Clinical application of repetitive transcranial magnetic stimulation in stroke rehabilitation.

Proper stimulation to affected cerebral hemisphere would promote the functional recovery of patients with stroke. Effects of repetitive transcranial magnetic stimulation on cortical excitability can be can be altered by the stimulation frequency, intensity and duration. There has been no consistent recognition regarding the best stimulation frequency and intensity. This study reviews the intervention effects of repetitive transcranial stimulation on motor impairment, dysphagia, visuospatial neglect and aphasia, and summarizes the stimulation frequency, intensity and area for repetitive transcranial magnetic stimulation to yield the best therapeutic effects.

Abbreviations:

rTMS, repetitive transcranial magnetic stimulation; TBS, theta burst stimulation; fMRI, functional magnetic resonance imaging

INTRODUCTION

Many interventions, such as pharmacological treatments, physical and behavioral therapies have been largely applied to improve diverse post-stroke neurologic deficits, including visuospatial neglect, aphasia, dysphagia and motor impairment, using brain plasticity[1]. Occurrence of cortical reorganization after stroke[2] has been reported in previous studies and suggested that increased cortical activity in the affected cerebral hemispheres would promote recovery from stroke[3].

Repetitive transcranial magnetic stimulation (rTMS) has gained growing importance in the field of stroke rehabilitation. Based on the ability to modulate excitability, the useful therapeutic effect of TMS has been proposed by many studies as a potential treatment for various disorders of stroke[4], such as motor performance[5678], dysphagia[9] and neglect[10]. Although the concepts of interhemispheric inhibition and change in synaptic plasticity are considered possible mechanisms, no complete interpretation of rTMS has been formulated[1112].

Effects of rTMS on cortical excitability can be altered by the stimulation frequency, intensity and duration[13]. However, a recent review concluded that the evidence remains insufficient regarding the optimum frequency or ‘dose’ of rTMS[14]. In this article, the effects of rTMS on various dysfunctions post-stroke in adult patients are reviewed.

EFFECTS OF RTMS ON MOTOR IMPAIRMENT POST-STROKE

(Table 1)

Table 1

rTMS on motor impairment

Natural recovery is no longer expected in many chronic stroke patients. Sustained improvement for up to 3 months was observed in one study[15]. rTMS exhibits much stronger and longer lasting effects in patients with acute stroke. Khedr et al[16] demonstrated that the yield of 1 Hz rTMS was 48.3% larger than sham stimulation, in terms of Barthel index in acute stroke, and the sustained effects lasted for over 1 year beyond the stimulation[6].

The results were consistent regardless of injury site, showing improvement in not only cortical, but subcortical stroke and pontine hemorrhagic patients, due to the control of corticospinal excitability, although no direct stimulation to deep structure prevailed[717]. However, some studies showed a gap between the functional results and corticospinal excitability[518], and the results of rTMS in stroke patients did not seem to be related to the regional blood flow. The mechanism underlying the motor recovery after rTMS application needs further investigation. Various frequencies between 1 to 20 Hz had been tried. Typically, low-frequency rTMS reduces excitability while high-frequency rTMS exerts facilitatory effects.

The use of 1 Hz rTMS has been proven to be effective, although there is some discrepancy among measurement tools[78151617]. Low-frequency rTMS on unaffected hemisphere produces better effects than high-frequency rTMS on affected hemisphere[1516]. Khedr et al[16] reported that 1 Hz stimulation on unaffected hemisphere exhibited greater motor function improvement than 3 Hz stimulation. This occurs because 3 Hz stimulation over the affected hemisphere only increased the cortical excitability of affected hemisphere without changes of unaffected hemisphere. Evidence exists that low-frequency rTMS changed cortical excitability in bilateral hemispheres at the same time, whereas high-frequency rTMS is successful only on the affected hemisphere[7]. Application of high-frequency rTMS alone on the affected hemisphere would not produce consistent results[1920]. Recently, high-frequency combined with low-frequency rTMS is better than low-frequency rTMS alone in recovery of motor function[19]. Theta burst stimulation (TBS) is a pattern of rTMS that can facilitate M1 excitability when delivered intermittently or suppress M1 excitability when delivered continuously. Although interhemispheric inhibition is considered as one of the mechanisms of rTMS, continuous TBS (cTBS) shows no relationship between motor function improvement and interhemispheric inhibition[2], and even causes deteriorated motor activity[121]. Therefore, TBS is considered against the simple application of the concept of interhemispheric inhibition.

Nonetheless, TBS has an advantage in safety, because it employs lower intensity compared to the usual rTMS, allowing it to be safer and more effective after securing the underlying mechanism and the precise usage[22]. Therefore, further research is needed to find out the optimal TBS protocols to induce better motor recovery. The intensity of the stimulation might play a highly critical role. Subthreshold stimulation may act by local effect of the stimulation area, but a suprathreshold stimulation may change not only the stimulation area but also the opposite homogenous motor cortex[5678]. Pomeroy et al[18] found that motor function did not change after suprathreshold (120%) stimulation. Effects of rTMS using different measurement tools were various, as shown in one study where outcomes of dexterity and grip strength were different[17].

Choice reaction time, grip strength and cognitive performance[723] are seldomly improved, which can be explained by activation of more distributed cortical areas for such parameters.

rTMS has been shown to be effective for spasticity, hyperkinetic disorder and simple motor function[2224]. rTMS has proved its effectiveness while conventional rehabilitation results in unsatisfactory outcomes, so rTMS is speculated to be a superb[25]. Task-specific rehabilitation combined with rTMS can improve motor function[2324]. The capability of rTMS to facilitate motor learning through motor cortex excitability[26] is one of the important elements of rehabilitation, and therefore can be applied as an adjuvant therapy for the treatment of motor impairment. A limitation of the review is the relatively small sample size. The outcomes are collected and analyzed within a short period of time. In addition, the protocols in each study[56781516171819202122] are not standardized. For this reason, the rTMS effects would not be better demonstrated. Further studies involving larger number of patients and using standard assessment of functional outcomes are needed.

EFFECTS OF RTMS ON DYSPHAGIA

(Table 2)

Table 2

rTMS on dysphagia

Dysphagia can be managed by modified diet, compensatory swallowing technique and training[9]. rTMS has been tried to treat dysphagia, since it can stimulate the cortical input to the swallowing center[9]. Khedr et al[9] reported that post-stoke dysphagia improved after daily treatment sessions using rTMS, with an excitatory frequency of 3 Hz on the esophageal motor area of affected hemisphere.

Based on this research, Khedr et al[27] tested the effects of 3 Hz rTMS on the esophageal motor cortex of bilateral hemispheres in patients with vertebrobasilar stroke, which induces severe dysphagia symptoms. All patients recovered swallowing function immediately after five sessions of rTMS applied on bilateral hemispheres with the same frequency. Although swallowing is controlled by bilateral hemispheres, interhemispheric asymmetry exists[2930], allowing swallowing function to be controlled by either direct inhibition of unaffected hemisphere or facilitation of affected hemisphere. Based on the interhemipheric rivalry concept, the unaffected hemisphere was stimulated with a frequency of 1 Hz[28], whereas the esophageal motor cortex of affected hemisphere was stimulated with a frequency of 3 Hz[9]. According to the studies, various recovery mechanisms have been suggested, including increased excitability of the corticobulbar projections from bilateral hemispheres[9] or activation of remaining unaffected premotor cortex and contralateral medulla[27]. While motor area of cerebral cortex controls swallowing initiation, brainstem has a role in the swallowing reflex. Since rTMS exerts corticobulbar tract through the cerebral cortex, it activates both oral and pharyngeal stages of the coordinated swallowing process at the same time[28]. Thus, rTMS might be a better adjuvant therapy, compared to the swallowing functional electrical stimulation.

EFFECTS OF RTMS ON VISUOSPATIAL NEGLECT

(Table 3)

Table 3

rTMS on visuospatial neglect

Visuospatial neglect is a common, yet frequently overlooked, neurological disorder following stroke characterized by a deficit in attention and appreciation of stimuli on the contralesional side of the body[31]. It is common, with an incidence of 24.7% in acute stroke patients[32]. In addition to a number of treatments attempted over the last few years, including scanning, limb activation, eye patching, neck vibration and prism[33], rTMS also has been used as an adjuvant therapy for neglect.

Initial studies demonstrated that rTMS of the unaffected hemisphere during the execution of a line bisection task transiently decreased the magnitude of neglect, followed by sustained effects after stimulation[34353637383940]. 25 Hz rTMS on unaffected parietal cortex was performed by means of an online approach by Oliveri et al[34]. Low-frequency rTMS on the unaffected hemisphere was used in five[3536373839] out of seven[363738394041] studies, and the effect was explained by the control of interhemispheric rivalry. Koch et al[35] applied 1 Hz rTMS (600 pulses, 90% resting motor threshold) over the unaffected posterior parietal cortex, to evaluate the sustained effect after the stimulation period, and showed reduction of the pathological hyperactivity of the intact hemisphere and improved performance immediately after rTMS.

Song et al[36] applied low-frequency rTMS on unaffected posterior parietal cortex twice a day for 2 weeks, (0.5 Hz, 90% motor threshold) and reported significant improvement for 2 weeks. In contrast, Oliveri et al[34] revealed the beneficial effects of high-frequency rTMS on unaffected hemisphere, contradicting the findings from other studies. The same parameters should be applied with caution, since it was the only study to use high-frequency rTMS for neglect. Nyffeler et al[40] applied cTBS over unaffected hemisphere with different numbers of trains and reported that repeated applications of TBS over the contralesional posterior parietal cortex on the same day specifically and significantly improved the perception of visual targets presented on the left side up to 32 hours. Most of the studies demonstrated positive results for visuospatial neglect[33343536383940]. Song et al[36] demonstrated that 20 sessions of low-frequency rTMS over the unaffected hemisphere improved line cancellation and bisection. However, dissociations were found among different types of measurement tools, because of varying sensitivities for the diagnosis of neglect[41]. It was well known that there were dissociations between two cardinal diagnostic tests, i.e. cancellation and line bisection[42].

Furthermore, the heterogeneous mechanisms of visuospatial neglect are thought to be another cause for such differences in results. Attention was one of the mechanisms that build up neglect, and rTMS improved attention to greater extent than visuospatial function[3738].

Long-term effect after rTMS stimulation was investigated[363739]. Peak behavioral inattention test score of the patients at 6 weeks after rTMS stimulation remained better than that of pre-rTMS stimulation[36]. rTMS exhibits longer after-effect on visuospatial neglect than other treatment modalities. A direct comparison would be necessary, since no such comparison has been attempted. Koch et al[35] found that 1 Hz rTMS over left primary parietal cortex inhibited the over-excitability of left posterior parietal cortex-primary motor cortex circuits and also impacted visual neglect. However, the improvement did not correlate with the size of the normalization of the over-excitability. Further studies are necessary to elucidate the effects and mechanisms, and to establish the optimal protocols of rTMS for visuospatial neglect improvement. The effects of affected hemisphere stimulation should be identified, and it is preferable to compare the effects of rTBS with other types of TMS.

EFFECTS OF RTMS ON APHASIA

(Table 4)

Table 4

rTMS on aphasia

The effectiveness of widely used speech-language therapy is generally known to diminish, as gradually moving on from acute to chronic stage in stroke patients[50]. Studies showed that recovery could take place for extended period of time after stroke in patients who received conventional aphasia rehabilitation[515253]. rTMS can be considered a novel therapy for aphasia because it can promote recovery of chronic aphasia. rTMS used for up to 11 years showed effects in chronic aphasic patients[434445], even in those who showed stable deficits of elicited propositional speech[46]. rTMS has effects on a variety of language problems, ranging from naming difficulty to speech arrest, depending on the stimulation parameter and area of rTMS.

rTMS applied to an anterior portion of right Broca's homologue has shown to affect language behaviors, including naming, in stroke patients with chronic, nonfluent aphasia[434445], with a frequency of 1 Hz. The mechanism of this protocol is to reduce interhemispheric inhibition towards the left hemisphere, which contains the main language area. On the contrary, Kakuda et al[47] applied low-frequency rTMS at Wernicke's area for sensory aphasia patients. Winhuisen et al[48] applied 4 Hz rTMS over activated region on positron emission tomography during semantic matching task and showed improvement of verb generation. Bilateral hemispheres were thought to have roles in supporting language recovery as proved by functional magnetic resonance imaging (fMRI) study[49]. Therefore, rTMS using the simple concept of interhemispheric inhibition might deteriorate recovery of aphasia. Kakuda et al[47] revealed that 1 Hz rTMS applied to the area homologous to the most activated site on fMRI seemed to be a feasible approach for post-stroke aphasic patients. The frameless stereotaxic system was used to guide the specific area on the scalp during rTMS application for aphasia.

Bashir et al[54] proved the superiority of navigated rTMS in terms of both physiologic and behavioral effects[54] by maximizing accurate and consistent targeting.

Therapeutic applications of rTMS are expected to benefit greatly with navigating electric field. As the recovery of language function was mediated by different parts of brain at different stages in terms of time[55], new therapeutic strategies, combining with fMRI or neuronavigation system, should be established for enhanced aphasia treatment in the future.

CONCLUSION

Reviewing the studies on effects of rTMS in post-stroke patients, the role of rTMS as an adjuvant therapy can be reaffirmed despite some conflicting outcomes. Nonetheless, it is imperative to further establish rTMS protocols including frequency, intensity and location to maximize the benefits of rTMS.

All the studies with the use of 1 Hz rTMS were proven to be effective and low-frequency rTMS on unaffected hemisphere was better than high-frequency rTMS on affected hemisphere[1516]. In addition, high-frequency combined with low-frequency rTMS was better than low-frequency rTMS alone, in terms of motor training[19]. Suprathreshold stimulation may change not only the stimulation site but also the opposite homogenous motor cortex[567818] and showed less effectiveness compared with the subthreshold stimulation.

Swallowing function can be controlled by either direct inhibition of unaffected hemisphere or facilitation of affected hemisphere[2829]. A previous study using TMS demonstrated that post-stroke dysphagia recovery was asymmetric between two hemispheres and was accompanied with activation of the unaffected hemisphere[56]. Application of 5 Hz rTMS to the unaffected pharyngeal motor cortex increased pharyngeal cortical excitability and improved swallowing behavior[57]. Similar findings have been identified by other studies using brain imaging techniques[5859]. Such asymmetries have been explained by a lack of transcallosal inhibition between hemispheres in swallowing function[6061]. Based on the interhemipheric rivalry concept, the unaffected hemisphere was stimulated with low-frequency rTMS[30], whereas the affected hemisphere was stimulated with high-frequency rTMS[9]. Both hemispheres have roles and recently rTMS is applied over the most activated sites on functional image during task[52]. In the future, optimal stimulation of rTMS using fMRI or neuronavigation system should be established for clinical application for the stroke patients.