Analgesia-enhancing effects of repetitive transcranial magnetic stimulation on neuropathic pain after spinal cord injury:An fNIRS study

Abstract

Background:

Repetitive transcranial magnetic stimulation (rTMS) is a promising treatment for chronic intractable neuropathic pain in patients with spinal cord injury (SCI). However, the analgesia-enhancing effects of rTMS on conventional interventions (e.g., medications), and the underlying mechanisms remain poorly understood.

Objective:

To investigate the enhancement of analgesia and change of cortex activation by rTMS treatment on neuropathic pain following SCI.

Methods:

A double-blind, sham-controlled, clinical trial was performed. Twenty-one patients with neuropathic pain after SCI were randomized (2:1) to receive a session of rTMS (10 Hz, a total of 1200 pulses at an intensity of 80% resting motor threshold) or sham treatment over the left primary motor cortex (M1) corresponding to the hand area daily for six weeks with a one-day interval per week. At T0 (before rTMS treatment), T1 (after the first session rTMS), T2 (after one week), T3 (after two weeks), T4 (after four weeks) and T5 (after six weeks), activations in the bilateral M1, primary somatosensory cortex (S1), premotor cortex (PMC) and prefrontal cortex (PFC) during the handgrip task were measured using functional near-infrared spectroscopy (fNIRS). In addition, the numerical rating scale (NRS) was used to assess pain.

Results:

The pain intensity or activation in PFC, PMC, M1 or S1 was not remarkably changed at T1. Along with the time, the pain intensity gradually decreased in both the rTMS and sham groups. The real rTMS, compared with the sham, showed more pain relief from two weeks (T3) to six weeks (T5), and the activations of the motor-related areas M1 and PMC were remarkably suppressed.

Conclusions:

The findings of this preliminary study with a small patient sample suggest that the analgesia-enhancing effects of high-frequency rTMS might be related with the amelioration of M1 and PMC hypersensitivity, shedding light upon the clinical treatment of SCI-related neuropathic pain.

Keywords

rTMS neuropathic pain spinal cord injury fNIRS

1 Introduction

Neuropathic pain is a highly disabling and common complication after spinal cord injury (SCI) that significantly diminishes patients’ quality of life and has limited effective pharmacological therapies available (Widerström-Noga, 2017). Growing evidence suggests that cortical hyperexcitability caused by peripheral and central sensitization contributes to the development of neuropathic pain (Nees, Finnerup, Blesch, & Weidner, 2017; Topka, Cohen, Cole, & Hallett, 1991). Therefore, brain stimulation techniques that alter the excitability of the cortex are capable of controlling neuropathic pain (Moreno-Duarte et al., 2014). A recent functional magnetic resonance imaging (fMRI) study on rats with neuropathic pain caused by spinal cord lesions revealed that motor cortex stimulation (MCS) alleviates pain intensity partially by suppressing hyperactivity in the primary somatosensory cortex (S1) and prefrontal cortex (PFC) (Jiang et al., 2014). Compared with MCS, repetitive transcranial magnetic stimulation (rTMS) is a noninvasive and economical neuromodulatory technique, emerging as a promising analgesic treatment for neuropathic pain (Nardone et al., 2014).

Several studies have investigated the analgesic effects of rTMS on neuropathic pain following SCI, but without yielding clear conclusions (Gao et al., 2017; Tazoe & Perez, 2015). Ten real (5 Hz, a total of 500 pulses at an intensity of 115% motor threshold) or sham daily motor rTMS treatments induced a similar and salient reduction in pain intensity in patients with SCI by the end of the treatment series, but only the treatments with real rTMS produced continued pain relief during the follow-up period (Defrin, Grunhaus, Zamir, & Zeilig, 2007). Another study found that five consecutive days of motor rTMS treatments (10 Hz, a total of 1000 pulses at an intensity of 80% resting motor threshold) did not significantly decrease the pain sensation compared with the sham treatments (Kang, Shin, & Bang, 2009). Therefore, the duration of rTMS treatments may be a critical factor influencing the analgesic effects, but did not last more than ten days in previous studies. Besides, it has not been reported whether rTMS enhances the analgesic effects of conventional rehabilitation interventions and medications. Also, the lack of understanding of the mechanisms underlying rTMS-induced analgesia may partially contribute to the mixed results and absence of improvement in effectiveness.

Functional near-infrared spectroscopy (fNIRS) is a non-invasive technique that can detect changes in cortical oxyhemoglobin (HbO) in real-time. Changes in HbO concentration are closely related to the activation state of the brain (Ren et al., 2017). Compared with other neuroimaging techniques, such as fMRI and positron emission tomography, fNIRS is portable and does not have stringent motor and physical constraints, thus being especially suitable for patients with SCI (Zhu et al., 2015).

In the present study, we performed a clinical trial to investigate the enhancement of analgesia by treatment with rTMS on neuropathic pain following SCI. Furthermore, the underlying mechanisms were explored by fNIRS.

2 Methods

2.1 Participants

According to the inclusion criteria, a total of 24 right-handed patients with neuropathic pain following SCI were recruited at the inpatient unit of Department of Rehabilitation Medicine in Xijing Hospital, Xi’an, China between March 2018 and September 2018. Three participants were excluded due to refused participation (n = 2) or at least one exclusion criteria (n = 1). Thus, a total of 21 patients entered the trial.

The study inclusion criteria included: (1) complete or incomplete SCI; (2) neuropathic pain at or below the lesion level with an average pain intensity, as assessed by the numerical rating scale (NRS), where 0 is “no pain” and 10 is “the worst possible pain”, of no less than 4 at baseline and before rTMS treatment. The neuropathic pain was diagnosed by a rehabilitation expert according to its definition and characteristics (Treede et al., 2008). NRS was chosen because it is recommended by the National Institute on Disability and Rehabilitation Research–sponsored consensus group regarding SCI-related pain (Bryce et al., 2007) and by the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials consensus group for use in pain clinical trials (Dworkin et al., 2005); (3) pain not attributable to any other causes (e.g., rheumatologic disorders or diabetes). In consideration of the contraindications for rTMS and fNIRS, patients with the following conditions were excluded: (1) any metallic implant in the head, (2) an implanted stimulator, including cardiac pacemaker or drug delivery system, or (3) a personal or familial historyof epilepsy.

This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Xijing Hospital. All participants provided written informed consent before participation.

2.2 Study design

As shown in Fig. 1, this study was a randomized, double-blind, sham-controlled study. Using the computer-generated randomization method, eligible patients were assigned to the rTMS and the sham group at a 2:1 ratio according to the previous studies (Mishra, Nizamie, Das, & Praharaj, 2010; Quan et al., 2015). Since the duration of rTMS treatment on neuropathic pain following SCI did not last more than ten days in the previous studies, the 2:1 ratio was chosen to obtain more safety information on the application of rTMS treatment for six weeks and to enhance enrollment in this study.

Fig.1

Treatment study flowchart.

All patients received a session of rTMS daily for six weeks with a one-day interval per week. The NRS was assessed and fNIRS was performed at baseline (T0), after the first session rTMS (day 1, T1), and at the end of week 1 (T2), 2 (T3), 4 (T4), and 6 (T5). The patients and the researchers evaluating the patients were blind to the grouping and treatment. Conventional rehabilitation interventions and medications, including antiepileptic drugs (AED) and non-steroidal anti-inflammatory drugs (NSAID), were not affected during the entire experiment.

2.3 Navigated rTMS

The rTMS was delivered using a figure-of-eight magnetic coil connected to a CCY-1 stimulator (YIRUIDE Medical Equipment Company, Wuhan, China) as described in our previous study (Long et al., 2018). To increase the accuracy and replicability of coil localization, the Visor2 TMS navigation system (Advanced Neuro Technology, Enschede, Netherlands) was used (Rossini et al., 2015). Each patient’s own T1-weighted magnetic resonance image was integrated into the navigation system and a three-dimensional reconstruction of the brain was automatically generated. Using an infrared camera, the location and orientation of the TMS coil and patient’s brain were co-registered in the system and visualized in real-time.

The coil was first navigated over the left primary motor cortex (M1). Then, it was carefully moved to the optimal position, which elicits the largest motor-evoked potential (MEP) in the abductor pollicis brevis muscle of the right hand, defined as the “hotspot”. To determine the resting motor threshold (RMT), the minimal amount of machine output producing MEPs≥50μV in at least 5 of 10 trials was used. For the rTMS group, 10 Hz rTMS was delivered to the hotspot under 80% of the RMT for 1.2 s at an intra-train interval of 3 s with a total of 1200 pulses. The magnetic stimulus had a biphasic waveform and the length of pulse was 340μs±20μs. The coil of the sham group was held to the scalp at a 90° angle, with the same site and stimulation parameters as the rTMS group. The rTMS was administered at the same time of a day. Throughout the study, the patients were carefully monitored in medical and neurological examinations to assess the safety of rTMS treatment.

2.4 fNIRS measurement

An fNIRS system (FOIRE-3000, Shimadzu Corporation, Kyoto, Japan) was used to detect the absorption of near-infrared light at the three wavelengths of 780, 805, and 830 nm. The optical data were then transformed into changes in HbO, deoxyhemoglobin (Hb), and total hemoglobin (HbT) concentrations according to the modified Beer–Lambert law (Baker et al., 2014). The measured areas and optical channel locations are shown in Fig. 2. We located the corresponding channels on the regions of interest (ROIs), bilateral M1, S1, premotor cortex (PMC), and PFC according to the international 10–20 system (Table 1). The fNIRS data were recorded concurrently with performance of the handgrip task.

Fig.2

Localization of light emitters (red squares) and detectors (blue squares) and measurement channels (yellow squares) for near-infrared spectroscopy (L: left hemisphere, R: right hemisphere). See Table 1 for cortical ROIs corresponding to the channels. ROIs, regions of interest.

Table 1

Cortical ROI corresponding to the channels

Cortical regions	Channels
L-PFC	1, 3, 6, 8, 11
R-PFC	2, 5, 7,10, 12
L-PMC	13, 14
R-PMC	25, 26
L-M1	17, 19
R-M1	27, 30
L-S1	24
R-S1	35

ROIs, regions of interest; L-PFC, left prefrontal cortex; R-PFC, right prefrontal cortex; L-PMC, left premotor cortex; R-PMC, right premotor cortex; L-M1, left primary motor cortex;R-M1, right primary motor cortex; L-S1, left primary somatosensory cortex; R-S1, right primary somatosensory cortex.

The subjects were trained to perform the handgrip task before the fNIRS examination according to a previous study with modifications (Morishita et al., 2015). Briefly, the handgrip task consisted of the (1) left task, where fists were clenched for 5 seconds and then loosened for 5 seconds with three repetitions, (2) a resting state of 15 seconds, (3) right task, which followed the same protocol as the left task, and (4) a resting state of 15 seconds (Fig. 3).

Fig.3

Handgrip task design.

2.5 Data analysis

The differences of demographics were tested using the Mann-Whitney U test for the continuous variables and the χ² test for categorical variables. The repeated-measures analyses of variance (ANOVA) were conducted to assess the effect of rTMS (real, sham) on the values of NRS (Defrin et al., 2007). The analysis of fNIRS data was performed as our previously study (Li et al., 2015). Since HbO is the most sensitive indicator of activity-dependent changes in blood flow in brain regions, only HbO was assessed in this study (Miyai et al., 2003). A hemodynamic response function filter and a wavelet-minimum description length detrending algorithm were used to remove physical noise and artifacts for each participant. The precoloring method was then used to attenuate high frequency components and smooth the fNIRS data. General Linear Model (GLM) analysis was applied to the time course of HbO concentration changes by using the NIRS-SPM toolbox in Matlab. Two-sided Student t-tests and repeated-measures ANOVA were used for further statistical analyses of imaging data (Schecklmann et al., 2014). First, brain activation was assessed by two-sided Student’s t-test against zero and interpolated for all channels over the whole probe set according to T-maps. Second, intervention-specific instant effects were evaluated by ANOVA for independent factors (rTMS vs sham) and time as the dependent factor (T0 vs T1). Third, intervention-specific cumulative effects were evaluated by repeated-measures ANOVA for the independent factors (rTMS vs sham) and time as the dependent factor (T0, T2, T3, T4, and T5). All statistical analyses were performed with SPSS 20.0.0.1 (SPSS Inc., USA) and NCSS 11.0.9 (NCSS, Kaysville, UT).

3 Results

A total of 21 eligible patients were randomly assigned to the rTMS group (n = 14) and the sham group (n = 7). Three patients in the rTMS group and one patient in the sham group withdrew early from the study because of personal reasons. Thus, seventeen patients completed the treatment protocol and entered into the final analysis (Fig. 1). No patients complained discomfort during or after the rTMS treatment. No pathologic symptoms, such as seizure, were reported by any patient during the clinical trial. The real rTMS was as well tolerated as the sham rTMS during the whole experiment. The baseline characteristics of the enrolled patients are shown in Tables 2 and 3. The demographics and average pain intensity during the baseline period were indistinguishable between the two groups.

Table 2
Patient characteristics

All patients (n = 17) rTMS group (n = 11) Sham group (n = 6) P

Age (y) 37.00 (23.00–54.50) 47.00 (31.00–60.00) 41.00 (20.50–48.50) 0.580

Sex (male/female) 15/2 10/1 5/1 1.000

Type of SCI (complete/incomplete) 12/5 8/3 4/2 1.000

NRS at baseline 5.00 (4.00–5.00) 5.00 (4.00–5.00) 4.50 (4.00–5.50) 0.615

	All patients (n = 17)	rTMS group (n = 11)	Sham group (n = 6)	P
Age (y)	37.00 (23.00–54.50)	47.00 (31.00–60.00)	41.00 (20.50–48.50)	0.580
Sex (male/female)	15/2	10/1	5/1	1.000
Type of SCI (complete/incomplete)	12/5	8/3	4/2	1.000
NRS at baseline	5.00 (4.00–5.00)	5.00 (4.00–5.00)	4.50 (4.00–5.50)	0.615

Data are median (interquartile range) for continuous variables (differences were tested using the Mann-Whitney U test) and number for categorical variables (differences were tested using the χ² test). SCI, spinal cord injury; NRS, numerical rating scale; y, year; mo, month.

Table 3

SCI levels and pain locations of the patients

Patient	Group	ASIA	SCI Level	Type of pain	Pain Locations	NRS at baseline
1	rTMS	A	T1	Below	Bilateral lower limbs	6
2	rTMS	A	T4	Below	Bilateral lower limbs	4
3	rTMS	A	L1	At	Left shoulder, Thoracic, Back	6
4	rTMS	A	C5	At	Back	5
5	rTMS	A	T1	At	Trunk	4
6	rTMS	D	C6	Below	Limbs, Thoracic, Abdomen	5
7	rTMS	A	T9	Below	Bilateral lower limbs	4
8	rTMS	C	T1	Below	Bilateral lower limbs	4
9	rTMS	C	C3	Below	Shoulders, arms, hands	5
10	rTMS	A	L1	At	Back	5
11	rTMS	A	C5	Below	Limbs, Trunk	5
12	Sham	D	T1	Below	Bilateral lower limbs	5
13	Sham	A	T2	Below	Bilateral lower limbs	4
14	Sham	A	L1	At	Abdomen	4
15	Sham	A	T12	At	Back	4
16	Sham	C	T4	At	Back	7
17	Sham	A	C6	Below	Bilateral lower limbs	5

SCI, spinal cord injury; NRS, numerical rating scale; mo, month.

The analgesia-enhancing effects of single-session rTMS were firstly evaluated and the results showed that the main effect of group (rTMS and sham), the main effect of time (T0 and T1) and the interaction group by time were not significant (Fig. 4). Then the changes of NRS ratings throughout the whole experiment were analyzed. The pain intensity gradually decreased over time in both the rTMS group and sham groups. The main effects of time and group were significant (P < 0.01), so was the interaction group by time (P < 0.05). Post-hoc analysis further revealed the difference between the rTMS and sham groups became significant starting at week 2 (T3, P < 0.05; Fig. 4).

Fig.4

Changes in the NRS scores over time. NRS, numerical rating scale. Bars denote mean±standard deviation (SD). *P < 0.05 compared with the rTMS group at respective time points. rTMS, repetitive transcranial magnetic stimulation.

To determine the instant effects of the single-session rTMS on regional cerebral blood flow, we performed fNIRS scans before and immediately after the first session rTMS. Figure 5 shows the activation map of seven ROIs, including the PFC, left PMC (L-PMC), left M1 (L-M1), left S1 (L-S1), right PMC (R-PMC), right M1 (R-M1), and right S1 (R-S1), during left or right handgrip task before and after the first session rTMS. The comparison with the T0-rTMS and the T1-rTMS from the left handgrip task showed that there was no obvious difference in the PFC, L-PMC, L-M1, or L-S1 but that there was a slight difference in the R-PMC, R-M1, and R-S1 (Fig. 5A). We summed the values of channels within each ROI for statistical analysis and found no significant difference in any ROI (Fig. 5B). The T0-rTMS was compared with T1-rTMS during the right handgrip task. There were no obvious differences in the R-PMC, R-M1 R-S1, L-S1, or L-PMC, but a slight difference was observed in the PFC and L-M1 (Fig. 5C). Similarly, the calculation results revealed no significant difference in any ROI (Fig. 5D).

Fig.5

HbO activation (t-scores) mapping by fNIRS during handgrip task before (left column, T0) and after (right column, T1) the first session rTMS (A, C). Black numbers on the maps represent the measured channels with the same distribution as in Table 1. (B and D) Statistical analysis of A and C, respectively, based on calculated HbO changes within the ROIs. Bars denote mean±standard deviation (SD). fNIRS, functional near-infrared spectroscopy; HbO, oxyhemoglobin.

Nextly, fNIRS was performed at 1 (T2), 2 (T3), 4 (T4), and 6 weeks (T5) to assess the cumulative effects of multisession rTMS on blood oxygen. Repeated-measures ANOVA with HbO as the dependent variable, condition (rTMS vs sham) as a between-subjects independent variable, and time (T0, T2, T3, T4, and T5) and task (left and right hand) as within-subject independent variables revealed a significant condition×task interaction (P < 0.05) in the L-PMC. The post-hoc analyses revealed a pronounced decrease in HbO changes in L-PMC when performing the left task after multisession rTMS and the calculated HbO changes were significantly different (Fig. 6). Additionally, significant condition×task×time interactions (P < 0.05) were revealed in the L-M1. Post-hoc analyses revealed that at week 2, the right handgrip task of the sham group induced a distinct activation and the calculated HbO changes were significantly different between left and right handgrip tasks in the L-M1 (P < 0.05), while multisession rTMS decreased the magnitude of activation, resulting in no significant difference in the calculated HbO changes (Fig. 7).

Fig.6

Comparison of HbO activation (t-scores) mapping during left handgrip task between multisession rTMS and sham treatmentss (A). (B) HbO changes within left premotor cortex calculated for statistical analysis of A. Bars denote mean±standard deviation (SD). ***P < 0. 001, compared with the sham group during left handgrip task.

Fig.7

Comparisons of HbO activation (t-scores) mapping during handgrip task between multisession rTMS (right column) and sham (left column) treatments after 2 weeks (A). (B) HbO changes within left primary motor cortex after 2 weeks calculated for statistical analysis of A. Bars denote mean±standard deviation (SD). *P < 0.05, compared with the sham group during left handgrip task after 2 weeks.

4 Discussion

This study confirmed that single-session rTMS did not enhance the analgesic effects, or induce the blood oxygen changes in the PFC, PMC, M1, or S1 during the handgrip task, while multisession rTMS for at least two weeks mitigated pain more greatly than sham, and the activations of motor-related areas, including the M1 and the PMC, were suppressed when performing the handgrip task

The addition of single-session real rTMS treatment failed to significantly reduce the pain intensity compared with the sham rTMS treatment, suggesting the single-session rTMS treatment has limited analgesic-enhancing effects. Supporting this, Ruth Defrin et al. reported that the pain alleviation by one rTMS (5 Hz) treatment is probably due to the placebo effect (Defrin et al., 2007). However, it was also found that the analgesic effect of single-session rTMS treatment is closely related to pain origin. A single rTMS (10 Hz) over the M1 reduces by approximately 40% the pain intensity in patients suffering from trigeminal nerve lesions, while patients with SCI react relatively worse to rTMS (Lefaucheur et al., 2004). This is possibly due to the relatively damaged descending modulatory pathways in the spinal cord, which affects the efficacy of corticothalamic descending control of the rTMS treatment (Basbaum & Fields, 1984). Alternatively, the fNIRS study revealed that there were no changes in blood oxygen in cortex after single-session rTMS, suggesting single-session rTMS may not induce enough changes in cortical activation to affect the modulatory pathways of pain.

Unlike the single-session rTMS, it was found that starting at two weeks post-initiation of treatment, the rTMS group has a greater pain relief than the sham group, which lasted till the end of the experiment. Previous studies reported that ten daily treatments with the real rTMS (10 Hz or 5 Hz) induced an analgesic effect comparably to the sham rTMS treatments, although only the real rTMS group experienced pain relief during the follow-up period (Defrin et al., 2007; Yılmaz, Kesikburun, Yaşar, & Tan, 2013). In another blind crossover study, 11 SCI patients with neuropathic pain received the rTMS (10 Hz for 5 consecutive days) and sham treatments over the M1 hand area for 12 weeks apart. At one week post-treatment, the pain relief did not differ between treatment types (Kang et al., 2009). One explanation for this disparity may be the experimental design. To better conform with clinical practices, we did not exclude conventional rehabilitation intervention or medication, nor did we consider whether the patients were sensitive to conventional intervention. Most of the previous studies only included patients with chronic intractable neuropathic pain (a minimum duration of 12 months), who were resistant to conventional intervention (Defrin et al., 2007; Kang et al., 2009; Lefaucheur et al., 2004; Yılmaz et al., 2013). Another explanation may be the duration of the rTMS treatments. In this study, rTMS treatments were performed for up to six weeks, while the previous studies investigated the cumulative effects of the rTMS treatments for no more than ten days. Therefore, our results highlight the importance of the sufficient duration of the rTMS treatments for generating analgesia-enhancing effects.

Besides the duration, the rTMS target also plays a critical role in generating analgesic effects. The M1, S1, PMC, and supplementary motor area were targeted using the navigation-guided rTMS in 20 patients experiencing pain. M1 was found to be the only efficient target (Hirayama et al., 2006). It is well known that human motor cortex contains an overlapping sequence of body part representations from the tongue in a ventral location to the foot in a dorsal location. Some authors believe that the rTMS treatment directly induces local excitability changes in the target, which should therefore be in the area corresponding to the pain (Andre-Obadia et al., 2006; Defrin et al., 2007; Lefaucheur, Drouot, Menard-Lefaucheur, Keravel, & Nguyen, 2006; Moseley & Flor, 2012; Yılmaz et al., 2013). However, pain relief was better obtained with facial pain than upper limb pain when the target was in the motor hand area (Lefaucheur et al., 2004). Fanny Jetté et al. compared the analgesic effects of a single rTMS treatment over the motor hand or leg area on the pain following SCI and found a significant but equivalent reduction of pain in both areas, suggesting that the targeting of the M1 area corresponding to pain may not be necessary (Jetté, C Té, Meziane, & Mercier, 2013). In our study, most of the patients suffered from neuropathic pain outside the upper limbs, but we targeted the rTMS treatment to the hand area and found remarkable analgesia-enhancing effects. Compared with other motor areas such as the area corresponding to the lower limbs, the hand area is closer to the scalp and wider, which allows for a higher intensity induced electrical field during the rTMS treatment and is much easier to accurately stimulate. Accordingly, our results suggest that the hand area may be an efficient target for rTMS to generate analgesic-enhancing effects.

It was interesting to find that multisession rTMS dampened the activations in the motor-related cortex M1 and PMC, suggesting that the motor cortex may be involved in the cumulative analgesia-enhancing effects of the rTMS treatment. Previous studies suggested that neuropathic pain might partially result from a “bottom-up” mechanism, in which SCI results in maladaptive neuroplasticity, central sensitization, and hypersensitivity of the motor cortex and, ultimately, the sensory pathway (Nees et al., 2017; Topka et al., 1991). The activation of M1 during motor imagery activates the pain neuromatrix and ongoing pain in subjects with SCI (Gustin, Wrigley, Henderson, & Siddall, 2010). Additionally, in the patients with neuropathic pain, allodynic stimuli significantly increase activation in the M1 and premotor areas (Peyron et al., 2004). It is well established that the M1 is closely connected with the PMC (Teitti et al., 2008). In our study, the PMC and M1 were markedly activated during the handgrip task in the sham group, which is consistent with a previous study (Freund et al., 2011). However, after multisession rTMS treatments, the activations of the PMC and M1 were significantly suppressed and accompanied by a decrease in pain intensity. Notably, the suppressed PMC and M1 areas were on the left side, which is the same as the rTMS treatment, suggesting a direct effect from the rTMS treatment. Our study also assessed the cognition-associated PFC and sensory S1, but did not found that there were significant differences between the sham and rTMS groups. The PFC and S1 are implicated in nociceptive processing (Garcia-Larrea & Peyron, 2013). Through fMRI scans, Li Jiang et al. found that blood oxygenation level dependent (BOLD) signals in the PFC and S1 were significantly attenuated after a session of MCS treatment (50μA, 50 Hz, 300μs, for 30 min) in anesthetized rats with spinal cord lesions (Jiang et al., 2014). Also, the fMRI study revealed that the pain intensity experienced by patients with SCI was significantly correlated with the magnitude of activation in the right dorsolateral PFC (Gustin et al., 2010). The discrepancy may lie in whether anesthesia was utilized and the differences between fMRI and fNIRS. Anesthesia suppresses metabolic activity and disrupts cortical networks (Peeters, Tindemans, De Schutter, & Van der Linden, 2001). Compared with fMRI, fNIRS has the advantage of a high temporal resolution, but the disadvantage of a low spatial resolution.

The main limitation of this study was the small sample size. To observe the cumulative effects, the rTMS treatment was performed for as long as six weeks, making it difficult to enroll more patients who could receive treatment for the entire experimental period. Future studies with a larger sample size are needed. Second, depression is closely related to neuropathic pain and rTMS is an effective intervention for it (Campbell, Clauw, & Keefe, 2003). In this study, we did not investigate comorbid depression in the patients, since it was previously reported that the real and sham rTMS treatments could similarly reduce depression, which was not sufficient to explain pain reduction (Defrin et al., 2007).

5 Conclusions

Our results suggest that high-frequency rTMS treatment over the hand area of the left M1 for more than two weeks effectively enhances the analgesic effects of conventional rehabilitation and pharmacological therapies on neuropathic pain after SCI, which may partially result from the amelioration of M1 and PMC hypersensitivity. These findings contribute to our understanding of the analgesic effects of rTMS and have important implications for the relief of SCI-related neuropathic pain in the clinic.

Disclosure

The authors have no financial conflicts of interest.

Footnotes

Acknowledgments

This study was supported, in part, by the International Science and Technology Cooperation Program of China (grant no. 2013DFA32610), PLA Military Medicine Innovative Engineering Project (grant no. 16CXZ022), and Young Elite Scientists Sponsorship Program by CAST (grant no. 2018QNRC001).

References

Andre-Obadia,

, Peyron,

, Mertens,

, Mauguiere,

, Laurent,

, ... Garcia-Larrea,

(2006). Transcranial magnetic stimulation for pain control. Double-blind study of different frequencies against placebo, and correlation with motor cortex stimulation efficacy. Clinical Neurophysiology, 117(7), 1536–1544.

Baker,

W.B.

, Parthasarathy,

A.B.

, Busch,

D.R.

, Mesquita,

R.C.

, Greenberg,

J.H.

, ... Yodh,

A. G.

(2014). Modified Beer-Lambert law for blood flow. Biomedical Optics Express, 5(11), 4053–4075.

Basbaum,

A.I.

, Fields,

H. L.

(1984). Endogenous pain control systems: Brainstem spinal pathways and endorphin circuitry. Annual Review of Neuroscience, 7, 309–338.

Bryce,

T.N.

, Budh,

C.N.

, Cardenas,

D.D.

, Dijkers,

, Felix,

E.R.

, Finnerup,

N.B.

, ... Widerstrom-Noga,

(2007). Pain after spinal cord injury: An evidence-based review for clinical practice and research. Report of the National Institute on Disability and Rehabilitation Research Spinal Cord Injury Measures meeting. The Journal of Spinal Cord Medicine, 30(5), 421–440.

Campbell,

L.C.

, Clauw,

D.J.

, Keefe,

F.J.

(2003). Persistent pain and depression: A biopsychosocial perspective. Biological Psychiatry, 54(3), 399–409.

Defrin,

, Grunhaus,

, Zamir,

, Zeilig,

(2007). The Effect of a Series of Repetitive Transcranial Magnetic Stimulations of the Motor Cortex on Central Pain After Spinal Cord Injury. Archives of Physical Medicine and Rehabilitation, 88(12), 1574–1580.

Dworkin,

R.H.

, Turk,

D.C.

, Farrar,

J.T.

, Haythornthwaite,

J.A.

, Jensen,

M.P.

, Katz,

N.P.

, ... Witter,

(2005). Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain, 113(1-2), 9–19.

Freund,

, Weiskopf,

, Ward,

N.S.

, Hutton,

, Gall,

, Ciccarelli,

, ... Thompson,

A.J.

(2011). Disability, atrophy and cortical reorganization following spinal cord injury. Brain, 134(Pt 6), 1610–1622.

Gao,

, Chu,

, Li,

, Yang,

, DU

, Li,

, ... Chan,

(2017). Repetitive transcranial magnetic stimulation for pain after spinal cord injury: A systematic review and meta-analysis. Journal of Neurosurgical Sciences, 61(5), 514–522.

10.

Garcia-Larrea,

, Peyron,

(2013). Pain matrices and neuropathic pain matrices: A review. Pain, 154, S29–S43.

11.

Gustin,

S.M.

, Wrigley,

P.J.

, Henderson,

L.A.

, Siddall,

P.J.

(2010). Brain circuitry underlying pain in response to imagined movement in people with spinal cord injury. Pain, 148(3), 438–445.

12.

Hirayama,

, Saitoh,

, Kishima,

, Shimokawa,

, Oshino,

, Hirata,

, ... Yoshimine,

(2006). Reduction of intractable deafferentation pain by navigation-guided repetitive transcranial magnetic stimulation of the primary motor cortex. Pain, 122(1), 22–27.

13.

Jetté,

, C Té,

, Meziane,

H.B.

, Mercier,

(2013). Effect of Single-Session Repetitive Transcranial Magnetic Stimulation Applied Over the Hand Versus Leg Motor Area on Pain After Spinal Cord Injury. Neurorehabilitation and Neural Repair, 27(7), 636–643.

14.

Jiang,

, Ji,

, Voulalas,

P.J.

, Keaser,

, Xu,

, Gullapalli,

R.P.

, ... Masri,

(2014). Motor Cortex Stimulation Suppresses Cortical Responses to Noxious Hindpaw Stimulation After Spinal Cord Lesion in Rats. Brain Stimulation, 7(2), 182–189.

15.

Kang,

B.S.

, Shin,

H.I.

, Bang,

M.S.

(2009). Effect of repetitive transcranial magnetic stimulation over the hand motor cortical area on central pain after spinal cord injury. Archives of Physical Medicine and Rehabilitation, 90(10), 1766–1771.

16.

Lefaucheur,

J.P.

, Drouot,

, Menard-Lefaucheur,

, Keravel,

, Nguyen,

J.P.

(2006). Motor cortex rTMS restores defective intracortical inhibition in chronic neuropathic pain. Neurology, 67(9), 1568–1574.

17.

Lefaucheur,

J.P.

, Drouot,

, Menard-Lefaucheur,

, Zerah,

, Bendib,

, Cesaro,

, ... Nguyen,

J.P.

(2004). Neurogenic pain relief by repetitive transcranial magnetic cortical stimulation depends on the origin and the site of pain. Journal of Neurology Neurosurgery and Psychiatry, 75(4), 612–616.

18.

Li,

, Zhu,

, Gao,

, Xu,

, Li,

, Hu,

, ... Konstruktion,

E. T. O.

(2015). Using functional near-infrared spectroscopy (fNIRS) to detect the prefrontal cortical responses to deception under different motivations. Biomedical Optics Express, 6(9), 3503–3514.

19.

Long,

, Wang,

, Zhao,

, Duan,

, Feng,

, Hui,

, ... Yuan,

(2018). Effects of combining high- and low-frequency repetitive transcranial magnetic stimulation on upper limb hemiparesis in the early phase of stroke. Restorative Neurology and Neuroscience, 36(1), 21–30.

20.

Mishra,

B.R.

, Nizamie,

S.H.

, Das,

, Praharaj,

S. K.

(2010). Efficacy of repetitive transcranial magnetic stimulation in alcohol dependence: A sham-controlled study. Addiction, 105(1), 49–55.

21.

Miyai,

, Yagura,

, Hatakenaka,

, Oda,

, Konishi,

, ... Kubota,

(2003). Longitudinal Optical Imaging Study for Locomotor Recovery After Stroke. Stroke, 34(12), 2866–2870.

22.

Moreno-Duarte,

, Morse,

L.R.

, Alam,

, Bikson,

, Zafonte,

, ... Fregni,

(2014). Targeted therapies using electrical and magnetic neural stimulation for the treatment of chronic pain in spinal cord injury. NeuroImage, 85, 1003–1013.

23.

Morishita,

, Hyakutake,

, Saita,

, Takahara,

, Shiota,

, ... Inoue,

(2015). Pain reduction associated with improved functional interhemispheric balance following transcranial direct current stimulation for post-stroke central pain: A case study. Journal of the Neurological Sciences, 358(1-2), 484–485.

24.

Moseley,

G.L.

, Flor,

(2012). Targeting cortical representations in the treatment of chronic pain: A review. Neurorehabilitation and Neural Repair, 26(6), 646–652.

25.

Nardone,

, Höller,

, Leis,

, Höller,

, Thon,

, Thomschewski,

, ... Trinka,

(2014). Invasive and non-invasive brain stimulation for treatment of neuropathic pain in patients with spinal cord injury: A review. The Journal of Spinal Cord Medicine, 37(1), 19–31.

26.

Nees,

T.A.

, Finnerup,

N.B.

, Blesch,

, Weidner,

(2017). Neuropathic pain after spinal cord injury. Pain, 158(3), 371–376.

27.

Peeters,

R.R.

, Tindemans,

, De Schutter,

, Van der Linden,

(2001). Comparing BOLD fMRI signal changes in the awake and anesthetized rat during electrical forepaw stimulation. Magnetic Resonance Imaging, 19(6), 821–826.

28.

Peyron,

, Schneider,

, Faillenot,

, Convers,

, Barral,

, Garcia-Larrea,

, ... Laurent,

(2004). An fMRI study of cortical representation of mechanical allodynia in patients with neuropathic pain. Neurology, 63(10), 1838–1846.

29.

Quan,

W.X.

, Zhu,

X.L.

, Qiao,

, Zhang,

W.F.

, Tan,

S.P.

, Zhou,

D.F.

, ... Wang,

X.Q.

(2015). The effects of high-frequency repetitive transcranial magnetic stimulation (rTMS) on negative symptoms of schizophrenia and the follow-up study. Neuroscience Letters, 584, 197–201.

30.

Ren,

, Lu,

, Liu,

, Shen,

, Zhang,

, Ma,

, ... Liu,

(2017). Decreased prefrontal brain activation during verbal fluency task in patients with somatoform pain disorder: An exploratory multi-channel near-infrared spectroscopy study. Progress in Neuro-psychopharmacology & Biological Psychiatry, 78, 153–160.

31.

Rossini,

P.M.

, Burke,

, Chen,

, Cohen,

L.G.

, Daskalakis,

, Di Iorio,

, ... Ziemann,

(2015). Non-invasive electrical and magnetic stimulation of the brain, spinal cord, roots and peripheral nerves: Basic principles and procedures for routine clinical and research application. An updated report from an I.F.C.N. Committee. Clinical Neurophysiology, 126(6), 1071–1107.

32.

Schecklmann,

, Giani,

, Tupak,

, Langguth,

, Raab,

, Polak,

, ... Fallgatter,

A.J.

(2014). Functional Near-Infrared Spectroscopy to Probe State- and Trait-Like Conditions in Chronic Tinnitus: A Proof-of-Principle Study. Neural Plasticity, 2014, 1–8.

33.

Tazoe,

, Perez,

M.A.

(2015). Effects of repetitive transcranial magnetic stimulation on recovery of function after spinal cord injury. Archives of Physical Medicine and Rehabilitation, 96(4 Suppl), S145–S155.

34.

Teitti,

, Määttä,

, Säisänen,

, Könönen,

, Vanninen,

, Hannula,

, ... Karhu,

(2008). Non-primary motor areas in the human frontal lobe are connected directly to hand muscles. NeuroImage, 40(3), 1243–1250.

35.

Topka,

, Cohen,

L.G.

, Cole,

R.A.

, Hallett,

(1991). Reorganization of corticospinal pathways following spinal cord injury. Neurology, 41(8), 1276–1283.

36.

Treede,

R.D.

, Jensen,

T.S.

, Campbell,

J.N.

, Cruccu,

, Dostrovsky,

J.O.

, Griffin,

J.W.

, ... Serra,

(2008). Neuropathic pain: Redefinition and a grading system for clinical and research purposes. Neurology, 70(18), 1630–1635.

37.

Widerström-Noga,

(2017). Neuropathic Pain and Spinal Cord Injury: Phenotypes and Pharmacological Management. Drugs, 77(9), 967–984.

38.

Yılmaz,

, Kesikburun,

, Yaşar

, Tan,

A. K.

(2013). The effect of repetitive transcranial magnetic stimulation on refractory neuropathic pain in spinal cord injury. The Journal of Spinal Cord Medicine, 37(4), 397–400.

39.

Zhu,

, Li,

, Fan,

, Li,

, Huang,

, He,

, ... Konstruktion,

E. T. O.

(2015). Atypical prefrontal cortical responses to joint/non-joint attention in children with autism spectrum disorder (ASD): A functional near-infrared spectroscopy study. Biomedical Optics Express, 6(3), 690–701.