The Silent Period Induced by Transcranial Cranial Nerves

Home , Genioglossus, Mentalis

5865tournal ofNeurology, Neurosurgery, and Psychiatry 1995;59:586-596 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from The silent period induced by transcranial magnetic stimulation in muscles supplied by cranial nerves: normal data and changes in patients

K J Werhahn, J Classen, R Benecke

Abstract magnetic stimulation in limb muscles a period The silent period induced by transcranial of electromyographic (EMG) silence (silent magnetic stimulation of the sensorimotor period) lasting some 200 ms in small hand cortex (Magstim 200, figure of eight coil, muscles can be found if the target muscle is loop diameter 7 cm) in active muscles tonically active.'--8 The duration of the silent supplied by cranial nerves (mentalis, period follows a proximodistal gradient in the sternocleidomastoid, and genioglossus) upper limb muscles, being longer in small was studied in 14 control subjects and hand muscles than in proximal arm muscles. nine patients with localised lesions of the As to precisely what structures are involved sensorimotor cortex. In the patients, in the generation of the silent period there measurements of the silent period were is an increasing amount of data suggesting that also made in the first dorsal interosseus it is not due to changes in a motoneuron and tibialis anterior muscles. In the con- activity but to changes in pyramidal cell trols, there was a silent period in con- excitability.' 68-11 tralateral as weil as ipsilateral cranial Apart from the study of distal limb muscles, muscles and the duration of the silent several reports exist on activation of cortico- period increased with increasing stimulus nuclear fibres supplying cranial nerve intensities. The mean duration of the motoneurons using transcranial magnetic silent period was around 140 ms in con- stimulation.'2 13 With bilateral recordings two tralateral mentalis muscle and around types of responses can be seen in muscles sup- 90 ms in contralateral sternocleidomas- plied by cranial nerves: purely ipsilateral short toid muscle at 1-2 x threshold stimula- latency responses and bilateral responses of tion strengths. Whereas the duration of longer latency. On the basis of latency, ampli- the silent period in ipsilateral mentalis tude, configuration, and reaction to facilitatory muscle was shorter than on the contralat- manoeuvres, these two types were considered eral side it was similar on both sides in to originate from unilateral direct excitation of sternocleidomastoid muscle. In patients cranial nerves at proximal sites and the corti- with focal lesions of the face associated conuclear system respectively.'2 13 The bilateral primary motor cortex and corresponding (corticonuclear) long latency responses reflect http://jnnp.bmj.com/ central facial paresis, the silent period in the bilateral cortical input to the brainstem mentalis muscle was shortened whereas it nuclei of cranial nerves. was unchanged or prolonged in limb mus- The first aim of the present study was to cles (first dorsal interosseus, tibialis ante- describe the characteristics of the silent period rior) with stimulation over the affected induced by transcranial magnetic stimulation hemisphere. By contrast, in a patient with in facial, neck, and tongue muscles in normal a lesion within the parietal cortex, the subjects. If it is assumed that the silent period on September 23, 2021 by guest. Protected copyright. silent period in mentalis muscle was pro- in these muscles is also generated cortically, a longed with stimulation of the affected silent period should also exist in cranial mus- side. cles ipsilateral to the site of stimulation, given the bilateral corticonuclear projection. (3 Neurol Neurosurg Psychiatry 1995;59:586-596) Secondly, we wanted to examine changes in the duration of the silent period in muscles supplied by cranial nerves in patients with uni- Department of Keywords: transcranial magnetic stimulation, cranial lateral cortical lesions. Alterations of the silent Neurology, Heinrich- nerves, sensorimotor cortex, silent period period in distal limb muscles of Heine Universitat patients with Dusseldorf, Germany cortical and subcortical or thalamic lesions K J Werhahn Transcranial magnetic stimulation has been have been described.45 " In patients with J Classe widely used to study the function of the cor- lesions of the primary motor cortex itself the R Benecke ticospinal pathways in humans.' The silent period was shortened'4 or abolished'0 in Correspondence to: Professor Dr R Benecke, responses it evokes are thought to be gener- muscles contralateral to the hemispheric Department of Neurology, ated by excitation of corticomotoneuronal lesions. In patients with lesions in other motor Universitat Rostock, Gehlsheimer Strasse 20, pathways at the cortical or just subcortical competent cortical areas which are afferent to 18055 Rostock, Germany. level in the motor cortex and to be mediated the primary motor cortex a prolongation of the Received 7 October 1994 by fast conducting corticospinal pathways. 2 silent period occurred.14 and in final revised form 16 June 1995 After the early compound muscle action Some of the data from the present investiga- Accepted 18 August 1995 potential (CMAP) evoked by transcranial tion have appeared in abstract form .15 The silent period induced by transcranial magnetic stimulation in muscles supplied by cranial nerves: normal data and changes in patients 587 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from Methods was established by MRI and CT performed Experiments were performed in 14 control during the stay in hospital. subjects and nine patients. All gave their informed consent for the study and the project STIMULATION PROCEDURE AND RECORDINGS IN was approved by the local ethics committee. MUSCLES SUPPLIED BY CRANIAL NERVES IN CONTROLS AND PATIENTS CONTROLS Subjects and patients were seated in a comfort- Fourteen control subjects (11 men, three able chair. Transcranial magnetic stimulation women, mean age 34-2 (range 19-52) years) was performed with a conventional Nova- were chosen from among the authors and metrix 200 stimulator (Fa Madaus, Magstim other members of the department. Subjects Company UK) with a maximum output of 2-0 with a history of hypertension, headache, a Tesla, connected to a figure of eight magnetic cerebrovascular event, or subjects over the age coil, each loop having a diameter of 7 cm. of 60 were excluded. This last criterion was Technical information about the stimulator used to reduce the possibility of subclinical can be found elsewhere.'1 16 cerebrovascular abnormalities. The coil was held over the skull with the maximum current in the coil flowing in an PATIENTS anteroposterior direction. The centre of the Table 1 lists the clinical details of the patients. coil was placed consecutively over both hemi- Patients (five men, four women, mean age spheres about 9 cm lateral and 3 cm anterior 59-1 (range 41-75) years were selected among to a line connecting the vertex and the external consecutive patients seen as inpatients in our auditory meatus. The precise coil position was department, who had a unilateral hemispheric then adjusted to give a response of maximum lesion due to vascular accident or tumour. size in the contralateral cranial muscle at a Patients unable to cooperate or confined to given stimulus intensity. The optimal position bed were excluded. Clinical data and results of of the coil was marked on the skull for better investigations were obtained from the patients' retrieval. Before each recording session the records. Two patients (3 and 5) reported pre- threshold for evoking cortical muscle vious cerebrovascular events. Patient 3 had responses was determined for each muscle had a left hemispheric infarct with global dys- with the subject or patient at rest and the coil phasia and right facial weakness one year held over the optimal spot. Threshold was before the study and patient 5 had had a left defined as the lowest stimulus intensity (as a hemispheric infarct with right sided hemipare- percentage of the maximal output of the sis, including the face, with complete remis- device) with which a defined cortically evoked sion after a few weeks, four years previously. response of at least 100 ,uV peak to peak The localisation of the lesions in the patients amplitude could be obtained on three consecu-

Table 1 Clinical datafor patients CMCTIAmplitude

Patient Age (y) Sex Localisation Type Interval Clinicalfindings Hand Leg http://jnnp.bmj.com/ 1 49 F Right parietal TU* 33 months CN normal, mild Normal Normal sensory deficits left arm, no dysphasia 2 61 M PMC Infarct 4 months Initially mild lower T CMCT Normal facial weakness, mono- left paresis left lower arm, no dysphasia 3 75 F Left fronto- Infarct 5 days CN normal, in 1992 4 Amp Normal parietal dysphasia and right facial right

weakness, sensorimotor on September 23, 2021 by guest. Protected copyright. dysphasia, paresis right arm 4 65 F Right fronto- Infarct 4 years Minor hemiparesis left Normal Normal parietal arm > leg, predominant left facial weakness, no dysphasia, left hemi- neglect and sensory deficits 5 64 M Left postcentral Infarct 1 year CN normal, no dysphasia, 4 Amp Normal gyrus spastic monoparesis right right arm, transient right hemiparesis (including face) in 1989 6 54 M Left PMC and Infarct 2 8 months Right facial weakness and No response 4 Amp striatocapsular minor motor aphasia, right t CMCT spastic hemiparesis right Right 7 73 F Left fronto- Infarct 17 days Discrete right hemiparesis 4 Amp Normal parietal (arm > leg) sensory dys- right phasia, mild facial weakness 8 41 M Right fronto- Infarct 17 days Left hemiparesis t CMCT t CMCT parietal (face > arm > leg) left Left global dysphasia 9 55 M PMC Infarct 3 days Right hemiparesis Normal Normal (arm + face) resolving after 4 days, motor dysphasia Amplitude and CMCT = Amplitude and central motor conduction time for the first dorsal interosseus muscle (in patient 2 the finger extensors were abnormal while variables in the first dorsal in interosseus muscle were normal) and tibialis anterior muscle; t = prolongation; 4 = reduction, CN = cranial nerves; Interval = the time between the occurrence of the lesion and the time of study; PMC = primary motor cortex. *Haemangioma as proved by histology, operated on in June 1990. 588 Werhahn, Classen, Benecke J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from Figure 1 Raw data from A two normal subjects Mentalis SCM Genioglossus showing examples ofthe silent period in muscles supplied by cranial nerves. In each tracefive sweeps are superimposed. (A) Recordings are shown in the muscles contralateral and ipsilateral to the side II ofstimulation. Note that in mentalis and Lhimk ilI A. I JJ1 I genioglossus muscles during .ML. -II-LM Tl 1 " the silent period, EMG TIRr, background activity is not decreased to zero. (B) the r~~~~P17mi'rwl relation between the stimulus intensity (expressed relative to the 0.5 mV 0.5 mV threshold (T) for 0.5 mV contralateral responses at rest), and the duration of 100 ms 100 ms 100 ms the silent period in contralateral muscles is shown. The silent period was longer with higher Ipsilateral u stimulus intensities only in to the side mentalis and genioglossus of stimulation -_ S muscle; it was unchanged P11 Wi,m r ,rwnr- in sternocleidomastoid T r]' (SCM).

llitzikl1.1ii ii Iill .1I 1-2x U n- 1lLIm*

0-5 mV http://jnnp.bmj.com/

100 ms 100 ms

1-5x - r r yr, -- on September 23, 2021 by guest. Protected copyright.

tive trials in the contralateral target muscle. patients, except 7 and 8, there was no inter- Usually a stimulation strength of 1-5 x hemispheric difference in the threshold of threshold was used during the recordings. In stimulation in the muscles under study. In some experiments (subjects with a high corti- patient 7 the threshold for cortical responses cal threshold at rest or in whom the effect of was 5% higher with stimulation over the changing the stimulus intensity was under affected hemisphere in both mentalis and ster- investigation) recordings were also made using nocleidomastoid muscle. By contrast, in 1-2 x threshold or 2-0 x threshold stimulus patient 8 the threshold stimulating over the intensities. In the controls the threshold of affected hemisphere was 10% below the stimulation between the two hemispheres dif- threshold of the unaffected hemisphere. fered by 5% of the maximum output of the In most cases recordings were made with device in one subject for mentalis muscle, in concentric needle electrodes (Medelec, dis- two subjects for sternocleidomastoid muscle, posable type DFC 25) inserted into the target and in one subject for genioglossus muscle. In muscles. In some subjects who did not tolerate The silent period induced by transcranial magnetic stimulation in muscles supplied by cranial nerves: normal data and changes in patients 589 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from needle recordings surface electrodes (Dantec clearly define the offset, EMG activity of the 1 3L20) with the active electrode over the mus- single traces was rectified and measurements cle belly and the reference electrode 2-3 cm of the amplitude were made before and after apart were used. In controls mentalis muscle the stimulus. These measurements were used was recorded from bilaterally in 11 subjects to determine the onset of EMG background and stemocleidomastoid muscle in nine sub- activity after the silent period . For statistical jects. Recordings from the tongue musculature analysis unpaired Student's t tests were used. were made in two subjects. Responses from mentalis muscle were STIMULATION PROCEDURE AND RECORDINGS IN recorded with needle electrodes in seven LIMB MUSCLES IN PATIENTS patients and with surface electrodes in two In the patients recordings were also made (patients 7 and 9 in tables 2 and 3). from the first dorsal interosseus and tibialis Recordings were made bilaterally except in anterior muscles. Surface electrodes (Dantec patient 9. In addition, in patients 6, 7, and 9 13L20) with the active electrode placed over bilateral recordings were also made from stern- the muscle belly and the reference electrode ocleidomastoid muscle. placed either over the index finger or the After definition of the optimal site of stimu- medial aspect of the tibial bone were used. lation and threshold for a given muscle, bilat- Stimulation was performed with a Novametrix eral recordings were made in tonically active Magstim 200 conventional stimulator con- muscles. The level of preactivation (20% of nected to a round coil (12 cm outer diameter) maximum) was monitored by audiovisual con- for stimulation of the first dorsal interosseus trol by the subjects or patients. In both sub- and a large angled figure of eight coil (outer jects and patients at least 20 stimuli were loop diameter 13 cm) for stimulation of leg applied to each hemisphere for every muscle muscles. The coils were held with the centre of under study. The EMG signals were amplified each coil over Cz. Threshold at rest was deter- using a Toennies Myograph IIR with band- mined as the stimulation strength which pass filtering set between 20 and 3000 Hz. induced a CMAP in the contralateral muscle The data were digitalised by a CED 1401+ in at least half of five consecutive trials when (Cambridge Electronic Design, UK) interface, the muscle under study was at rest. During sampled (CED, Signal Averager), and stored recordings a stimulus intensity of 1-5 x with a sampling rate of 5000 Hz using a per- threshold was used with the muscle tonically sonal computer for later off line analysis. active and under audiovisual control of the For analysis and measurements of the silent level of preactivation. Sampling and analysis period duration the mean (SD) of 20 sweeps procedures were otherwise identical to those of 600 ms total duration was measured off used for muscles innervated by cranial nerves. line. The stimulus in each sweep was given Normal data regarding the silent period dura- 100 ms after the start of the sweep to record tion in limb muscles have been reported in a the level of EMG background activity. Trials previous study.8 with insufficient background activity were rejected from analysis. Because the silent period in cranial muscles most often did not Results terminate abruptly the duration of the silent CONTROLS period in these muscles was defined as the In the controls, transcranial magnetic stimula- http://jnnp.bmj.com/ time between the onset of the magnetic stimu- tion of one hemisphere evoked a silent period lus and the time when EMG background in ipsilateral and contralateral cranial muscles. activity reached at least 50% of the prestimu- Table 2 shows the mean durations of the silent lus level. In most cases the offset of the silent period in mentalis, sternocleidomastoid, and period could be determined by visual inspec- genioglossus muscles at various stimulus tion (fig 1). In cases in which it was difficult to intensities. Figure IA shows a representative

example of the silent period in these muscles. on September 23, 2021 by guest. Protected copyright. The silent period in contralateral mentalis muscle in most cases consisted of an initial Table 2 Duration and interhemispheric difference in ms (mean (SD)) of the silent period in normal subjects period lasting around 100 ms during which the EMG background activity was suppressed No of Muscle subjects T (stim int) Contralateral Ipsilateral and a second phase in which some tonic activity returned but was still clearly below the Stimulation of right hemisphere Mentalis 9 1-2 x (68) 140-0 (30 2) 128-2 (26 4) prestimulus level (fig 1A). 9 1.5 x (85) 184-8 (17 2) 150-5 (324) As illustrated in fig 1A and shown in table 2 Stemocleidomastoid 9 1-2 x (93) 92-5 (18 7) 96-9 (14-2) in Genioglossus 2 1-2 x (45) 61 9 (4-0) 59-4 (5 0) the mean duration of the silent period ipsi- 2 1-5 x (57) 1224(10-0) 117-7(14-0) lateral mentalis muscle was somewhat shorter Stimulation ofleft hemisphere Mentalis 9 1.2 x (67) 133-7 (31-6) 118-9 (34 8) than in contralateral muscles (stimulation of 9 1-5 x (85) 1822 (200) 1514 (446) right hemisphere: duration = 34-3 ms, P = Sternocleidomastoid 8 1-2 x (93) 87-0 (23-1) 97-9 (18-5) 0 Genioglossus 2 1-2 x (42) 68-7 (2 8) 57-7 (11-6) 0 13; stimulation of left hemisphere: duration 2 15 x (53) 140-2 (8Q5) 135 3 (1 1) = 30-8 ms, P = 0 077, t test). This was not Interhemisphenic difference Mentalis 9 1 2 x 12 9 (10-4) 20-8 (12 2) true for sternocleidomastoid or genioglossus 9 1-5 x 17 2 (7-7) 19-4 (14-1) muscles, in which the duration on average was Stemocleidomastoid 8 1-2 x 9-8 (8-7) 5-4 (3-5) Genioglossus 2 1-2 x 6-9 (1-6) 6-6 (1-0) the same in ipsilateral and contralateral mus- 2 1-5 x 18-5 (25-1) 176 (15-1) cles (see also fig 1A). x duration of the T = Stimulus intensity in x threshold at rest; stim int = average stimulus intensities (in paren- At 1 2 threshold the theses) as percentage of the maximum output of the stimulating device. silent period was significantly longer 590 Werhahn, Classen, Benecke J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from Table 3 Duration in ms (mean (SD)) of the silent period in patients with unilateral cerebral lesions Mentalis SCM FDI TA Affected Unaffected Patient (ipsilat) (ipsilat) Affected Unaffected Affected Unaffected Affected Unaffected 1 179 (18)*** 104 (28) - - 222 (23)*** 183 (13) 258 (16)*** 193 (50) 102 (32) 89 (27) 2 109 (11)*** 211 ± 14 - - 212 (14)*** 170(6) - - 73 (19) Not measurable 3 95 (3)*** 171 (12) - - 241 (33)*** 200 (15) 174 (38)*** 126 (11) 108 (25) 162 (29) 4 63 (15)*** 153 ± 9 - - 602 (82)*** 229 (19) 348 (29)*** 284 (32) Not measurable 149 ± 8 5 124 (10)*** 176 ± 6t - - 194 (27) 169 (7) 230 (34)*** 163 (10) 102 (22) 163 ± 13 6 96 ± 8*** 203 (7) 106 (10) 104 (7) No response 177 (7) No response 108 (16) 100 (6) 201 (7) 7 117 (17)*** 197 (7) 53 (3) 55 (3)t 159 (15)*** 193 (6) 244 (6) 250 (8) 113 (13) 184 (19) 8 153 (9)*** 332 (14) - - 355 (17)*** 270 (37) 389 (40) 387 (45) 144 (25) 332 (15) 9 58 (5)*** 163 (15) 97 (11) 96 (7) 204 (12)** 186 (10) 135 (17) 125 (17)

** p < 0-01; *** p < 0-001 (t test for interhemispheric difference). t 1-2 x threshold stimulus intensity. Silent period duration in muscles ipsilateral to the side of stimulation are given in bold. SCM = stemoclerdomastoid; FDI = first dorsal interosseus; TA = tibialis anteroid; - = not done.

(P < 0 01) in mentalis muscle compared with sities (1-2 v 1U5 threshold; P < 0 05). At 1-5 stemocleidomastoid and genioglossus muscle threshold differences in duration between both for ipsilateral and contralateral responses mentalis and genioglossus muscle were not (table 2). significant. Stemocleidomastoid muscle could The duration of the silent period in not generally be tested at 1-5 threshold due to genioglossus and mentalis muscle ipsilateral high absolute threshold and limitations of the and contralateral to the side of stimulation stimulus device. However, in one subject depended on the stimulus intensity (fig 1B, activity in stemocleidomastoid muscle showed table 2), being longer at higher stimulus inten- only a slight increase in duration of the silent

Figure 2 Raw data from mentalis, A sternocleidomastoid (SCM), andfirst dorsal Mentalis interosseus (FDI) muscles contralateral to the side of stimulation in two patients. In each graph five sweeps are superimposed. In Affected patient 6(A) the silent (right) period in mentalis muscle

(left two panels) is http://jnnp.bmj.com/ significantly shorter and less wellformed and the amplitude ofthe early direct response is reduced Unaffected 250 V with stimulation of the (left) affected compared with the I unaffected hemisphere. By contrast, there is no 100 ms difference in the silent period duration in on September 23, 2021 by guest. Protected copyright. sternocleidomastoid (right two panels). Comparing the duration ofthe silent B period in mentalis muscle, patient 1(B) shows that it is significantly longer after stimulation ofthe affected compared with the Affected EL. ].U unaffected side. By (left) contrast, the duration of the silent period in the first dorsal interosseus muscle (right two panels in (B)) is prolonged with stimulation of the affected compared with the unaffected side. Unaffected (right) 1 mV V2 Wr?j12500mV 100 ms 100 ms The silent period induced by transcranial magnetic stimulation in muscles supplied by cranial nerves: normal data and changes in patients 591 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from Figure 3 Raw data from 11 Over right hemisphere mentalis musclefrom 4 Over left (affected) patient 7. Stimulation over hemisphere the affected (left) hemisphere produced a shorter silent period both in the contralateral and ipsilateral mentalis muscles compared with the silent period after stimulation of the unaffected hemisphere. Note that the direct Left mentalis A b- I- --AL.-- responses in muscles - --.T- ipsilateral to the side of stimulation are partly Il masked by the stimulus artifact, which was bigger on the side ipsilateral to the magnetic coil.

II, I Right mentalis I I& 6

I 0-2 mV

100 ms

period at 1-5 x threshold stimulus intensity Shortening of the duration of the silent period in (Fig 1B). facial muscles The duration of the silent period was vari- In five patients there was evidence of a facial able between subjects (see SDs in table 2). weakness clinically at the time of the

study http://jnnp.bmj.com/ This was particularly true at low stimulus (patients 4 and 6-9) whereas three patients intensities (1 2 x threshold). Comparing the had had only a history of a central facial weak- duration of the silent period between the two ness (2, 3, and 5) and seemed clinically unaf- hemispheres, there was no statistically signifi- fected at the time of the examination. In one cant difference when all subjects were taken patient (2) the facial involvement had been together. The mean interhemispheric differ- seen at the time when the ischaemic lesion ence in the duration of the silent period in the occurred, four months before the study. In the muscles contralateral to the side of stimulation other two patients (3 and 5) facial weakness on September 23, 2021 by guest. Protected copyright. in all subjects was 12-9 (10-4) ms (mean(SD)) had been noted as part of a previous transient at 1-2 x threshold and 17-7 (7 7) ms at 1-5 x ischaemic episode (three years and one year threshold for mentalis muscle and 9 8 (8 1) ms respectively) and had resolved completely. for stemocleidomastoid muscle (table 2). In all patients with either facial weakness at the time of the study or a history of facial PATIENTS involvement the duration of the silent period Table 3 summarises the results for the patients in mentalis muscle was clearly shorter with and figs 2 and 3 give examples of the raw data. stimulation of the affected than the unaffected According to this definition, there was a signif- hemisphere. There was no difference in the icant difference in the duration of the silent amount of shortening of the silent period period in mentalis muscle in all patients. between the group with clinically evident facial In patient 1 the silent period was prolonged weakness and the group with only a history of in mentalis muscle with stimulation over the facial involvement. An accompanying sensory affected compared with the unaffected hemi- or motor dysphasia was present in four but not sphere whereas it was shortened in the other present in three of the patients, and its pres- eight patients. The absolute latency of the ence did not influence the amount of shorten- contralateral corticonuclear response in men- ing of the silent period. Figure 2 shows raw talis muscle was normal in all patients, data from patient 6, in (A) illustrating the dif- although needle recordings did not always ference in the duration of the silent period in allow a precise estimation of latency. contralateral muscle after stimulation of the 592 Werhahn, Classen, Benecke J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from Figure 4 Cranial MRI and mean duration of the silent period (SP) in mentalis, first dorsal interosseus (FDI), and tibialis anterior (TA) muscles in patient 6. T2 weighted axial sections of the brain are shown at two different levels illustrating that the medial aspects of the primary motor cortex are preserved while the main pathogically affected area extends laterally involving theface associated primary motor cortex. Bars indicate the SD ofsilent period duration. ***p < 0 001. http://jnnp.bmj.com/

affected and unaffected hemispheres. By con- in table 3. Figure 3, illustrating patient 7, on September 23, 2021 by guest. Protected copyright. trast, there was no difference in the duration of shows that the duration of the silent period the silent period in stemocleidomastoid mus- was abnormally short ipsilaterally and con- cle (fig 2A right panel). tralaterally with stimulation of the affected On the basis of MRI or CT there were lesions hemisphere whereas it was of normal length on including the face associated primary motor both sides when the stimuli were given over cortex in all eight patients with shortening of the the normal hemisphere. silent period in mentalis muscle (fig 4). EFFECTS IN STERNOCLEIDOMASTOID MUSCLE COMPARISON OF CONTRALATERAL AND In three patients (6, 7, and 9), in whom IPSILATERAL RESPONSES IN MENTALIS MUSCLE recordings were taken from the sternocleido- In the patients with shortening of the silent mastoid muscle, there was no interhemi- period in mentalis muscle (except 9 in whom spheric difference in the duration of the silent only contralateral muscles were recorded period after stimulation of the two hemi- from) silent period duration in both ipsilateral spheres. An example illustrating raw data from and contralateral muscles was of similar length patient 6 is given in fig 2A and the results are on both sides-that is, they were long bilater- shown in table 3. In patient 7 of table 3 the ally with stimulation of the unaffected and stimulus intensity was only 1 2 x threshold, short with stimulation of the affected side. because of the high threshold for responses in This is illustrated in fig 3 and the absolute val- stemocleidomastoid muscle in this case. ues of silent period duration in ms are shown Correspondingly, duration of the silent period The silent period induced by transcranial magnetic stimulation in muscles supplied by cranial nerves: normal data and changes in patients 593 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from Figure S Cranial MRI and mean duration of the silent period (SP) in mentalis, first dorsal interosseus (FDI), and tibialis anterior (TA) muscles in patient 1. Tl weighted axial and coronal sections are shown, illustrating the extent ofthe right parietal lesion. Bars indicate SD ofsilent period duration. ***p < 0-001.

Patient No 1

300 r _ Affected - (left) - -... Unaffected (right)

-i E 200 -.v.+ I EL 4 LO i 4-a 0 I 0: co :3

100 http://jnnp.bmj.com/

0 on September 23, 2021 by guest. Protected copyright. Mentalis FDI TA

in this patient was therefore shorter than in the from mentalis muscle, stimulus intensity was other two. In all patients in whom sternoclei- 1-2 x threshold. domastoid muscle was recorded from, there was a significant shortening of the silent period RECORDINGS FROM FIRST DORSAL INTEROSSEUS in mentalis muscle after stimulation of the AND TIBIALIS ANTERIOR MUSCLES affected hemisphere (table 3). In six patients (1-4, 8, 9) duration of the silent period in the first dorsal interosseus muscle PROLONGATION OF SILENT PERIOD DURATION was significantly prolonged with stimulation of IN FACIAL MUSCLES the affected compared with the unaffected Patient 1 had a lesion in the right parietal cor- hemisphere. One patient (6) did not show tex well outside the primary motor cortex (fig either a primary muscle response or a silent 5). There was no facial palsy. This patient period. In the remaining patients there was a showed a significant prolongation of the silent shortening of silent period duration in the first period with stimulation over the affected dorsal interosseus muscle in one (7) and no hemisphere in all muscles recorded (table 3). significant difference in duration of the silent This phenomenon was especially pronounced period between the hemispheres in the other in mentalis muscle (fig 2B, left panel). It (5). Table 3 summarises the data for silent should be noted that in this patient, recording period measurements in limb muscles. 594 Werhahn, Classen, Benecke J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from In tibialis anterior muscle there was an inhibition induced by peripheral nerve stimu- interhemispheric difference of duration of the lation, and the silent period after transcranial silent period in four patients (1 and 3-5) with magnetic stimulation have been performed.817 prolongation after stimulation of the affected As far as the silent period in cranial muscles is compared with the unaffected side (mean 61 concerned, similar experiments can hardly be (9), range 48-67 ms, P < 0-001). A concomi- done except in the orbicularis oculi muscle tant increased duration of the silent period in recording the blink reflex. Leis et al'8 found a the first dorsal interosseus muscle was present normal excitability of motoneurons innervat- in three of these patients (1, 3, and 4). Lack of ing the orbicularis oculi muscle during the interhemispheric difference of silent period silent period induced by transcranial magnetic duration in tibialis anterior muscle was noted stimulation as tested by modulation of the Rl in three patients (7-9). component of the blink reflex. Peripheral stim- Measurements of the central motor conduc- ulation of the facial nerve is followed by a tion time (CMCT) and of the peak to peak silent period which is much shorter than that amplitude of magnetically evoked potentials after transcranial magnetic stimulation.69 This (MEP) in limb muscles (table 1) were normal finding, however, does not really exclude a in three patients (1, 4, and 9). The other six medullary origin of the silent period after tran- patients showed either a decreased amplitude scranial magnetic stimulation as it cannot be (compared with peripheral electrical stimula- ruled out that corticonuclear fibres excite tion) of the cortically evoked responses or a medullary inhibitory interneurons which are pathologically prolonged CMCT (for details not recruited by peripheral facial nerve stimu- see table 1). lation. In the patients, the duration of the silent period was pathologically shortened in eight Discussion out of nine patients (2-9) with stimulation of The present study shows measurements of the the affected compared with the unaffected silent period evoked by transcranial magnetic hemisphere both in the contralateral clinically stimulation in muscles supplied by cranial affected and in the ipsilateral unaffected men- nerves in normal control subjects and nine talis muscle (fig 3). In all patients with silent patients with hemispheric brain lesions includ- period shortening in mentalis muscle there ing the face associated areas of primary motor was a central facial weakness clinically either cortex. The results indicate that not only the at the time of the study or on the basis of the long latency corticonuclear responses are pre- patient's history. Furthermore, the face associ- sent bilaterally after unilateral transcranial ated primary motor cortex was affected magnetic stimulation but also silent periods. according to CT and MRI (fig 4). Only In patients with unilateral central facial weak- recently, von Giesen and coworkers,'4 ness due to ischaemic lesions of the face associ- analysing 30 patients with lesions of different ated primary motor cortex, stimulation of the brain areas, showed that the duration of the affected hemisphere induced a pathological silent period depends on whether the primary shortening of the silent period both in the con- motor cortex is directly involved in the lesion tralateral clinically affected and in the ipsilat- or whether there is damage to areas projecting eral unaffected mentalis muscle. On the other to the primary motor cortex (for example, hand, pathologically prolonged silent periods thalamus, supplementary motor cortex). http://jnnp.bmj.com/ occur both in limb and facial muscles when Patients with the former type of lesion showed the lesion lies either outside the primary motor a shortening of the silent period in limb mus- cortex or in only a limited area of primary cles. Similar to their conclusion we therefore motor cortex which is not directly associated suggest that a shortening of the silent period in with the muscle in which the silent period is the patients is due to direct damage to measured. As discussed below these findings inhibitory intracortical interneurons, which are

favour a cortical origin of the silent period also considered to generate the silent period. on September 23, 2021 by guest. Protected copyright. in cranial muscles and suggest that this inhibitory action is bilaterally organised, as has BILATERAL CORTICAL ORGANISATION OF been shown for corticonuclear responses CRANIAL MUSCLES induced by transcranial magnetic stimula- Cranial motor nuclei receive bilateral cortical tion. 12 input via corticonuclear fibres.19 Bilateral EMG responses could be recorded from orbic- ORIGIN OF THE SILENT PERIOD IN CRANIAL ularis oris, orbicularis oculi, mentalis, stern- MUSCLES ocleidomastoid, masseter, and tongue muscles As in the discussion of the origin of the silent with a delay compatible with excitation of cor- period in limb muscles the question arises as ticonuclear pathways.'2 20 21 to whether the silent period is induced by seg- The present investigation shows that after mental interneurons in the medulla or by local stimulation of the primary motor cortex pro- inhibitory actions in the primary motor cortex. jecting to cranial nuclei, in addition to bilateral In limb muscles spinal motoneuron corticonuclear responses, silent periods in the excitability during the silent period was tested contralateral and ipsilateral cranial muscles using the H reflex technique and was found to occur. In mentalis muscle the duration is be enhanced or unchanged.8 Such experiments somewhat longer in the contralateral than the made a spinal origin of the silent period after ipsilateral side. transcranial magnetic stimulation unlikely. It could be argued that the occurrence of a Furthermore, comparisons between the spinal bilateral silent period is not compatible with a The silent period induced by transcranial magnetic stimulation in muscles supplied by cranial nerves: normal data and changes in patients 595 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from cortical origin. Taken for granted that tonic silent period in the left mentalis muscle stimu- activity in cranial muscles results from both lating over the right affected hemisphere excitation of the contralateral and ipsilateral abnormally long compared with the contralat- primary motor cortex it may be surprising that eral duration. a silent period in ipsilateral muscles after uni- As has been shown by von Giesen et al'4 in lateral transcranial magnetic stimulation limb muscles, a pathologically prolonged silent occurs at all. If the silent period originates period may occur if there is damage to areas from intracortical inhibitory intemeurons the projecting to the motor cortex (thalamus, pari- excitatory drive of the stimulated hemisphere etal cortex). Other studies have also high- should be reduced, leaving the excitatory drive lighted the influence of corticocortical of the other non-stimulated hemisphere inner- connections on the excitability of primary vating ipsilateral muscles uninfluenced. The motor cortex.2223 Although one has to inter- inhibitory drive generated by transcranial pret the results in this patient only tentatively magnetic stimulation to one hemisphere may until further supportive data are obtained, the reduce the net excitation at the motoneuronal prolongation of the silent period in this patient level so that for some motoneurons the may be due to disfacilitation of the primary remaining excitation from the other hemi- motor cortex caused by a lack of afferent corti- sphere may not be sufficient to reach firing cal input from the parietal cortex as already threshold. If this is the case, the ipsilateral proposed by von Giesen et al.'4 silent period would be expected to be incom- As opposed to facial muscles, the dominant plete. Indeed, inspection of original recordings finding in limb muscles in the remaining (see also fig 1) indicates that EMG activity is patients was also a significant prolongation of not decreased to zero during the silent period, the silent period (table 3). The question arises but rather in most cases a low level of EMG as to why such a uniform prolongation occurs. activity persists. As both the duration and the A common denominator may be that all such intensity (the reduction of the remaining patients presented with a central facial weak- EMG activity) of the silent period is more pro- ness and had a dominant lesion to the face nounced in contralateral than in ipsilateral associated primary motor cortex. Face associ- muscles especially in mentalis and genioglos- ated areas of primary motor cortex project to sus muscle (fig 1), the contralateral hemi- the limb associated primary motor cortex via sphere apparently has a stronger influence on corticocortical connections and may therefore motoneuron activation than the ipsilateral influence the activity of the pyramidal cells one. By contrast, in sternocleidomastoid mus- and local interneurons of these primary areas cle the ipsilaterally and contralaterally induced of the motor cortex. Synaptic contacts with silent period seems to be of similar duration pyramidal cells and with excitatory or and intensity-that is, both hemispheres have inhibitory interneurons are made by cortic- equal influence. This is in line with the clinical ocortical fibres in the superficial layer of pri- finding that-for example, in acute stroke-a mary motor cortex as has been shown by contralateral paresis occurs in lower facial and electron microscopy.2426 The prolongation of tongue muscles but not in sternocleidomastoid the silent period durations in limb muscles in muscle. the patients with a dominant lesion of the face associated area of primary motor cortex and a PROLONGATION OF THE SILENT PERIOD IN corresponding shortening of the silent period http://jnnp.bmj.com/ FACIAL AND LIMB MUSCLES in mentalis muscle may therefore be caused by In patient 1 the duration of the silent period deafferentation due to lesions of corticocorti- was pathologically prolonged with stimulation cal afferents from the face area to the arm or of the affected compared with the unaffected leg area of the primary motor cortex. The hemisphere (table 3, figs 2 and 5). Further- presence of pyramidal signs in the limbs (table more, silent periods in the first dorsal 1) does not, therefore, always imply shorten-

interosseus and tibialis anterior in this patient ing of the silent period as there was prolonga- on September 23, 2021 by guest. Protected copyright. were also prolonged with stimulation over the tion of the silent period in limb muscles even affected hemisphere compared with the unaf- with pyramidal signs clinically in some fected one (table 3). One may argue that the patients as has been shown previously.' 014 duration of the silent period in mentalis mus- In summary our data in normal subjects cle on the unaffected side was abnormally show that there is a bilateral silent period in short compared with the controls. It should be tonically active muscles supplied by cranial noted, however, that firstly in this patient a nerves after unilateral magnetic stimulation. stimulus intensity of only 1-2 x threshold The findings cannot conclusively prove a corti- (72% of the maximum output of the device) cal origin of the silent period. The experiments was used when recording from mentalis mus- in normal subjects and the results in patients, cle. Secondly, on statistical grounds (t test) however, strongly suggest that the silent this difference reached a P value of only 0017 period results from local inhibitory actions in compared with normal on the unaffected side the primary motor cortex. In patients, changes whereas the P value for the affected side was in silent period duration in different muscles < 0-001. In this patient MRI showed a right seem to depend on the area primarily involved parietal lesion due to resection of a haeman- in the lesion. Stimulation of the "epicentre" of gioma and no abnormalities in the left hemi- a lesion in the motor cortex, where direct sphere (fig 5). Clinically there were no motor damage to the pyramidal cells and surround- but only sensory deficits affecting the left arm. ing interneurons occurred, seems to give rise For this reason we consider the duration of the to a shortening of the silent period in related 596 Werhahn, Classen, Benecke J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.59.6.586 on 1 December 1995. Downloaded from muscles. By contrast, the lack of corticocorti- nial magnetic stimulation is of exclusive cortical origin: cal projections from the centre of the lesion evidence from isolated cortical ischemic lesions in man. to Neurosci Lett 1994; 180: 180:41-5. adjacent areas may lead to an increase in silent 11 Uozumi T, Ito Y, Tsuji S, Murai Y. Inhibitory period fol- period duration in muscles supplied by these lowing motor potentials evoked by magnetic cortical stimulation. Electroencephalogr Clin Neurophysiol 1992;85: adjacent motor cortical regions. Further studies 273-9. must concentrate on serial observations of 12 Benecke R, Meyer BU, Schoenle P, Conrad B. Trans- cranial magnetic stimulation of the human brain: variables reflecting influences between differ- responses in muscles supplied by cranial nerves. Exp ent motor cortical areas to better understand Brain Res 1988;71:623-32. 13 Benecke R, Meyer BU. Magnetic stimulation of corticonu- the pattern of functional reorganisation of the clear systems and of cranial nerves in man: physiological brain after structural damage. The silent basis and clinical application. Electroencephalogr Clin Neurophysiol (suppl) 199 1;43:333-43. period after transcranial magnetic stimulation 14 von Giesen HJ, Roick H, Benecke R. Inhibitory actions of in different muscles as an easily obtainable motor cortex following unilateral brain lesions as studied by magnetic brain stimulation. Exp Brain Res 1994;99: variable may hereby be useful. 84-96. 15 Werhahn KJ, ClaI3en J, Benecke R. Silent period induced by magnetic stimulation: Investigation of muscles innervated by cranial nerves. Mov Disord 1994;271:91. This study was supported by the Deutsche Forschungs- 16 Barker AT, Jalinous R, Freeston IL. Non-invasive magnetic gemeinschaft (SFB 194). stimulation of the human motor cortex. Lancet 1985;ii:1 106-7. 17 Inghilleri M, Berardelli A, Cruccu G, Manfredi M. Silent period evoked by transcranial magnetic stimulation of the 1 Rothwell JC, Thompson PD, Day BL, Boyd S, Marsden human motor cortex and cervicomedullary junction. J CD. Stimulation of the human motor cortex through the Physiol 1993;466:521-34 scalp. Exp Physiol 199 1;76:159-200. 18 Leis A, Kofler M, Stokic D S, Grubwieser G J, Delapasse J 2 Day BL, Dressler D, Maertens de Noordhout A, Marsden S. Effect of the inhibitory phenomenon following mag- CD, Rothwell JC, Thompson PD. Electric and magnetic netic stimulation of cortex on brainstem motor neuron stimulation of human motor cortex: surface EMG and excitability and on the cortical control of brainstem single motor unit responses. J Physiol 1989;412:449-73. reflexes. Muscle Nerve 1993;16:1351-8. 3 Fuhr P, Agostino R, Hallett M. Spinal motor neurone 19 Jenny AB, Saper CB. Organisation of the facial nucleus and excitability during the silent period after cortical stimula- corticofacial projection in the monkey: a reconsideration of tion. Electroencephalogr Clin Neurophysiol (suppl) 1991 ;81: the upper motor neurone facial palsy. Neurology 1987; 257-62. 37:930-9. 4 Haug BA, Schonle PW, Knobloch C, Kihne M. Silent 20 Muellbacher W, Mathis J, Hess CW. Die elektrophysiolo- period measurement revives as a valuable diagnostic tool gische Untersuchung der zentralen und peripheren with transcranial magnetic stimulation. Electroencephalogr motorischen Bahnen zu den Zungenmuskeln (The elec- Clin Neurophysiol 1992;85: 158-60. trophysiological investigation of central and peripheral 5 Kukowski B, Haug B. Quantitative evaluation of the silent motor pathways to tongue muscles). Zeitschrift fuir EEG period, evoked by transcranial magnetic stimulation dur- und EMG 1993;21:130. ing sustained muscle contraction, in normal man and in 21 Meyer BU, Fauth C, Liscic R, Bischoff C, Conrad B. patients with stroke. Electroencephalogr Clin Neurophysiol Organisation of descending motor tracts to motoneu- 1992;32:373-8. rones of lower facial nerves, neck muscles, tongue mus- 6 Uncini A, Treviso M, Di Muzio A, Simone P, Pullman S. cles, and proximal and distal arm muscles in man. Mov Physiological basis of voluntary activity inhibition Disord (suppl 1) 1992;7:19. induced by transcranial cortical stimulation. 22 Kujirai T, Caramia MD, Rothwell JC, Day BL, Thompson Electroencephalogr Clin Neurophysiol 1993;89:211-20. PD, Ferbert A, Wroe S, Asselmann P, Marsden CD. 7 Cantello R, Gianelli M, Civardi C, Mutani R. Magnetic Corticocortical inhibition in human motor cortex. J brain stimulation: the silent period after the motor Physiol 1993;471:501-19. evoked potential. Neurology 1992;42: 1951-9. 23 Brown P, Day BL, Rothwell JC, Thompson PD, Marsden 8 Roick H, Giesen HJV, Benecke R. On the origin of the CD. Intrahemispheric and interhemispheric spread of post-excitatory inhibition seen after transcranial magnetic cerebral cortical myoclonic activity and its relevance to brain stimulation in awake human subjects. Exp Brain epilepsy. Brain 1991;114:2333-51. Res 1993;94:489-98. 24 Jones EG. Varieties and distribution of non-pyramidal cells 9 Cammarota A, Brasil-Neto J, Valls-Sole J, Pascual-Leone in the somatic sensory cortex of squirrel monkey. J7 Comp A, Cohen LG, Hallett M. Differential inhibitory properties Neurol 1975;160:205-68. of transcranial magnetic, transcranial electric and percu- 25 Szentagothai J. Basic circuitry of the neocortex. Exp Brain taneous electric brainstem stimulation. Neurology 1993; Res (suppl) 1976;1:282-7. 43:A261. 26 Jones, EG, Porter, R. What is area 3a? Brain Res (review) http://jnnp.bmj.com/ 10 Schnitzler A, Benecke R. The silent period after transcra- 1980;2: 1-43. on September 23, 2021 by guest. Protected copyright.