Keywords

1 Introduction

Transcranial Magnetic Stimulation (TMS) is a neurophysiological technique that allows a noninvasive, painless stimulation of the human brain through the intact scalp.

Different brain areas can be targeted by TMS, depending on the position of the coil. TMS effects on motor areas have been better characterized compared to nonmotor areas since the output produced by the stimulation of the primary motor area of one side can be easily recorded from muscles of the contralateral side of the body.

The application of noninvasive TMS to the human brain for assessing central motor pathways was described for the first time in 1985, in the Lancet journal, by A.T. Barker, R. Jalinous and I.L. Freeston, from the University of Sheffield [1].

The new TMS technique had a unique potential and some advantages compared to noninvasive transcranial electrical stimulation (TES), which was developed in 1980 by P.A. Merton and H.B. Morton [2]. Compared to TMS, TES requires high current densities to overcome the skull and to generate action potentials, resulting in painful and low tolerable stimulation.

The interest in TMS raised during the years and a consistent number of studies on this topic have advanced our knowledge of the human brain [3], even if many limitations exist due to the artificial nature of the stimulation. So far, many protocols of TMS stimulation have been tested and described, and different cortical circuits activated by TMS have been characterized [4, 5]. TMS can be used alone or in combination with other techniques in order to test corticospinal and cortico-cortical connectivity and brain plasticity, to map brain functions, and study specific cortical functions by inducing a “virtual lesion” in a targeted area [6,7,8].

A milestone in TMS history has been the demonstration that protocols based on repetitive TMS (rTMS) can induce prolonged effects, which outlast the period of stimulation [9, 10]. This evidence opened exciting research and clinical scenarios in which rTMS protocols are used for neuromodulatory/therapeutic purposes.

To date, TMS has a recognized role in the clinical and research settings. Stimulation protocols have been standardized, and safety limits of TMS stimulation have been established [11, 12]. Indeed, specific rTMS protocols received Food and Drug Administration (FDA) approval for the treatment of drug-resistant unipolar major depression.

In this chapter, we will review the evidence and the hypotheses on the neurophysiological bases and on the mechanisms of action of TMS, focusing on TMS application to the primary motor cortex.

2 How TMS Is Delivered

TMS is based on the Faraday's principle of electromagnetic induction, according to which a time-varying magnetic field will induce an electric current [13]. In TMS, a brief electric current is delivered through a capacitor to a coil, made of loops of copper wire embedded in a plastic case. Perpendicularly to the coil plane, a focal magnetic field is induced, which penetrates the scalp and the skull without attenuation and generates an electric current. If sufficiently strong, the induced electric current will change the electrical potential of the conductive superficial neuronal membranes leading to an action potential [14, 15].

The most widespread TMS devices can provide monophasic or biphasic pulse shapes with a determined width. More recently, TMS devices with controllable pulse parameters have been introduced [16].

Different types of coil exist, for superficial and deep targets of stimulation, and their effects have been modelled [17, 18]. Among the most frequently used coils, there are the figure-of-eight coil (which induces a more focal stimulation) and the circular coil (which induces a nonfocal stimulation of the brain) [4].

Focal coils can be oriented so as to induce currents in the brain with different directions: more commonly, the coil is kept perpendicularly to the central sulcus, and a posterior-to-anterior (PA) directed current is induced in the brain.

TMS spatial resolution and corticospinal output vary depending on several factors, including the shape of the stimulating coil, its position above the scalp, coil orientation, stimulation intensity, pulse waveform, ongoing voluntary muscle contraction, and other variables [19,20,21,22].

3 Single-Pulse TMS

The responses that can be recorded at the muscular level after TMS are named as motor-evoked potentials (MEPs) [1, 23,24,25] (Fig. 2.1). The optimal scalp location to evoke MEPs in the targeted muscle is defined as “hot-spot”, while the minimum TMS stimulation intensity able to elicit consistent MEPs (with peak-to-peak amplitudes of at least 50 μV in each trial) in at least 5 out 10 consecutive TMS stimuli at rest is defined as resting motor threshold or RMT [12]. For each MEP, objective measures such as onset latency, peak latency, amplitude, and area can be obtained (Fig. 2.2). MEP amplitude, usually measured peak-to-peak, has an intrinsic variability of multifactorial origin [26, 27]. The mechanisms through which primary motor cortex TMS produces MEPs are partially understood due to the complexity of cortical circuits and the difficulty in assessing the interactions between the induced current in the brain and the neural networks, which are composed of different cell types, with different orientations and sizes. The physiological effects produced by motor cortex stimulation have been characterized first in animals, using direct electrical stimulation of the motor cortex together with the direct recording of the evoked corticospinal activity from the high cervical cord. These recordings revealed that a single electrical stimulus delivered to the motor cortex could produce a high-frequency (>600 Hz) repetitive discharge of corticospinal axons originating both from direct and indirect activation of corticospinal cells [28,29,30]. The earliest wave that is still recordable after cerebral cortex ablation was thought to originate from direct activation of the corticospinal axons and has therefore been termed the “D” wave [29]. The following waves that require the integrity of the cerebral cortex were thought to originate from indirect, trans-synaptic, activation of corticospinal neurons and were termed “I” waves. They were numbered in order of their appearance (I1, I2, I3, …). The interval between I-waves is about 1.5 ms, which corresponds to a discharge frequency of about 600 Hz. The same high-frequency corticospinal activity was subsequently recorded in humans after motor cortex TMS through epidural high cervical electrodes implanted for the treatment of chronic pain. This unique setting has provided relevant insight [31]. Indeed, it has been shown that also in humans the TMS-induced corticospinal descending activity is made by multiple descending high-frequency waves. Several studies showed that the composition of the corticospinal volleys in terms of D- and I-waves is influenced by the parameters of stimulation (stimulation intensity, coil type, and coil orientation) and by changes in cortical excitability (e.g., changes induced by voluntary contraction) [31, 32]. When the stimulating coil is aligned to induce a current perpendicularly to the line of the central sulcus (approximately posterior–anterior in the brain; PA), TMS evokes the earliest trans-synaptic response that, in analogy with animal recordings, is termed I1-wave. At higher intensities, this wave is followed by later waves numbered in order of their appearance (I2, I3, etc.) [31]. Only at very high stimulus intensity, a short-latency D-wave is evoked. When the induced current flows parallel to the line of the central sulcus (approximately lateral-to-medial in the brain; LM), only a D-wave is preferentially recruited. If the orientation of the induced current is kept perpendicular to the line of the central sulcus, but it is reversed (approximately anterior–posterior in the brain; AP), the evoked activity is less synchronized, with some later peaks of latencies compared to those of the I-waves evoked by PA stimulation [31]. Similar findings have been obtained with biphasic stimulation (a PA-induced current followed by an AP-induced current): using biphasic TMS discharges, a corticospinal activity with a frequency that is half of that of the I-waves (about 330 Hz) has been recorded in some patients [4] (Fig. 2.3). These findings suggest that motor cortex TMS may activate not only the corticospinal neurons responding with a high-frequency discharge at I-wave frequency, but also different populations of corticospinal neurons responding at lower frequencies. However, these activities are usually not evident in volleys recorded at the epidural level because, as in animals, these volleys are dominated by fast conducting axons whose discharge is larger and more synchronous, particularly at high stimulation intensity. Only at lower intensities, different corticospinal outputs can be detected. Indeed, at high intensities of stimulation, the high-frequency I-waves represent the only output that is recorded with all the directions of the induced current in the brain and by both focal and nonfocal coils [4, 31] (Fig. 2.3).

Fig. 2.1
figure 1

TMS-induced responses at different recording levels

Full size image
Fig. 2.2
figure 2

Motor-evoked potential (MEP) elicited by single-pulse Transcranial Magnetic Stimulation (TMS) at 110% resting motor threshold (RMT) intensity, recorded from superficial electromyography (EMG) at the level of the contralateral first dorsal interosseous muscle

Full size image
Fig. 2.3
figure 3

Epidural recordings from the cervical cord of descending volleys evoked by lateromedial (LM), posterior–anterior (PA), anterior–posterior (AP), or biphasic (PA-AP) transcranial magnetic stimulation (TMS) at low and high intensity in patients with cervical epidural electrodes. At lower intensities of stimulation, the different orientations of the induced current evoke different corticospinal activities: LM TMS evokes D-waves; PA TMS elicits three I-waves; AP TMS evokes a dispersed activity, and no clear waves can be identified; biphasic TMS (PA followed by AP) evokes longer latency and lower frequency I-waves. At high intensity, all the directions of the induced current only evoke the high-frequency I-waves

Full size image

Thus, the direct recording of corticospinal activity in humans and in animals demonstrates that different activities can be produced by transcranial stimulation, suggesting the presence of multiple independent cortical circuits within the motor cortex projecting to the lower motor neurons [4].

Interestingly, the simultaneous recording of TMS and electroencephalography (EEG), known as TMS-EEG, is emerging as a very useful clinical tool to assess cortico-cortical connectivity together with corticospinal connectivity. In this case, the TMS-evoked responses are recorded through the EEG electrodes as positive and negative deflections in the EEG signal and are called TMS-evoked potentials (TEPs) [33].

4 Paired-Pulse Stimulation

In paired-pulse TMS protocols, pairs of stimuli are delivered using two connected TMS stimulators. Depending on the interstimulus interval and stimulus intensity, the interaction between pairs of stimuli delivered to the primary motor cortex can be inhibitory or facilitatory, as assessed by MEP amplitude.

Specific paired-pulse TMS protocols have been described. Among the most frequently used in research, for their proposed role as an indirect measure of interneuronal function, there are the short-interval intracortical inhibition (SICI) and the intracortical facilitation (ICF) protocols. SICI and ICF are elicited by pairing a subthreshold conditioning stimulus and a suprathreshold test stimulus, delivered at 1–5 ms (SICI) or 8–30 ms (ICF) interstimulus interval (ISI), respectively. The result is a suppression (SICI) or a facilitation (ICF) of MEP amplitude [34, 35]. SICI has been mainly related to the activation of GABA-A receptors and to a reduction of late I-waves [36,37,38], while ICF has been in part attributed to glutamatergic NMDA receptor activation, even if it is less well understood [39, 40]. Other paired-pulse protocols are the short-interval intracortical facilitation (SICF) and the long-interval intracortical inhibition (LICI) (for more details see [4]).

Several other TMS protocols are used in research, being TMS a very versatile tool. These protocols include the interhemispheric inhibition (IHI), in which two TMS coils (one for each hemisphere) are used, and the very interesting protocols in which TMS is paired with peripheral electrical stimulation: short-latency afferent inhibition (SAI), long-latency afferent inhibition (LAI), and paired associative stimulation (PAS). For a more comprehensive list and description of TMS protocols, see [12]. Interestingly, epidural recordings in humans have shown that inhibitory protocols only suppress the later components of the corticospinal volley with no effect on the I1-wave [4]. This observation provides further support to the existence of independent cortical circuits producing different corticospinal activities with only some of them under a GABAergic inhibitory control.

5 Repetitive TMS (rTMS)

In rTMS, a repetitive stimulation, with biphasic or monophasic stimuli, is delivered over the scalp. rTMS targeting primary motor area showed to be able to induce prolonged effects on corticospinal excitability, which outlasted the stimulation from several minutes to some hours [9, 41]. The mechanisms underlying rTMS effects are still largely unknown. rTMS application on motor areas is commonly studied through the analysis of MEPs size before and after rTMS stimulation. In contrast, rTMS effects over nonmotor areas have more indirect outcome measures, including EEG and MRI connectivity measures and behavioral tests, whose interpretation requires more caution.

To date, existing evidence suggests that rTMS might induce changes in cortical and subcortical neurotransmitter release, with consequent prolonged changes in synaptic activity [42, 43].

rTMS applied to the dorsolateral prefrontal cortex (DLPFC), as in the treatment of depression, is thought to act not only on the stimulated area but also in distant regions, which are anatomically and/or functionally connected [44, 45].

rTMS classical protocols include low-frequency (LF) rTMS (<1 Hz) and high-frequency (HF) rTMS (>1 Hz). Other popular rTMS protocols are the continuous theta-burst stimulation (cTBS) and the intermittent theta-burst stimulation (iTBS) (Fig. 2.4). Classically, LF rTMS and cTBS were considered inhibitory protocols, able to induce long-term depression (LTD)-like plasticity, whereas HF rTMS and iTBS were considered excitatory protocols, able to induce long-term potentiation (LTP)-like plasticity [9]. However, it is now known that their effect is mixed and it depends on many variables, including the number of stimuli [46, 47], the intensity of stimulation, and the baseline cortical activation state [9, 48]. The after-effects of the different rTMS protocols are commonly described in terms of the changes that are produced in threshold or size of evoked MEPs, and the different protocols are simply classified as inhibitory or facilitatory, assuming that the physiological basis of all the inhibitory and of all the excitatory protocols are similar. However, epidural recordings in humans, performed before and after different rTMS protocols, have shown that, even though most protocols selectively modulate the late components of the corticospinal volleys, some of them could selectively modulate the earliest component or the inhibitory cortical circuits [25]. Thus, epidural recordings revealed that the effects of different protocols on cortical circuits are not homogeneous and that distinct protocols can modulate specific neural elements in distinct layers of the cortex. Different patterns of modulation have been demonstrated: (1) the most commonly observed change after rTMS is a selective modulation of late I-waves with no change in the amplitude of the I1-wave (i.e., inhibition is obtained after low-frequency rTMS (1 Hz), while a selective enhancement of late I-waves with no change in the amplitude of the I1-wave is observed after iTBS). This pattern indicates a more pronounced effect on cortico-cortical interneurons projecting on corticospinal cells with no change in the excitability of corticospinal cells; (2) after high-frequency rTMS (5 Hz), all the volleys are enhanced including the D-wave. This pattern highlights how that the excitability of corticospinal neurons is enhanced; (3) the cTBS protocol suppresses the I1-wave selectively, while later I-waves are much less affected. This suggests that cTBS has its major effect on a single source of inputs to corticospinal cells, which is responsible for the I1-wave production; (4) a very low-intensity and high-frequency stimulation has no effect on corticospinal volleys but suppresses intracortical inhibitory activity, as evaluated with paired-pulse stimulation, suggesting that this form of stimulation selectively modulates the excitability of GABAergic inhibitory networks in the motor cortex [25]. Thus, epidural recordings have shown that it might be possible to modulate specific cortical circuits using rTMS, and this could be extremely relevant because neural circuits that are differentially affected in various neuropsychiatric disorders can be targeted quite selectively with rTMS.

Fig. 2.4
figure 4

Protocols of repetitive Transcranial Magnetic Stimulation (rTMS). Cf. text for details

Full size image

Extensive evidence supports the potential therapeutic applications of rTMS in specific neurological and psychiatric disorders [9].

The main clinical application of rTMS is drug-resistant unipolar major depression, for which rTMS received FDA approval in 2008. The optimal stimulation parameters for a safe and effective administration of rTMS in the treatment of depression have been recently reviewed [49]. The standard rTMS protocol used for the treatment of depression is the 10 Hz stimulation (trains of 4-second duration, with an intertrain interval of 26 seconds) delivered through a figure-of-eight coil, over the left DLPFC at an intensity of 120% relative to RMT. The total number of pulses per session is 3000. Each session lasts about 37 minutes. The total number of sessions is 20 (5 working days/week for 4 consecutive weeks).

In 2018, a randomized noninferiority trial, which included more than 400 patients (the largest trial of brain stimulation ever done), demonstrated that iTBS effectiveness is noninferior to that of the 10 Hz treatment, with very similar tolerability and safety profiles [50].

Since one iTBS session has a duration of about 3 minutes, approximately 10 times shorter than the standard 10 Hz rTMS session, the new protocol is advantageous in practical terms. However, the total number of sessions tested in the trial is still 20, which requires high patients’ compliance.

Systematic clinical studies are still needed to define all the clinical indications of therapeutic rTMS and to identify effect predictors. Further research is also needed to clarify the mechanisms of action and to optimize the stimulation parameters.