relief following MCS; %R rTMS = percentage of pain relief following rTMS.

Downloadable materials

  • Supplementary Fig. 1 (PDF 973 KB)

Article Information

Correspondence Benjamin Pommier: Centre Hospitalier Régional Universitaire de Saint-Etienne, Hôpital Nord, Saint-Priest-en-Jarez, France.

INCLUDE WHEN CITING Published online May 18, 2018; DOI: 10.3171/2017.12.JNS171333.

Disclosures The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

© AANS, except where prohibited by US copyright law.



  • View in gallery

    Electrode positioning according to pain topography. A: Pain covering the whole hemibody: a first electrode was oriented anteroposteriorly, parallel to the parasagittal line, to treat pain in the lower limb. A second electrode was oriented superoinferiorly, along the main axis of the primary motor cortex, to treat pain in the face and/or upper limb. B: Pain covering the upper limb and face: the two electrodes were oriented anteroposteriorly, perpendicular to the primary motor cortex, in front of the corresponding somatotopic representations of the painful areas. Copyright Benjamin Pommier. Published with permission.

  • View in gallery

    Percentage (upper) and duration (lower) of pain relief according to rTMS session number in rTMS responders. Percentage of pain relief: Friedman test, F = 20.08, p = 0.00048. Duration of pain relief: Friedman test, F = 19.77, p = 0.00055. Square brackets indicate a significant Wilcoxon signed-rank post hoc test (p < 0.05).

  • View in gallery

    Percentage of pain relief after rTMS and MCS in patients categorized according to their final %R MCS (results are presented as mean ± standard deviation). FU = follow-up; M = month; S = session.

  • View in gallery

    Correlation observed between percentage of pain relief and percentage decrease of postoperative NRS (R = 0.94, p = 7.9E−6).

  • View in gallery

    Simple linear regressions models. Left: Association between %R rTMS and %R MCS. Right: Association between rTMS pain relief duration and %R MCS. Gray circles correspond to the first rTMS session (nonsignificant results) and black squares to the fourth rTMS session (significant results).

  • View in gallery

    Receiver operating characteristic (ROC) curve showing an area under the curve of 1 and an optimal %R rTMS threshold of 10% at the fourth rTMS session. Figure is available in color online only.

  • View in gallery

    Positive (PPV) and negative (NPV) predictive values of rTMS at the first (upper) and fourth (lower) rTMS sessions for MCS efficacy. Results are given considering a positive response threshold of 10%R for rTMS (rTMS+) and 40%R for MCS (MCS+; see text for details). Patients presenting with %R lower than these scores are respectively represented as rTMS− and MCS−. The %R rTMS at the fourth rTMS session provided statistically significant results (p = 0.0045) with high PPV and NPV. Instead of considering results at the fourth rTMS session, if we considered results after the first rTMS session, results were nonsignificant (p = 1) with a fairly high PPV of 80% but a low NPV of 29%.




All Time Past Year Past 30 Days
Abstract Views 296 296 164
Full Text Views 181 181 7
PDF Downloads 118 118 9
EPUB Downloads 0 0 0


Google Scholar