Motor responses evoked by magnetic brain stimulation in Huntington's disease

Electroencephalography and clinical Neurophysiology, 85 (1992) 197-208

197

© 1992ElsevierScientificPublishers Ireland, Ltd. 0924-980X/92/$05.00

ELMOCO 91644

Motor responses evoked by magnetic brain stimulation in Huntington's disease Bernd-Ulrich Meyer, Johannes Noth a, Herwig W. Lange b Christian Bischoff, Jochen Machetanz, Adolf Weindl, Simone R6.richt c, Reiner Benecke c and Bastian Conrad Departments of Neurology, Technical University, Munich (F.R.G.), and ~Alfried Krupp Krankenhaus, Essen (F.R.G.), and the Departments of b Psychiatry and ¢ Neurology, University of Diisseldorf, Diisseldorf (F.R. G.) (Accepted for publication: 9 December 1991)

Summary

In 34 patients with manifest Huntington's disease (HD), and in 21 first-degree offspring without clinical signs or symptoms, the sizes, central motor latencies (CMLs) and variation in latencies of EMG responses (MEPs) following transcranial magnetic brain stimulation were studied in muscles of the upper and lower extremities. In subgroups of patients and their offspring median and tibial nerve somatosensory evoked potentials (SEPs) and electrically elicited long-loop reflexes (LLRs) in hand muscles were also investigated. Increased MEP thresholds were observed in 10% of the HD offspring, while CML, latency variability and MEP amplitudes always lay within normal range. In contrast, SEPs were abnormal in 33%. In HD patients MEPs were found to be abnormal in up to 72% of patients when all available response parameters were taken into consideration. MEP abnormalities correlated with the duration of motor symptoms and the severity of choreic motor activity. When both MEPs and SEPs were evaluated, abnormalities could be detected in 91% of all HD patients. We suggest that abnormal MEPs might reflect an altered excitability of the cortico-spinal system as a consequence of basal ganglia dysfunction, rather than a structural damage of the investigated descending pathways. To localize the pathological mechanism responsible for altered LLRs, a "loop analysis" was performed by recording LLRs, MEPs and SEPs in the same patients. Alterations of LLRs correlated best with abnormal SEPs and might therefore be explained by reduced somatosensory input to the motor cortex. Key words: Huntington's disease; Magnetic brain stimulation; EMG responses; Somatosensory evoked potentials; Long-loop reflexes; Pathophys-

iology

In Huntington's disease (HD) electrophysiological investigations using long-loop reflexes (LLRs) have revealed a reduction of the amplitude of, or an absence of, the second reflex component to be a characteristic finding (Noth et al. 1985; Deuschl et al. 1989). These abnormalities have been attributed to a degeneration of CNS neurons within the transcortical reflex pathway with its lemniscal afferents and efferents along the pyramidal tract (Marsden et al. 1983; Wiesendanger 1986). Since an impairment of somatosensory evoked potentials (SEPs) has also been found in HD (Noth et al. 1984; Bollen et al. 1985), the well known loss of neurons in the thalamus (Lange 1981; Roos 1986) could be the cause of abnormal LLRs and SEPs. Magnetic brain stimulation (Hess et al. 1987) provides a new technique for the evaluation of the functional integrity of the efferent cortico-spinal system in HD. In

Correspondence to: Dr. Bernd-Ulrich Meyer, Department of Neurology, Technical University of Munich, Moehlstr. 28, 8000 Munich 80 (F.R.G.). * Supported by the Bundesministerium fiir Forschung und Technologie, 01 KL 9001.

contrast to previous studies of HD patients which reported normal motor responses (MEPs) in conduction studies using electrical (Berardelli et al. 1988; Caramia et al. 1988; Thompson et al. 1988a) and magnetic brain stimulation (Eisen et al. 1989; H6mberg and Lange 1990), this study revealed abnormal MEPs following transcranial magnetic brain stimulation when a wider range of response parameters was taken into account. Similarly, changes in MEPs have been reported for other diseases primarily affecting the basal ganglia such as Wilson's disease (Berardelli et al. 1990; Meyer et al. 1991), progressive supranuclear palsy (Abbruzzese et al. 1991), and in some cases of Parkinson's disease (Eisen et al. 1990). Investigation of the discharge characteristics of single motor units during stationary isometric contractions of small hand muscles has revealed irregular firing patterns in patients with HD (Dengler et al. 1986). The question arises as to whether fluctuations in the excitation level of motoneurons can also be detected in relaxed hand muscles and as to whether or not such fluctuations could be used as a diagnostic criterion. To answer these questions the variability of motor responses following repetitive magnetic brain stimulation has been evaluated. This paradigm has recently been

198 introduced in the investigation of activated hand muscles in order to investigate impaired cortico-spinal impulse transmission in patients with multiple sclerosis (Britton et al. 1991). We investigated the function of the cortico-spinal system in patients with H D and some of their offspring (subjects at risk) by evaluating different parameters of motor responses (response threshold, central motor latency, response amplitude, latency variability) elicited by magnetic brain stimulation. Additionally, SEP and LLR have been studied in a subgroup of patients to estimate the usefulness of motor responses in detecting abnormalities and to identify the pathological mechanism which produces the abnormal long-loop reflexes.

Methods

Patients and general procedure Thirty-four patients with a clear diagnosis of H D based on history and on clinical and radiographical findings and 21 first-degree offspring of H D patients (subjects at risk) were examined using magnetic stimulation. In addition, 3 patients with obvious choreic movements were examined using transcranial electrical brain stimulation following the principles described previously by Benecke et al. (1988). All 21 subjects at risk (mean age 34 years, S.D. 10; age range 17-51; 9 males) were free of clinical indications of HD. A subgroup of H D patients and subjects at risk also underwent a standardized examination of SEPs (n = 22 H D patients, n = 15 subjects at risk) and electrically elicited LLRs ( n - - 9 H D patients, n = 7 subjects at risk). The H D patients were placed in 2 groups on the basis of visual observation of the extent of the choreic movements. Patients in group I (n = 12; mean age 39 years, S.D. 9; age range 27-59, 6 males) showed sporadic mild hyperkinesia of the distal limbs, often in the form of stereotyped flexion/extension movements of the fingers. The duration of choreic movements was 2.1 years (mean, S.D. 1.4; range 0.5-5). One patient out of this group showed hyperkinesia only after ingestion of alcohol, another only during light sleep. One patient was on medication with tiapride. Group II included 22 patients (mean age 51 years, S.D. 12; age range 33-70; 14 males) with obvious moderate to severe choreic movements, often involving the cranial, proximal limb, and trunk muscles. H D II patients were clearly older than H D I patients and had a longer history of choreic movements (mean 6.6 years, S.D. 3.8; range 2-15 years). In this group 13 patients were on medication with tiapride (n = 4), sulpiride (n = 3), neuroleptics (n = 2), nitomane (n = 1), tranquillizers (n = 5), or tricyclic antidepressants (n = 3), either as monotherapy or in combination.

B.-U. MEYER ET AL.

Routine SEP and LLR Routine SEPs and LLRs were recorded following electrical stimulation of the median nerve (for details of stimulation and recording procedure refer to Deuschl et al. (1989) and Noth et al. (1984) respectively). Magnetic stimulation and recording The motor cortex was stimulated using the 1.5 Tesla version of the Magstim 200 (Novametrix). The large circular coil (consisting of 19 turns of copper band, inner diameter 5.5 cm, outer diameter 11.6 cm) was centered over the vertex (for hand muscle responses) or 3 cm anteriorly (for leg muscle responses). For examination of responses in right-sided muscles the coil was laid flat on the skull so that the direction of the current was anticlockwise around the coil when viewed from above. In this paper, the direction of coil currents is given to the newest information available (Day et al. 1990). For examination of left-sided responses the coil was turned over so that the coil current flowed clockwise. Conduction studies. In the determination of motor thresholds the subjects were asked to keep the muscles under investigation relaxed. During the conduction studies the subjects were asked to make a sustained voluntary contraction of the first dorsal interosseus (FDI) and anterior tibial (TA) muscles of about onethird of the individual maximal tonic force by squeezing a force transducer with the hand or by elevating the forefoot. Stimulation was then performed with 1.5 times the threshold or with maximal stimulation strengths when the threshold was higher than 66% of the maximum output of the magnetic stimulator. Peripheral conduction times were obtained by magnetic stimulation of the motor roots with the same coil. Central motor latencies (CMLs) were calculated by subtracting the longest peripheral conduction time following magnetic root stimulation from the onset latency of the fastest cortically evoked muscle compound action potential. A minimum of 5 stimuli was applied to the brain and nerve roots in order to be certain of obtaining responses with the shortest onset latency following brain stimulation and the longest conduction time following root stimulation. Onset latencies were determined to the first negative deflection. Amplitudes (peak to peak) of the MEPs were expressed as a percentage of the size of the maximal M wave elicited by stimulation of the appropriate peripheral nerve. Variability studies. The variability in latency of the cortically elicited MEPs was described as the mean consecutive difference (MCD) between the onset latencies of 20 responses of the right FDI. Again, stimulation strengths were set at 1.5 times the excitation threshold. The repetition rate was between 0.4 and 0.3 stimuli/sec, depending on the stimulation strength.

MOTOR RESPONSES IN HUNTINGTON'S DISEASE

199

Records were made with the target muscle at rest, and were obtained with A g / A g C I electrodes placed over the muscle belly. The signals were amplified by a Toennies Myograph II with bandpass filtering between 20 Hz and 3000 Hz. Data were collected and stored on a hard disk using a Tandon personal computer and a

Amplitude [% M]

CED data collection program (Sigavg, sampling frequency 5000 Hz/channel).

Normal ualues for CML and response amplitude In order to define normative data a group of 57 normal subjects (age range 21-81) was investigated in

Amplitude [% M] 100

100

I I

FDI

80

FDI

80 I I

60

60

+

"•

•

40 •

%,

•

••

"•11

I

+ +

40

+

I

+'"+, .+, t

I+

+

I

, +, 4I-

+ +

|

20

20

+

+ 0

,

'1

4

,

6

i

I

!

8

10

I

0

i

12

u

4

u

4.+

J

i

I

8

6

-'1

10

i

OML [ms]

OML [ms] Amplitude [% M]

,

12

Amplitude

[9£ M]

100

1O0

TA

80

60

YA

80

60

+ + +++

+ •

40

+

+

40

+

o

"++

+

++ •

20

+

20 _

+ _

_

-4-, + 0

10

14

18

22

26 CML [ms]

,

10

14

.

n

-~

_+_

%'~, + , + ,

18

_1 •

+,

22

i

v

26 OML [ms]

Fig. 1. Motor responses elicited by transcranial magnetic brain stimulation in the first dorsal interosseus (FDI) and anterior tibial (TA) muscles on both sides in 21 subjects at risk (left diagrams) and in 34 patients with Huntington's disease (right diagrams). Patients with moderate to severe impairment in motor control are indicated by crosses, mildly affected patients by points. The central motor latency time is plotted against the peak-to-peak response amplitude, expressed as a percentage of the size of the M response. Vertical and horizontal lines indicate the normal range of CML and amplitudes, respectively. Responses were regarded as pathological when they lay below the horizontal lines or to the right of the vertical lines. Responses were absent in 4 TA muscles of patients and are not displayed in the diagram.

200

B.-U. M E Y E R ET AL.

TABLE I Excitation thresholds for motor responses elicited by magnetic brain stimulation in the relaxed first dorsal interosseus muscles of a control group (C), subjects at risk (R), and patients with manifest Huntington's disease presenting with mild ( H D I) or with moderate to severe impairment of motor control (HD II) (for group criteria see text). Subgroups of patients without (a) and with medication (b). Thresholds were considered to be increased when they were higher than 58% of the maximal stimulator output. Patient

Examined

Motor thresholds

group

subjects (n)

Increased in n subjects

% max. stimulator output m e a n + S.D. (range)

C

12

-

44.2+5.9 (36-53)

R

21

2 (10%)

51.0+8.6 (40-75)

HD I Total

Normal

FDI

C 12

4 (33%)

a

11

3 (25%)

b

1

H D II Total

MEPs elicited in the right FDI muscle was determined in 12 normal subjects with the muscle at rest and with stimulation strengths set at 1.5 times the excitation threshold. The mean consecutive latency difference (MCD) of 20 responses was 0.32 msec (S.D. 0.14). On the basis of these data and accepting a range of 2.5 S.D. as normal, MCDs of less than 0.7 msec were regarded as normal.

1

22

7 (32%)

a

9

3 (33%)

b

13

4 (31%)

55.8 + 13.6 * (43-90) 53.0+9.2 (43-70) 90 55.2+ 12.0 * (30-85) 51.1 + 8.6 (30-60) 58.6+ 13.8 (40-85)

20.8

C 26.6

* Significantly different from C, level < 0.03, Tukey H S D test.

HD

FDI

our laboratory following the described standardized examination procedure using magnetic brain and nerve root stimulation (for details see Kloten et al. 1992). In this study an age-related increase of the CML was found. To be sure that the determined CMLs were also valid for patients of higher age the upper limit of normality was defined on the basis of data obtained in 18 subjects older than 59 years. Under these circumstances the upper limits for CML to FDI and to T A motoneurons were 9.3 msec (mean (M): 6.5, single standard deviation (S.D.): 1.1 msec) and 20.9 msec (M: 16.1, S.D.: 1.9 msec) respectively, with a range of 2.5 S.D. being considered normal. The lower limits for the size of the responses (in relation to the maximal M wave) were 17% (M: 45, S.D.: 17) for the FDI and 13% (M: 40, S.D.: 22) for the TA, using the smallest amplitudes recorded among the entire population of normal subjects. Mean excitation thresholds were determined for the FDI muscle at rest and were 44% (S.D. 6) as determined by the output display of the stimulator (n = 12 subjects). Normal L,alues for response variability The

variability

of the

onset

latency

of consecutive

TA

R

- ~

2.5mV L

10 ms

Fig. 2. Motor responses in the first dorsal interosseus (FDI) and anterior tibial muscle (TA) elicited by magnetic stimulation of the cortex (C) and nerve roots (R) in a normal subject and in a patient with Huntington's disease (HD) and severe choreic movements. The cortically evoked responses of the patient were delayed and reduced in amplitude (FDI) or absent (TA), while the responses following root stimulation were normal. Age and body height of the normal subject and the H D patient were 21 and 35 years and 173 and 175 cm respectively.

MOTOR RESPONSES IN HUNTINGTON'S DISEASE

201

TABLE II Conduction studies: number of muscles with MEP abnormalities in patients with Huntington's disease (HD I and H D II, for group criteria refer to text) and subjects at risk (R). Subgroups of HD patients without (a) and with medication (b). Records were made bilaterally from the first dorsal interosseus and anterior tibial muscles. MEPs were considered abnormal when at least one of the response parameters (amplitude or central motor latency) was abnormal. Patient

Subjects

Subjects with increased CML a n d / o r reduced ampl. or absent responses

group

examined (n)

Total (n)

R

21

HD I Total a b HD II Total a b

1 muscle (n)

2 muscles (n)

3 muscles (n)

4 muscles (n)

0

0

0

0

0

12 11 1

6 (50%) 5 (45%) 1

3 3 0

1 1 0

0 0 0

2 1 1

22 9 13

15 (68%) 7 (78%) 8 (62%)

7 5 2

2 2 0

2 0 2

4 0 4

Results

Thresholds of motor responses In comparison to control subjects, the thresholds for cortically evoked MEPs in the FDI muscles were increased (i.e., higher than 58% of the maximal stimulator output) in about one-tenth of the subjects at risk and in about one-third of the H D I and H D II patients (Table I). Applying the Tukey HSD multiple comparisons test (ANOVA), H D I and H D II patients differed significantly from the normal subjects (significance levels < 0.03, < 0.02). In H D II patients motor thresholds tended to be higher in subjects on medication.

However, it cannot be excluded that the patients on medication were also more severely affected in terms of the impairment of motor control, at least before treatment.

CML and amplitudes of motor responses Fig. 1 and Tables II and III show the results of the standardized magnetic brain stimulation used to evaluate the CML and amplitudes of the compound muscle action potentials of the FDI and T A muscles. In all 21 subjects at risk the CML and response amplitude in FDI and T A muscles lay within the normal ranges. Out of 34 H D patients (groups H D I and

TABLE III Patterns of abnormalities in the central motor latency (CML) and the amplitude of motor responses elicited by magnetic brain stimulation in the first dorsal interosseus (FDI) and anterior tibial (TA) muscles. Subjects at risk and patients with manifest Huntington's disease were investigated (HD I and HD II, for group criteria see text). Subgroups of patients without (a) and with (b) medication. In each pair of figures, the first number refers to the FDI, the second to the TA muscle. Patient

Total

Any

Number of muscles with

group

number FDI/TA (n/n)

abnormality (n/n) (%)

Increased CML (n/n)

R

42/42

HD l Total

24/24

a

22/22

b

2/2

HD II Total

44/44

a

18,/18

b

26/26

0/0

Reduced ampl. (n/n)

Increased CML and red. ampl. (n/n)

Absent responses (n/n)

0/0

0/0

0/0

0/0

7/6 (29%/25%) 5/4 (23%/18%) 2/2

0/0

4/4

3/0

0/2

0/0

3/4

2/0

0/0

0/0

1/0

1/0

0/2

18/15 (41%/34%) 4/5 (22%/28%) 14/10 (54%/38%)

4/2

10/9

4/2

0/2

1/0

3/5

0/0

0/0

3/2

7/4

4/2

0/2

202

B.-U. M E Y E R ET AL.

II), 21 (i.e., 62%) showed abnormalities for at least 1 of the 2 response parameters in at least 1 of the 4 recorded muscles (Table II). These figures correspond with abnormalities in 46 out of 136 (i.e., 34%) investigated muscles (Table III). Abnormalities were more frequent in hand than in leg muscles (37% versus 31%) probably due to a better defined upper limit of normality for the FD1 muscles. An absence of response occurred only in leg muscles. Peripheral conduction times lay within the normal range for all subjects at risk as well as H D patients. Comparing the patients with mild dyskinesia (HD I) with those presenting with moderate to severe choreic movements (HD II), the H D II patients showed abnormal MEPs more often (68% versus 50%) and, for a given subject in a larger number of muscles (Tables II and III). In H D ] and H D II patients the most frequent patterns of abnormality were reduced response sizes for both the FDI and T A muscles. Fig. 2 displays original records of an H D II patient with abnormal cortically elicited MEPs; responses following nerve root stimulation were normal.

Variability of motor responses with magnetic brain stimulation To establish a parameter for the response variability, the mean consecutive difference (MCD) of the onset latencies of 20 consecutive responses in small hand muscles was calculated. In comparison to normal subjects, H D I and H D II patients had significantly different MCDs of latency (Tukey HSD multiple comparisons test, significance levels < 0.05 and < 0.002). Defining a normal range of 2.5 S.D. (MCD <0.7 msec), all subjects at risk lay within this range in contrast to 57% of the patients with HD (Fig. 4). In 3 out of 10 HD I patients and 10 out of 20 H D II patients the MCD lay above the 2.5 S.D. limit. Out of 13 H D II patients under medication, 7 had increased MCDs. Out of 7 patients without medication, 3 had an abnormal latency variability. Representative original records of an H D II patient with an increased MEP variability are displayed in Fig. 3. In 32 H D patients MEP abnormalities (prolonged CML a n d / o r reduced response amplitude) in the right FDI were compared with the variability of latency of

Normal Latency

~

27

2mV

[ms]

J

26 0.2mV

25 24 ~

~0ms

23~ I

10 ms

5ms

22

1

5

10

15

20

Trial

HD Latency 27

ImV

25 24 !~Oms

23 I

10 ms

I

1

5

10

15

20

Trial Fig. 3. Variability of motor responses in the FD1 muscle following magnetic brain stimulation in a normal subject and a patient with Huntington's disease (HD). Left box: superposition of 20 consecutive E M G responses. Center box: magnification of the onset of the E M G responses in the left box. Right box: trial-to-trial variation of the onset latency. In the patient the latency times are prolonged and show a non-systematic fluctuation.

M O T O R RESPONSES IN HUNTINGTON'S DISEASE

203

To establish whether the increased variability of latency observed in H D patients paralleled solely choreic motor activity, 5 normal subjects were asked to

MCD [ms] 1,6

Normal (~ MCD 0,28

1.2 []

d3

lq I

Iq

0.8 []

I

[

I

I

III

Normal MCD 0,27

[:3 F1

[]

0,4 f

n=12

n=14

n=10

n=20

R

HDI

HD II

, ,!,l ,

Fig. 4. Variability of latency times of consecutively elicited responses in the FDI muscle following magnetic brain stimulation. Control subjects (C), subjects at risk (R), and patients with Huntington's disease (HD 1 and HD II, for group criteria refer to text) were investigated. Patients receiving medication are indicated by an asterisk. Latency variation is expressed as the mean consecutive difference (MCD) of onset latency times of 20 motor responses. The vertical bars indicate the mean MCD plus/minus 1 S.D.

consecutive responses in the same muscle. Standard MEPs were abnormal in 10 patients (i.e., 31%); the latency variability was increased (MCD > 0.7 msec) in 15 patients (i.e., 47%). In 9 patients the response variability was increased while routine hand muscle MEPs were normal. When standard MEPs were evaluated for all muscles of the upper and lower extremities, abnormalities were detected in 59% of the patients in this group and in 72% when the latency variability in right hand muscles was also taken into consideration. To decide whether an increased latency variability resulted from jumps between two distinct latencies or from an irregular latency fluctuation, poststimulus time histograms of the response onset latency were evaluated for all subjects. In normal subjects, a clustering of latencies within a time span of 1 msec was observed for muscles at rest, which became even more pronounced when the muscle was voluntarily activated (Fig. 5). In all H D patients with an increased MCD of onset latencies the clustering of the latencies was reduced or even absent, while the latencies themselves varied by up to 6 msec (Fig. 5).

I

HD M©D 0,67

0 C

I

HD II MCD 0,88

I

I

I

1

III IIIII I

I

I

HD II MOD 1,55

6 4 2 I

15

I

I

I

I

20

I

II II IIIIII I! I

I

I

I

25

I

I

30

Latency [ms] Fig. 5. Poststimulus time histograms (PSTHs) of onset latency times of 20 consecutive responses elicited by magnetic brain stimulation in the first dorsal interosseus muscle. In the upper 2 boxes the PSTHs are given for the tonically contracted ( + ) and relaxed muscle in one normal subject. In the lower 3 boxes PSTHs are displayed for different HD patients with the option to keep the target muscles as relaxed as possible. The inserted figures give the mean consecutive latency difference (MCD) of the responses. With increasing MCD the onset latencies scattered over a broader range and latency clustering diminished (bin width 0.2 msec, stimulation strength 1.5 times excitation threshold for the relaxed muscle).

204

B.-U. MEYER ET AL.

imitate severe hyperkinetic movements. The MCDs of 50 consecutive E M G responses amounted at rest to 0.26, 0.35, 0.31, 0.42 and 0.29 msec and in the same subjects during imitated choreic movements 0.74, 1.18, 1.16, 1.08, and 0.97 msec. During imitated choreic movements the latencies varied in normal subjects to the same order of magnitude as in H D II patients with true chorea who were asked to keep the muscle as relaxed as possible during recording (Fig. 6). In 3 H D patients with an increased response variability 'at rest,' the latencies of 50 hand muscle re-

tHD MCDO,91

,,, !, Jl!J,l,llJJ!J,, , Normal

711111!J11! o I ,,,

6 4

Ih

,, ......

Normal MOD 1,12

2

!

16

I

1

I

I

20

IiI,II,i I,,I I

I

I

I

I

26

I

I

I

I

80

Latency [ms] Fig. 6. Poststimulus time histograms (PSTHs) of onset latency times of 50 consecutive responses elicited by magnetic brain stimulation in the right first dorsal interosseus muscle of 1 patient with HD (upper box),and 2 healthy neurologists. The HD patient was asked not to move, but the normal subjects were asked to imitate severe choreic movements of the right arm (bin width 0.2 msec, stimulation strength 1.5 times excitation threshold for the relaxed muscle).

sponses were plotted against the corresponding peakto-peak size. In all 3 patients the onset latency tended to shorten with increasing response amplitudes, as is shown for one patient in Fig. 7.

Variability of motor responses with electrical brain stimulation In 3 patients with H D anodal brain stimulation was also performed in order to decide whether the increased variability of latency observed with magnetic stimulation resulted mainly from fluctuations in the excitability at the motor cortical or spinal level. The electrical stimulation strength was adjusted so that the sizes of the magnetically and electrically elicited responses were of about the same amplitude. In comparison to magnetic stimulation, the latency variability of electrically elicited hand muscle responses was reduced. In the same patients the MCD amounted to 0.80, 0.72 and 0.91 msec (magnetic stimulation) and 0.41, 0.63 and 0.78 msec (electrical stimulation). The corresponding onset latencies of the muscle compound action potential were 22.1 + 0.6, 27.0 + 0.7, 25.2 + 1.3 msec (mean plus 1 S.D.; magnetic stimulation) and 21.4 + 0.4, 27.1 + 0.8, 24.8 + 1.1 msec (electrical stimulation). Obviously, no D-wave activation occurred during electrical stimulation. Correlation with SEP abnormalities In a subgroup of subjects at risk and H D patients CMLs and amplitudes of cortically elicited MEPs in the F D I and T A muscles were compared with SEPs following electrical stimulation of the median and tibial nerves. MEPs were regarded as pathological when the CML, the amplitude, or both, were pathological on at least one side. Median and tibial nerve SEPs were considered as pathological when the early potential was absent, the N20 or N30 latency prolonged, or the amplitude of the N 2 0 / P 2 5 or N 3 0 / P 4 0 component reduced on at least one side (for details of criteria see Noth et al. 1984). In subjects at risk and H D patients, abnormalities were more frequently observed for SEPs than for MEPs. One-third of 15 subjects at risk had abnormal SEPs, while none of these subjects had abnormal MEPs. In 17 out of 22 H D patients median a n d / o r tibial nerve SEPs were pathological (i.e., 77%) while only 11 patients out of this group had abnormal standard MEPs (i.e., 50%). The number of patients with MEP abnormalities increased to 15 (i.e., 68%) when an increased variability of latency of hand muscle responses was regarded as pathological. When the median and tibial nerve SEPs were compared separately with motor responses in the upper and lower extremities, the combined use of both techniques increased the number of detected abnormalities from 45% (median nerve SEPs alone) to 64% of the H D patients (SEP and MEP),

M O T O R R E S P O N S E S IN t t U N T I N G T O N ' S D I S E A S E

205

and MEPs and the other had only abnormal SEPs. In general, abnormal LLRs were predominantly associated with SEP abnormalities.

Amplitude [mY] 6

Discussion D

4 i

i 1 •

•

IN

| m m

0

i

20

t

l

22

i

• i

24

•

-" m " i

I

I

l

26 28 Latency [ms]

Fig. 7. Onset latency times plotted against the corresponding peakto-peak amplitude of 50 consecutively elicited E M G responses in the first dorsal interosseus muscle. Exemplary results of one patient with Huntington's disease and severe choreic movements. The patient was instructed to keep the recorded muscles as relaxed as possible. Stimulation strength was set at 80% of the maximal output of the stimulating device.

while there were no additional abnormalities detected by cortically elicited standard MEPs in leg muscles when compared with tibial nerve SEPs. However, when all available parameters for MEP abnormality (including latency variability in hand muscles) were evaluated in combination with median and tibial nerve SEPs, abnormalities could be detected in all but 2 patients (i.e., 91%).

Combined evaluation of LLR, SEP and MEP In 7 subjects at risk and 9 HD patients, cortically elicited motor responses in the FDI, median nerve SEPs and electrically elicited long-loop reflexes in thenar muscles were investigated. Out of the 7 subjects at risk two showed abnormal long-loop reflexes, one had abnormal SEPs, but none had abnormal motor responses. In the two subjects at risk with abnormal LLRs the SEPs were also pathological in one, while in the other SEPs and MEPs were normal. Out of the 9 H D patients, 7 had abnormal LLRs, 8 had abnormal SEPs and 2 had abnormal MEPs. Out of the 7 patients with abnormal LLRs, 6 had abnormal SEPs and 1 abnormal MEPs. One patient with abnormal LLRs had normal SEPs and MEPs. Of the 2 patients with normal LLRs, one had abnormal SEPs

Transcranial magnetic stimulation of the motor cortex is thought to excite cortical motoneurons transsynaptically and thereby to elicit descending impulse volleys in the fast conducting component of the pyramidal tract (Day et al. 1987; Hess et al. 1987). The descending impulses consist of multiple excitation volleys which lead to an activation of spinal alpha motoneurons and consecutive E M G responses. Abnormalities of such E M G responses, as observed in a high percentage of HD patients, could therefore reflect (1) an alteration of the excitability of motor cortical cells, (2) an impaired impulse propagation along the involved descending motor pathways, and (3) an imbalance of temporo-spatial integration of excitatory and inhibitory inputs at the level of spinal motoneurons. Which parts of the cortico-spinal system or structures influencing its function are affected in HD and could cause MEP abnormalities? The non-specific changes in the cortically elicited responses, with increased excitation threshold, reduced response size and sometimes moderately prolonged latency do not allow the affected structures to be identified. However, a severe demyelination in the descending motor tracts, which may be revealed in some cases of multiple sclerosis by markedly prolonged central motor latency (Ingram et al. 1988), can be ruled out in HD. An influence of the non-systematic fiber degeneration in the spinal cord (Hallervorden 1957) which sometimes occurs in H D cannot be excluded but is less probable since clinical pyramidal tract signs are rarely seen. Furthermore, an anterograde degeneration of the cortico-spinal pathways is not to be expected, since the primary motor cortex (area gigantopyramidalis) seems to be spared from the general cortical degenerative process. In contrast to all other cortical areas, area 4 was found to have a normal volume in morphometric studies (Lange 1981). At the spinal segmental level no losses of motoneurons have been observed (Spielmeyer 1928), which is in accordance with the normal peripheral conduction times seen in our study and with normal monosynaptic spinal reflex responses following mechanical stretches or peripheral nerve stimuli (Noth et al. 1985; Deuschl et al. 1989). It could therefore be postulated that the MEP abnormalities result from a functional and not from a structural change, which alters the excitability at the motor cortical or spinal level. This change in excitability could be secondary to basal ganglia dysfunction or an alteration of other facilitatory motor cortical inputs such as somatosensory

206 afferents, which are known to be impaired in H D (Noth et al. 1984; Bollen et al. 1985). Similar changes of standard MEPs were also observed in other basal ganglia disorders like Wilson's disease (Meyer et al. 1991). How could basal ganglia dysfunction explain increased motor thresholds, reduced sizes and increased latencies of cortically evoked E M G responses in HD? The basal ganglia, being part of a system of cortex-basal ganglia-cortex loops influence the activity and excitability of the motor and premotor cortex (Delong and Georgopoulos 1979). These loops include the striatum, the pallidum, the substantia nigra, the subthalamic nucleus and the thalamus, which can all be severely affected in HD (Dom and Malfroid 1976; Lange et al. 1976; Roos 1986). According to a recently developed model of basal ganglia dysfunction in early H D (Albin et al. 1989), disinhibition of lateral globus pallidus neurons, induced by degeneration of striatofugal neurons, causes diminished activity in the subthalamic nucleus. As is known to be the case from hemiballismus, this finally results in disinhibition of excitatory thalamo-cortical projections to the motor and premotor cortex. The net effect of this altered cortical input is a facilitation of cortico-spinal (CS) neurons, leading to the involuntary movements in HD. At a first glance, such a facilitation of CS neurons might be expected to lower their excitation threshold, rather than to raise it. However, our data obtained with magnetic brain stimulation do not support this simplistic view. If facilitation of CS neurons was the sole pathological change in the central motor system in HD, the slowness of voluntary movements seen in H D patients would be difficult to explain (Hefter et al. 1987; Thompson et al. 1988b). It thus appears that magnetic stimulation, with its presumed transsynaptic activation of CS neurons, resembles the pathological recruitment of CS neurons during voluntarily induced innervation as performed in the form of tonic muscle contraction during the standardized investigation of MEPs. Abnormalities of MEPs indicated manifest H D and paralleled the progression of the disease, as the number of abnormal responses correlated with the severity and duration of choreic movements. In subjects at risk without any motor symptoms, the MEPs lay within the normal range, but statistical analysis for the entire population of subjects at risk shows higher motor thresholds and higher response variability in comparison with normal subjects. Medication might have an effect on MEPs, especially D2 receptor antagonists, since excitation thresholds were higher in patients taking medication than those without medication, but with about the same degree of choreic movements. However, it is also possible that the patients on medication had a greater degree of motor impairment and basal ganglia dysfunction before treatment was

B.-U. MEYER ET AL. started. The variable patterns of MEP abnormality, with sometimes normal and sometimes abnormal cortically elicited responses within one patient, may reflect the often asymmetric and patchy distribution of the degenerative process in the basal ganglia of H D patients (Roos 1986). The finding of normal MEPs in HD by others (H6mberg and Lange 1990) may be due to the use of maximal stimulation strengths and therefore preferential D-wave activation and of different stimulators. Some of the patients described by H6mberg and Lange (1991) as having normal MEPs were also investigated in our laboratory and had abnormal MEP when using our standardized magnetic brain stimulation procedure. In addition to the abnormalities seen in the standardized conduction study, an increased latency variability was observed in about half of the HD patients after applying a repetitive stimulation paradigm. When the patients were asked to keep the recorded muscle totally relaxed and when stimulation conditions were the same in each trial, the onset latency of cortically elicited compound action potentials often varied irregularly within a wide range, as was apparent in the poststimulus time histograms. With the muscle at rest normal subjects showed a clustering of latencies within a narrow latency range which shifted to shorter latencies and became even more synchronized when the muscle was tonically activated. These latency peaks of the compound action potential are thought to reflect a suprathreshold excitation of fast conducting alpha motoneurons by different components (D and I waves) of the multiple descending excitation volley (Day et al. 1989). For HD patients it may be concluded that the descending volley was more desynchronized the more markedly the patients presented with choreic motor activity and the more the related descending activity influenced the excitation level of the CS system. The latency variability in H D patients must be attributed to fluctuations of excitability at the cortical and motoneuronal levels, since in comparison to magnetic stimulation the variability was only slightly smaller during electrical brain stimulation, with its assumed excitation site distal to the motor cortical cell soma (Day et al. 1986). That the latency variation parallels at least partly the motor activity could be confirmed by experiments in which healthy subjects imitated choreic movements. Under such conditions the response latencies varied, as was the case in H D patients with severe chorea. However, in some patients a high fluctuation of onset latencies could also be observed in phases when no choreic movements were visible or E M G activity was present in the recorded target muscle. Therefore, the latency variability might be a useful parameter to quantify the impairment of motor control in a distinct motoneuron pool, independent of visible movements and detectable E M G activity. Furthermore, in compar-

MOTOR RESPONSES IN HUNTINGTON'S DISEASE

ison with MEP conduction studies, the evaluation of response variability increases the diagnostic sensitivity of magnetic brain stimulation in detecting MEP abnormalities in HD patients. In order to answer the question of whether the impaired function of the descending motor pathways contributes to the lack of the second EMG component of long-loop reflexes in HD patients (Noth et al. 1985; Deuschl et al. 1989), a "loop analysis" was performed, on the assumption that long-loop reflexes follow transcortical pathways (Marsden et al. 1983; Wiesendanger 1986; Deuschl et al. 1991). To this end, the pathology of electrically elicited LLRs in hand muscles was correlated with the results of the investigation of lemniscal afferents (SEPs) and pyramidal efferents (MEPs) in the same patients. Absent or reduced components of the LLR were predominantly associated with a diminution of the early somatosensory evoked cortical potential. This supports the hypothesis that the alteration of LLRs is a direct consequence of the reduced somatosensory input to the cerebral cortex (Noth et al. 1985). In an earlier study on HD patients, no relationship between the size of the long-latency stretch reflexes of the flexor pollicis longus or the wrist muscles and the size of the median nerve SEP (N20/P25) was found (Thompson et al. 1988b). This finding does not necessarily contradict our assumption, as those experiments were performed on intermediate arm muscles, in which the receptor contribution to the long-latency stretch reflex may differ from that in distal hand muscles (Noth et al. 1991). Is there a role for MEPs in the neurophysiological diagnosis of HD and presymptomatic detection of subjects at risk? In comparison to routine median and tibial nerve SEPs and electrically and mechanically elicited LLRs, the percentage of detected abnormalities was lower when using a standardized conduction study with magnetic brain stimulation. Routine MEPs within the standardized conduction study were abnormal in 59% of patients with manifest HD. The percentage of patients with MEP abnormalities could be increased to 72% when the latency variability in hand muscles was also taken into account. The combined evaluation of all available MEP parameters, including latency variability and SEPs, increased the percentage of HD patients with abnormalities to 90%. In subjects at risk SEPs were abnormal in more than 30% (Noth et al. 1984; this study), while electrically elicited LLRs (Deuschl et al. 1989) and MEPs were normal. This raises the question as to whether or not the function of the cortico-spinal system is normal when MEPs are normal, which was found in about one-fourth of HD patients and in all the subjects at risk, or whether the test is insensitive in detecting abnormal motor function. The latter seems to be the case, at least for subjects at risk, since single motor

207

unit records during sustained firing can already reveal an impaired motor control in the form of a "microchorea" in such subjects (Petajan et al. 1979; Dengler et al. 1986).

Addendum

Recently we have investigated the effects of D2 antagonists on MEPs. In normal subjects no changes were observed. In HD patients response amplitudes further decreased and latencies increased. These findings suggest that D2 antagonists further reduce the already impaired input to the motor cortex from the basal ganglia in HD, rather than directly depress motor cortex excitability. We should like to thank Reinhold Sojer and Hermann Riescher for technical and computational assistance and Drs. John C. Rothwell and Stuart Fellows for reading the manuscript.

References Abbruzzese, G., Tabaton, M., Morena, M., Dall'Agata, D. and Favale, E. Motor and sensory evoked potentials in progressive supranuclear palsy. Movem. Dis., 1991, 6: 49-54. Albin, R.L., Young, A.B. and Penney, J.B. The functional anatomy of basal ganglia disorders. Trends Neurosci., 1989, 12: 366-375. Benecke, R., Meyer, B.-U., G6hmann, M. and Conrad, B. Analysis of muscle responses elicited by transcranial stimulation of the cortico-spinal system in man. Electroenceph. clin. Neurophysiol., 1988, 69: 412-422. Berardelli, A., Inghilleri, M., Priori, A., Accornero, A. and Manfredi, M. Electrical stimulation of motor cortex in patients with motor disturbances. In: P.M. Rossini and C.D. Marsden (Eds.), Non-lnvasive Stimulation of Brain and Spinal Cord: Fundamentals and Clinical Applications. Liss, New York, 1988: 219-230. Berardelli, A., Inghilleri, M., Priori, A., Thompson, P.D., Fabri, S., Fieschi, C. and Manfredi, M. Involvement of corticospinal tract in Wilson's disease. Movem. Dis., 1990, 5: 334-337. Bollen, E.L., Arts, R.J., Roos, R.A., Van der Velde, E.A. and Buruma, O.J. Somatosensory evoked potentials in Huntington's chorea. Electroenceph. clin. Neurophysiol., 1985, 62: 235-240. Britton, T.C., Meyer, B.-U. and Benecke, R. Variability of cortically evoked motor responses in multiple sclerosis. Electroenceph. clin. Neurophysiol., 1991, 81: 186-194. Caramia, M.D., Zarola, F., Spadaro, M., Pardal, A.M. and Bernadi, G., Neurophysiologic testing of the central impulse propagation characteristics in patients with sensorimotor disorders. In: P.M. Rossini and C.D. Marsden (Eds.), Non-Invasive Stimulation of Brain and Spinal Cord: Fundamentals and Clinical Applications. Liss, New York, 1988: 193-206. Day, B.L., Dick, J.P.R., Marsden, C.D. and Thompson, P.D., Differences between electrical and magnetic stimulation of the human brain. J. Physiol. (Lond.), 1986, 378: 36P. Day, B.L., Rothwell, J.C., Thompson, P.D., Dick, J.P.R., Cowan, J.M.A., Berardelli, A. and Marsden, C.D. Motor cortex stimulation in intact man. 2. Multiple descending volleys. Brain, 1987, 110: 1191-1209. Day, B.L., Dressier, D., Claus, D., Hess, C.W., Maertens de Noordhour, A., Marsden, C.D., Nakashima, K., Rothwell, J.C. and Thompson, P.D. Erratum: Direction of current in magnetic stim-

208 ulating coil used for percutaneous activation of brain, spinal cord, and peripheral nerve. J. Physiol. (Lond.), 1990, 430: 617. Delong, M.R. and Georgopoulos, A.P. Physiology of the basal ganglia - a brief overview. Adv. Neurol., 1979, 23: 137-154. Dengler, R., Wolf, W., Schubert, M. and Struppler, A. Discharge pattern of single motor units in basal ganglia disorder. Neurology, 1986, 36: 1061-1066. Deuschl, G., Liicking, C.H. and Schenk, E. Hand muscle reflexes following electrical stimulation in choreatic movement disorders. J. Neurol. Neurosurg. Psychiat., 1989, 52: 755-762. Deuschl, G., Michels, R., Berardelli, A., Schenk, E., Inghilleri, M. and Liicking, C.H. Effects of electric and magnetic transcranial stimulation on long latency reflexes. Exp. Brain Res., 1991, 83: 403-410. Dom, R. and Malfroid, M. Neuropathology of Huntington's chorea. Neurology, 1976, 26: 64-68. Eisen, A.A., Bohlega, S., Bloch, M. and Hayden, M. Silent periods, long-latency reflexes and cortical MEPs in Huntington's disease and at-risk relatives. Electroenceph. clin. Neurophysiol., 1989, 74: 444-449. Eisen, A.A., Siejka, S. and Calne, D. Amplitude changes of the MEP elicited by magnetic stimulation with aging and in Parkinson's disease (PD): further evidence for overlap between western PD and amyotrophic lateral sclerosis. Muscle Nerve, 1990, 13: 880881. Hallervorden, J. Huntingtonsche Chorea (Chorea chronica progressiva hereditaria). In: W. Scholz (Ed.), Handbuch der speziellen pathologischen Anatomie und Histologie, Vol. 13. Springer, Berlin, 1957: 793-822. Hefter, H., H6mberg, V., Lange, H.W. and Freund, H.-J. Impairment of rapid movement in Huntington's disease. Brain, 1987, 110: 585-612. Hess, C.W., Mills, K.R. and Murray, N.M.F. Responses in small hand muscles from magnetic stimulation of the human brain. J. Physiol. (Lond.), 1987, 388: 397-419. H6mberg, V. and Lange, H.W. Central motor conduction to hand and leg muscles in Huntington's disease. Movem. Dis., 1990, 5: 214-218. Ingram, D.A., Thompson, A.J. and Swash, M. Central motor conduction in multiple sclerosis: evaluation of abnormalities revealed by transcutaneous magnetic stimulation of the brain. J. Neurol. Neurosurg. Psychiat., 1988, 51: 487-497. Kloten, H., Meyer, B.-U., Britton, T.C. and Benecke, R. Magnetic stimulation: standardized determination of normal values for central and peripheral motor latencies regarding age-dependent changes. Z. EEG-EMG, 1992, in press.

B.-U. MEYER ET AL. Lange, H.W. Quantitative changes of telencephalon, diencephalon, and mesencephalon in Huntington's chorea, postencephalitic, and idiopathic parkinsonism. Verh. Anat. Ges., 1981, 75: 923-925. Lange, H., Th6rner, G., Hopf, A. and Schr6der, K.F. Morphometric studies of the neuropathological changes in choreatic diseases. J. Neurol. Sci., 1976, 28: 401-425. Marsden, C.D., Rothwell, J.C. and Day, B.L. Long-latency automatic response to muscle stretch in man: origin and function. In: Advances in Neurology. Motor Control Mechanisms in Health and Disease. Raven Press, New York, 1983: 509-539. Meyer, B.-U., Britton, T.C., Bischoff, C., Machetanz, J., Benecke, R. and Conrad, B. Abnormal conduction in corticospinal pathways in Wilson's disease. Movem. Dis., 1991, 4: 320-323. Noth, J., Engel, L., Friedemann, H.-H. and Lange, H.W. Evoked potentials in patients with Huntington's disease and their offspring. I. Somatosensory evoked potentials. Electroenceph. clin. Neurophysiol., 1984, 59: 134-141. Noth, J., Podoll, K. and Friedemann, H.H. Long-loop reflexes in small hand muscles studied in normal subjects and patients with Huntington's disease. Brain, 1985, 108: 65-80. Noth, J., Schwarz, M., Podoll, K. and Motamedi, F. Evidence that low-threshold muscle afferents evoke long-latency stretch reflexes in human hand muscles. J. Neurophysiol., 1991, 65: 1089-1097. Petajan, J.H., Jarcho, L.W. and Thurman, D.J. Motor unit control in Huntington's disease: a possible presymptomatic test. Adv. Neurol., 1979, 23: 163-176. Roos, R.A.C. Neuropathology of Huntington's chorea. In: P.J. Vinken, G.W. Bruyn and H.L. Klawans (Eds.), Handbook of Clinical Neurology, Vol. 49. Elsevier, Amsterdam, 1986: 315-326. Spielmeyer, W. Die anatomische Krankheitsforschung am Beispiel einer Huntingtonschen Chorea mit Wilsonschem Symptomenbild. Z. Neurol., 1928, 101: 701-728. Thompson, P.D., Berardelli, A., Rothwell, J.C., Day, B.L., Dick, J.P.R., Benecke, R. and Marsden, C.D. Pathophysiology of tics and chorea. In: R. Benecke, B. Conrad and C.D. Marsden (Eds.), Motor Disturbances, I. Academic Press, New York, 1988a: 213230. Thompson, P.D., Berardelli, A., Rothwell, J.C., Day, B.L., Dick, J.P.R., Benecke, R. and Marsden, C.D. The coexistence of bradykinesia and chorea in Huntington's disease and its implications for theories of basal ganglia control of movement. Brain, 1988b, 111: 223-244. Wiesendanger, M. Experimental evidence for the exsistence of a proprioceptive transcortical loop. In: H.-J. Freund, U. Biittner, B. Cohen and J. Noth (Eds.), Progress in Brain Research, Vol. 64. Elsevier, Amsterdam, 1986: 67-74.