Introduction

Rehabilitation of patients with severe traumatic brain injury (BMI) is an urgent medical and social problem, since this pathology remains one of the main causes of mortality and disability of the population of young working age [1]. The most common consequences of PMTCT include motor and cognitive impairment, as well as depression of consciousness [2, 3]. Over the past decade, transcranial electrical and magnetic stimulation (TMS) has been increasingly used as a neuromodulatory effect in clinical practice [4—12].

Historical reference

TMS has been used as a diagnostic tool since 90s of the twentieth century. This method is recognized as informative procedure for assessing anatomical integrity of corticospinal tract and functional state of human nervous system. The first foreign and national researches of TMS concerned analysis of motor cortex excitability and conduction along corticospinal pathways [4, 13, 14]. These studies showed that TMS is sensitive and perspective method, its diagnostic and curative capabilities are much wider and also related to functional assessment of visual cortex, speech and memory centers in both neurological and mental diseases [15]. TMS may be valuable to evaluate integrity and functional state of nervous pathways in various forms of central nervous system pathology [16]. Moreover, stimulation of different cortical areas in a healthy person and analysis of concomitant functional effects are useful to clarify structural and functional organization of the brain [17]. The last one is important for theoretical and scientific purposes.

Transcranial and peripheral rhythmic magnetic stimulation was observed to be effective curative measure besides diagnosis [18].

Currently, rhythmic TMS (rTMS) is one of the most common and popular procedures not only in neurology, but also in neurobiology as a whole. Significant increase of interest in this approach in recent years is shown in Fig. 1.

According to the PubMed database, the number of reports devoted to rTMS only for neurological diseases has been increased from single ones to 360—400 per year since 1988 (Fig. 1). Reviews account up to 20% of these publications (Fig. 2).

This report is based on analysis of the most significant publications and recommendations recognized by the scientific community. We also consider the reports of national and foreign authors and own experience of transcranial stimulation.

The use of rTMS is comprehensively described in reviews [4, 17, 18, 19]. rTMS is effective for relief of various neurogenic pain syndromes including phantom pain, post-stroke thalamic pain, compression cervical and lumbar pain syndromes, etc. [16, 18, 20]. Exposure points are various parts of the brain and spinal segments, as well as peripheral nervous system depending on zone of interest and expected functional effects. This method is used in the treatment of depression, pathological neurogenic tinnitus, epilepsy, migraine and migraine-like symptoms [4]. TMS is also applied to alleviate clinical manifestations of Parkinson's disease [21, 22]. Antispastic effect of some modes of magnetic stimulation is used for the management of spasticity, overactive bladder syndrome, hemifacial spasm, cerebral palsy and other diseases with similar pathogenesis [4]. Perhaps, rTMS is very actively used for the treatment of depression [23], neurogenic pain syndromes, as well as to compensate motor deficiency in patients with limb paresis [24]. There are effective rTMS modes for treatment of consequences of stroke [4, 25—27], including protocols incorporated into international guidelines [28]. rTMS protocols for head injury are less clear [5, 29, 30]. rTMS for restoration of patients with impaired consciousness after severe TBI is being studied [7, 29, 25, 31—33].

TMS mechanisms

Therapeutic effect of rTMS is based on change of cortical excitability, depolarization of neurons with subsequent appearance and distribution of action potential (AD) [4, 13]. TMS combines the advantages of non-invasiveness and painlessness. Moreover, this method is also characterized by more local effect compared to electrical stimulation [13, 19, 34]. More recent studies revealed that TMS affects some other processes such as redistribution of cerebral blood flow [36], release of brain-derived neurotrophic factor and dopamine, activity of some enzymes [17] besides excitation and inhibition per se [4, 35]. Perhaps, “non-electric” effects of TMS are associated with its initial action at the molecular, atomic and subatomic (quantum) levels [37, 38].

Rhythmic TMS allows non-invasive stimulation of brain neuronal structures with short magnetic pulses. Comprehensive studies using EEG and diffusion-tensor magnetic resonance imaging (DT MRI) showed improved functional and structural communications and integrative activity after brain stimulation [39].

Modern devices are capable to create 4 T electromagnetic fields and stimulation frequency over 100 Hertz. Magnetic impulse is generated in a plane perpendicular to the coil and non-invasively penetrates through the body’s tissues (skin, muscles, bone, etc.) to a depth of 10—15 centimeters. This impose causes own excitation of neuronal structures with subsequent spread of action potential [24].

Methodological features of rTMS

The effects of rTMS are determined by characteristics of electromagnetic field of the inductors. These features largely depend on the shape, dimensions, design of electromagnetic coil and its orientation in relation to the patient’s head. Annular, 8- and H-shaped coils of various diameters are the most common. Annular coils result quite powerful, but relatively poorly localized stimulation. At the same time, dual 8-shaped coils produce more focal (with an accuracy of 1 cm) but slightly less powerful pulses. Stimulation of the structures close to the scalp surface, for example cerebral cortex or cerebellum, may be performed using 8-shaped coil. In case of tangential placement of the coil in relation to scalp, the probability of stimulation is maximum in those areas oriented parallel to central segments of the coil [19]. Stimulation of larger areas is carried out using annular coils. Special H-shaped coils are used for stimulation of deep brain structures (hippocampus, subcortical structures, brain stem) [40]. Dual coils have recently appeared. In this devices, single Faraday’s coils are placed under certain angle to each other rather in parallel fashion. Development of stimulation technique and new coils for TMS is continued [4, 41].

Shape of magnetic pulse is important physical characteristic. It is monophasic or biphasic as a rule [34]. Monophasic impulses are preferable for basic researches, since it is believed that such stimuli excite relatively uniform columns of neurons in certain zone. Thus, more clear and obvious functional effects are obtained due to less involvement of adjacent structures [4, 13]. However, generation of this impulse requires complex magnetic stimulator. In addition, significant energy consumption also limits the widespread use of monophasic TMS.

Currently, foreign and national manufacturers offer various modifications of magnetic stimulators. It is necessary to emphasize separately the navigation systems for TMS. These systems are used to visualize stimulation zones. These complexes make it possible to load previously performed MRI scans of the brain into the system with subsequent 3D reconstruction using special cameras and sensors on the patient’s head and the coil. This system allows on-line virtual projection of coil position on the surface of cerebral cortex with fairly high accuracy after computer processing and special calibration to take into account errors and soft tissue thickness. These complexes open up new possibilities for both scientific researches and precise clinical use of TMS. The most significant drawbacks of such systems are extremely high cost (ten times more expensive than conventional “blind” TMS systems) and need to involve large personnel and time resources (specialists for radiation diagnosis). Until now, above described TMS navigation systems are used only in few laboratories [17, 27, 32].

Recommendations for therapeutic rhythmic transcranial magnetic stimulation: accumulated questions

Most of TMS effects as well as basic ideas about nature of impulses were obtained in researches of motor cortex in animals and, subsequently, humans as the most accessible system for evaluation and analysis. Clear, topographically determined motor response, determining its threshold for selection of individual stimulus power and relatively good reproducibility of the effects justified the choice of this zone for stimulation. Current guidelines [4, 13, 42] concern mainly diagnostic and therapeutic stimulation of motor cortex. At the same time, significant problems are associated with sensitivity threshold determination and stimulation intensity selection for other brain structures in the absence of a distinct functional (behavioral) response. Analysis of bioelectrical activity of the brain during rTMS may be perspective solution of this problem [11, 36, 43]. However, this method is not generally accepted.

Probably for this reason, some researchers use TMS parameters effective for the motor cortex during stimulation of various cortical zones that is not entirely correct. The problem is very relevant for pathology followed by with increased and reduced cortical excitability when interpretation of results and their comparison with normative data are difficult [4, 12].

TMS can cause the effect of short-term excitation and inhibition since TMS-induced action potential in a neuron propagates along the axon and is able to activate many surrounding neurons of various modalities through the synapses [40]. Certain general principles of therapeutic stimulation have been developed for a long follow-up period (2000—2015) and published in guidelines with various evidence levels [4, 18, 42].

Methodological issues also include the choice of rTMS intensity: suprathreshold, threshold or subthreshold. The last one varies in different studies from 90—80% [33, 27] to 50% of the motor threshold. Moreover, various researchers conduct stimulation using a wide range of power parameters in accordance with modern guidelines on TMS safety [45]. Therefore, significant variability in stimulation protocols is obtained.

Two rTMS frequency modes were distinguished (“inhibitory” and “activating”) according to the results of theoretical and practical studies in the 90s—2000s. The first mode was stimulation with a frequency of less than 2 Hz causing the so-called LTD effects (long-term depression), the second mode — stimulation with a frequency over 2—5 Hz causing LTP effects (long-term potentiation). These ideas became the basis for development of various rTMS protocols implying different frequencies and stimulation powers, as well as the number of stimuli (it is assumed that greater number of stimuli cause more persistent effects) and intervals between stimulus packets (different types of theta-burst stimulation) [4, 45, 46].

However, multiple rTMS data have been accumulated which do not fit into this scheme [47, 48]. A certain positive effect of rTMS was noted by various authors regardless of stimulation mode [5, 26, 39]. A.V. Chervyakov et al. [27] compared the effects of TMS of the motor cortex with frequencies of 1 Hz and 10 Hz in patients with stroke. The authors reported positive effect of various modes on motor functions. Some foreign authors confirmed these results [49, 50]. Thus, the idea about exciting physiological effects of high-frequency rTMS and inhibitory ones of low-frequency stimulation turned out to be very simplified [15, 48].

Perhaps, functional effects of rTMS are largely determined by the basal state of the brain tissue and its excitability. The last one may be significantly altered in patients with brain diseases including severe TBI. This assumption is consistent with the ideas of classical neurophysiology and confirmed by some recent trials. It is shown that potentiation or inhibition effect of stimulation substantially depends on initial functional state of neural networks (so-called “resting-state condition”). This condition depends on many factors which seem insignificant at first glance even in healthy people: sleep deprivation on the eve of the procedure, emotional state, hormone levels, dependence on circadian rhythms, etc.

In this regard, even recognized leading experts are currently reviewing the above-mentioned initial principles of rTMS although these approaches were considered fundamental 10—15 years ago [35]. There is a tendency towards refusal strictly regulated protocols to obtain reproducible effects in favor of predominantly individualized approach. This method considers initial cortical excitability [51] and ensures selection of rTMS parameters taking into account features and functional characteristics of initial state of central nervous system [52]. For example, reduced or absent reactivity in response to TMS in those brain areas with PET-confirmed metabolic impairment is described in the literature [53].

The protocols with different duration, power and number of sessions are used for most stimulation zones due to the absence of clear recommendations. The only common aspects for most of them are adherence to guidelines on safety of rTMS of the International Federation of Clinical Neurophysiology (IFCN) [44] and selection of TMS frequency in accordance with the above-described principle (classic LTP-LTD technique). Stimulation protocols for the treatment of neuropathic pain and depression are characterized by the highest evidence level (class A) [4].

An important problem of rTMS in patients with severe TBI followed by impaired motor sphere and consciousness is the choice of the “target” for stimulation.

Thus, positive effects of stimulation of primary motor and prefrontal cortex were reported in patients with motor disorders [17]. Stimulation of sagittal parts of motor area was effective in patients with Parkinson's disease [4]. Isolated potentiation rTMS of the motor cortex [4, 27, 30] in damaged hemisphere or in combination with inhibitory stimulation of the intact hemisphere [4, 25] is used after stroke. More recent studies have shown that the most effective motor recovery after stroke is ensured by stimulation of either both hemispheres [27,47] or low-frequency rTMS of only intact hemisphere [54]. It should be emphasized once again that these protocols are actively being reviewed [49, 50]. Inconsistencies in attempts to use standardized approaches have been previously noted. For example, Pino et al [55] proposed a bimodal balance model of stimulation where the choice of mode is associated with intact structural reserve.

Stimulation of occipital area [7], prefrontal cortex [33], as well as unilateral stimulation of primary motor cortex with more significant positive effect of left hemispheric stimulation are used [4, 31] for recovery of impaired consciousness [56]. rTMS of gyrus angularis was described and more significant effect was noted in patients with minimum consciousness rather vegetative state [32]. rTMS of sagittal parts of premotor cortex is perspective for activation of voluntary attention and motor sphere in case of vegetative state and akinetic mutism [29, 57]. Some authors note moderate positive effects of stimulation. Complications of rTMS [27] and individual characteristics of damaged brain response to stimulation are less commonly analyzed [53].

Considering above-mentioned references, the choice of a “target” for rTMS is often determined by general ideas about topographic anatomy or neurophysiology of impaired function. At the same time, an important but not always taken into account aspect of stimulation (largely due to insufficient knowledge) is functional integrity and features of interaction of stimulated structures with other ones that makes impossible the use of "mechanistic" approach. We cannot argue that functional effects of stimulation of certain brain area are due solely to direct effect on the structures in projection of magnetic coil. Interaction of individual brain areas, functional communications and pathways, orthodromic and antidromic TMS-induced spread of excitation, as well as accumulation effect should be also considered. Considering hierarchical organization of nervous system, stimulation of higher systems followed by functional modulation of underlying structures and vice versa are discussed within the concepts of positive and negative feedback [4,18].

It is important to emphasize that the problems of network and hierarchical organization of central nervous system in norm and disease are among the urgent for neuroscience. In this regard, various questions concerning systemic brain responses to stimulation, rTMS protocols for various CNS diseases (including the choice of individual “targets” for stimulation), as well as evaluation of results are still of significant research interest [4, 11, 58].