Transcranial magnetic stimulation (TMS) is based on Faraday’s principle of electromagnetic induction. A TMS coil is placed on the scalp, and current is passed through it that generates a rapidly fluctuating magnetic field that in turn induces electrical current flow toward the intracranial space. Activities of the cortical neurons are modulated by the electrical field produced by the magnetic field.8 There are several configurations of TMS coils, with circular and figure-of-eight coils being the two main models. Circular coils are the simplest, and they provide diffuse stimulation. In contrast, figure-of-eight or ‘butterfly’ coils are the most commonly used because they can induce highly focused electrical fields.

The five parameters of location, focality, frequency, pulse intensity, and duration should be considered when conducting TMS.8 There are several ways to ensure that the correct location is stimulated. When the eloquent cortex is stimulated, the induced response is expected to be based on functional mapping. A single pulse of TMS will induce muscle activation by stimulating the primary motor cortex (M1). Perception of a phosphene is expected if the primary visual cortex (V1) is stimulated. Stereotactic navigation techniques been recently become popular for stimulating more-complex cortical areas, such as the dorsolateral prefrontal cortex (DLPFC). The depth and focality of stimulation are critical functional parameters, and they are interlinked, with an increased depth resulting in reduced focality. The pulse intensity is correlated with the depth of stimulation, with a larger pulse intensity increasing the depth but reducing the focality.

TMS can modulate brain activities in various ways depending on the stimulation pattern and frequency. Single-pulse TMS can induce an immediate response, but it exerts no sustained effect on the stimulated cortex. This modality is normally used to record motor evoked potentials and to determine the motor threshold (MT).8 Paired-pulse TMS applies two pulses in rapid succession, and is used in basic and clinical neurophysiological research to measure inhibitory and excitatory central mechanisms.6,9 In contrast, rTMS can induce long-lasting neuroplastic changes, and is used for treating several neurological diseases. Low-frequency (1 Hz) stimulation has an inhibitory effect, while high-frequency (≥ 5 Hz) stimulation has an excitatory effect on the underlying cortex. Theta-burst stimulation (TBS) is another form of rTMS, which also has two protocols: (1) continuous TBS with an inhibitory effect and (2) intermittent TBS with an excitatory effect.10,11 TBS should induce longer lasting neuroplasticity compared with rTMS.8

The mechanism underlying the effects of applying rTMS to M1 in treating neuropathic pain is unclear. The motor cortex has connections to the thalamus,12 and motor cortex stimulation can inhibit thalamic activity that then indirectly activates the descending pain inhibitory pathway.13 These connections might explain the analgesic effect of rTMS on M1 against neuropathic pain.14 Another postulated mechanism of action is endogenous opioid release induced by M1 stimulation.14 A positron-emission tomography (PET) study evaluating the activity of opioid receptors found decreased signals after 7 months of direct electrical stimulation to M1, suggesting that endogenous opioids released by M1 stimulation could occupy opioid receptors.15 A recent PET study also demonstrated significant release of endogenous opioids within a hemispheric brain network by applying single session of 10-Hz rTMS to M1 in 10 healthy subjects.16

The International Federation of Clinical Neurophysiology and the European Federation of Neurological Societies reached a consensus on the therapeutic use of rTMS and reported evidence-based guidelines in 2014.17 They subsequently updated the recommendations taking into account all publications related to rTMS, including data prior to 2014 as well as currently reviewed literature until the end of 2018.7 They recommended level A evidence (definite efficacy) for high-frequency rTMS of M1 contralateral to the painful side for neuropathic pain. High-frequency rTMS of the left M1 or DLPFC for improving the quality of life or pain was relevant for level B evidence (probable efficacy).7

Since the early report on the effects of M1 stimulation by Tsubokawa et al.,18 many studies have been performed to define the appropriate rTMS parameters for treating neuropathic pain. Most studies demonstrating good efficacy of rTMS against neuropathic pain have applied similar rTMS parameters: high-frequency stimulation (10-20 Hz), 1,600 to 3,000 stimulations per session, intertrain interval of 10 to 84 seconds, intensity of 70% to 90% of the MT, and using a figure-of-eight coil (Table 1). The most significant results for the good efficacy of rTMS against neuropathic pain were obtained following M1 stimulation.

The most-effective M1 location at which to apply stimulation remains to be defined. Andre-Obadia et al.19 reported that rTMS was more efficacious when it is delivered over the hand motor area than the face area regardless of the pain location (hand or face). However, Ayache et al.20 showed that anatomical targeting using magnetic-resonance-imaging-guided navigation may provide a better effect than the motor hotspot. That study included patients with neuropathic pain at various locations, which is more likely to reflect real situations.

The second-most-studied cortical target for rTMS in chronic pain has been the DLPFC. Unfortunately, rTMS of the DLPFC showed a very weak effect on chronic pain in a recent randomized multicenter sham-controlled trial.21 Parieto-opercular cortex stimulation for treating orofacial pain produced significantly better analgesia than stimulation of the primary sensorimotor cortex or sham.22,23 However, these studies were based on the short-term effect of a single rTMS session, and so long-term results need to be investigated further. Stimulating other cortical areas such as insular cortex24 or middle cingulate cortex25 failed to improve chronic neuropathic pain.

rTMS is relatively safe. The most-common adverse effects of rTMS are headache, pain at the site of stimulation, increased bodily pain or paresthesia, fatigue, sleep disorders, nausea, anxiety/irritability/cognitive impairment, and muscle sensations during the stimulation.21 No serious treatment-related adverse effect has been reported. Serious adverse events such as seizures are rare, especially when utilizing focused coils and an intensity of 70% to 80% of MT.

Long-term effects of rTMS for maintenance treatment have not been clearly elucidated. The cumulative effect of rTMS sessions over the long term has been reported for various protocols of maintenance treatment. However, further study is needed to clarify the long-term effects of rTMS for maintenance treatment.26,27

Identifying new stimulation targets for rTMS would also be useful. The posterior insular cortex and anterior cingulate cortex are involved in central integration of the sensory-discriminative aspect of pain processing.28 Deep brain stimulation of the posterior insular can lead to antinociceptive effects dependent on endogenous opioids and cannabinoids in experimental models of peripheral neuropathic pain.29,30

Regarding the mechanism of action, TBS is a better method than high-frequency rTMS for inducing long-lasting effects with a lower stimulation intensity. However, published data on TBS have only been related to experimental or acute provoked pain,31,32 and so further studies of TBS for chronic neuropathic pain treatment are needed.