The pathophysiology of myoclonus

The pathophysiology of myoclonus

perspectives on disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thep...

889KB Sizes 0 Downloads 5 Views

Recommend Documents

No documents
perspectives

on disease

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

ThepaUIophy$iologyof myodonu$ Mark Hallett Myoclonus is defined as a quick involuntary movement generated by dysfunction of the CNS. There are so many types of myodonus that classification has been difficult. Recent investigations have shed some light on the physiology of some forms of myodonus and have led to a new physiological classification. Additionally, since it appears that some types of myodonus represent hyperactivity of normal brain pathways, analysis of myodonus may be useful in the understanding of normal brain mechanisms of motor control.

Involuntary movements are common presenting symptoms in neurological patients. These movements disrupt voluntary movement and are unattractive and disquieting. Myoclonus, one type of involuntary movement, is a quick muscle jerk arising from dysfunction of the CNS. There are many types of myoclonus, and there are no common etiological, physiological or therapeutic features that bind them together. It is clearly critical from a clinical point of view to be able to classify a case of myoclonus in order to understand or treat a patient. An etiological classification is valuable in that it may lead to specific therapy, but this is unfortunately not commonly possible. A physiological classification may lead to symptomatic therapy since the physiology and pharmacology of a movement disorder should be strongly related. Physiological studies aimed at determining a physiological classification have been fruitful, not only for the classification itself, but also for the insights that they have produced concerning the functioning of the brain. One reason for this is that many types of myoclonus appear to represent hyperexcitability of brain pathways that ordinarily participate in motor function. Hyperexcitability leads to involuntary rather than voluntary movement. The first reasonable approach to physiological classification was that of Halliday, which divided rnyoclonus into three types, pyramidal, extrapyrarnidal and segmental 1. On the basis of new data, a new scheme (Table I) has been proposed by Marsden, Hallett and Fahn which can be seen as an elaboration of the original 2-4. Pyramidal myoclonus is probably equivalent to cortical myoclonus; extrapyramidal myoclonus, characterized by longer EMG bursts and a less obvious EEG correlate than pyramidal myoclonus, includes a number of new categories; segmental myoclonus is preserved. Some of these types of myoclonus will be described in this article. One principal feature of the current scheme is the acceptance of the idea, first proposed by Muskins, that certain types of myoclonus are fragments of epilepsy5. The jerks of myoclonus can be viewed clinically as a small part of the convulsions of epilepsy, and the physiology can be identical. The basic building block of epilepsy is the paroxysmal depolarization shift (PDS), which causes a neuron to produce a rapid burst of action potentials. Groups of nearby neurons firing synchronously may be TINS- February 1987 [10]

associated with an interictal (or isolated) spike on the EEG. If the output of these hyperactive neurons is excitatory to the motor system, the result would be a rnusclejerk appreciated clinically as myoclonus. Local rhythmic recurrence would be epilepsia partialis continua (continuous rhythmic jerking of one body part); local spread of hyperexcitability would be a Jacksonian march (jerking that begins with one small part of the body and gradually spreads to involve additional adjacent body parts); full generalization would be a grand rnal seizure. There are three types of epileptic rnyoclonus, which will be described in detail: cortical reflex myoclonus, reticular reflex rnyoclonus, and primary generalized epileptic rnyoclonus.

Mark Hallettis at the NationalInstituteof Neurologicaland Communicative DisordersandStroke, Building10, Room 5N226, National Institutesof Health, Bethesda,MD 20892, USA.

Cortical reflex myoclonus Cortical reflex myoclonus is thought to be a fragment of focal or partial epilepsy6-8. Each rnyoclonic jerk involves only a few adjacent muscles, typically only an antagonist pair, although larger jerks with more muscles can also be seen. Many or all muscles in the body can be affected by different jerks, so that the rnyoclonus is 'rnultifocal' (the term 'generalized', which means all muscles are involved in a single jerk is not applicable). Myoclonus is spontaneous, and can be accentuated by both voluntary movement (action or intention myoclonus) and somatosensory stimulation (reflex myoclonus). Similar cases without reflex features can be referred to as spontaneous cortical rnyoclonus2. The genesis of cortical reflex rnyoclonus is thought to be hyperexcitability of sensorimotor cortex, with each jerk representing the discharge of a small region involving at most a few contiguous muscles. There are four pieces of evidence in favor of this hypothesis. (1) Only a few contiguous muscles are involved in any jerk, and this is what would be expected with TABLEI. Physiologicalclassificationof myoclonus Epileptic:

Cortical Reticular Primarygeneralizedepileptic

Segmental Palatal Spinal Peripheral Essential Ballisticmovementoverflow Dystonic Startlesyndromes Exaggeratedstartle Hyperekplexia Jumping Nocturnal myoclonus Asterixis Categoriesin italicare discussedin the text. For reviewof other categoriesand more in-depth coverage,see Refs 2-4. Non-epileptic:

© 1987, ElsevierScience Publishers B.V., Amsterdam 0378- 5912/87/$02.00

69

:~I~'.

.

.

-

;;~

.

~

.





.

~

.

~ .........,~;:; ~

.

~..........~,....

A.

R, WoEXT.

FPt

-- AI

T3

-- AI

TS

-- A1

F3

-- A1

C3

--A1

1:'3

--At

01

-- A1

F4

-- A2

C'l

--A2

P4

--A2

02

--A2

FP2

--A2

Fill

-

T4

-- A2

L - -

--

~

--

"

~

"

....

~

--

"

~

--

"

"

L[

. . . . .

.

.

.

_

-

A2

~

~

~

, --

.

-

,

.

.

.

.

.

.

E F

T,

~Tt

o

20

to

30

4o

so

M

tlo

~o

8o

Iio

" C,

P~

~oo

s

B.

R. W.EXT

F?

--A!

~ .....

._

_

T3

-- A;

......

;.

~;

F3

-- A!

P3

--A1

--

01

-- A!

-

F4

-- A2

..

~, ,

,

-

_

;-

-.42

-

....

02

- - ~

.

.

FP2 - A2 -_: --FI

-- A2

T4

--A2

11

- - A 2 - -

,

..

-

-

CA - A~ ..... - . PwI

,

.

.

_ .

.

.

_

.

z-:

. . . . . .

...... ~..; . . . . .

,.

Tj

C3

T~

P~

C,

T,

T

4.

M

o,1/

Fig. 1. (A) Electrical correlates o f spontaneous myoclonic jerking in patient with cortical reflex myoclonus. Top left is a sin&le E/vlG record. Bottom left is avera&ed EEG before and after the jerk. Upper right is a topo&raphic map of the EEG at the time indicated on the bottom left. Li&ht shadin& is ne&ative and dark shadin& is positive. (B) Electrical correlates of myoclonic jerks produced by right median nerve stimulation o f the same patient as in (A). (Taken from Ref. 3 and adapted, with permission, from Ref. 19.) 70

TINS- February 1987 [10]

hyperexcitability of a small cortical region. (2) When muscles innervated by cranial nerves are involved in a muscle jerk, the precise timing of onset of activation suggests that the set of impulses that generate the myoclonus travels the brainstem in a rostrocaudal direction. Specifically, the masseter (fifth cranial nerve) is active before the orbicularis oculi (seventh cranial nerve), which is itself active before the sternocieidomastoid (eleventh cranial nerve). (3) There is a focal, time-locked EEG event that precedes both spontaneous and reflex-induced myoclonic jerks. The focal area of cortex is the appropriate region of the sensorimotor strip contralateral to the involved muscles. The event, a positive-negative transient, appears to be the same for both spontaneous and reflex-induced jerks (Fig. I). (4) In at least two cases with focal myoclonus and epilepsia partialis continua, a small region of cortical hyperexcitability was identified by direct cortical recording, and the myoclonus was eradicated when the cortical region was excised. Moreover, in experimental animals it is possible to produce reflex myoclonus by application of either alumina cream or penicillin to a focal area of motor cortex. The precise details of the physiological processes underlying the cortical event are yet not known. The delay between the cortical discharge and the muscle jerks is similar to the delay between direct cortical stimulation and evoked muscle response, and ranges from 10--30 ms. Somatosensory-evoked potentials (SEP) measured on the scalp are giant in patients with this type of myoclonus. In the SEP produced by median nerve stimulation, the segment corresponding to the discharge in spontaneous jerks (the positive-negative transient) is the P1-N2 segment, and this is characteristically one of the enlarged segments. Dawson and Halliday had pointed out some years ago that the amplitude of the SEP correlated with the magnitude of the myoclonic jerk; more recently Rothwell et aL9 have demonstrated that the best correlation is with the P1-N2 segment when analysing individual jerks. These facts would be compatible with the hypothesis that the EEG event is due to a synchronous discharge in a focal area of motor cortex and that the underlying cellular event is a paroxysmal depolarization shift (PDS). The more neurons involved, the bigger the jerk. The larger the area of activation, the more muscles involved. The negativity of the correlative EEG component is not unexpected, since the EEG correlate of a PDS is usually negative, and the component of the premotor potential before voluntary movement thought to be due to motor cortex firing is also negative. It has been attractive to think that reflex generation of cortical myoclonus stems from the hyperexcitability of a normal transcortical reflex pathway. The existence of transcortical reflex pathways has been sugg~ ~ted as an explanation for the recently identified str :h reflexes that are longer in latency than the monosynaptic stretch reflex 1°. If these pathways exist and motor cortex is hyperexcitTINS- February 1987 [10]

Fig. 2.

1=3 O1 F4

- A1 - A1 - A2'

C4 - A2 P4 - A 2 02 - ~,

m o e

~

-

-

eo

~ -

_

.

iota

FP2F8 - ~'

Averaged EEG correlate of spontaneous myodonicjerkin& in patient with primary generalized epileptic myodonus. EMG correlate occurredat 250 ms. Topographic map of EEGat the time indicated (arrow) is made similar to that in Fig. 1. (Adapted,

withpermission,from Ref.13.)

.J J J J

J

f

able, then reflex myoclonus should be generated. A number of features suggest that it is the sensory rather than the motor cortex that is hyperexcitable. Indeed, if the sensory cortex could drive the motor cortex effectively through a direct pathway, the motor cortex itself would not need to be hyperexcitable. For example, some patients show selectivity in the type of stimulation that will produce the myoclonus. For some, stretch is the critical stimulus; for others it is light touch. This kind of selectivity reflects the sensory segregation known to occur in somatosensory cortex. Rothwell eta/. have recently demonstrated that the ~iant SEP is not linked obligatorily to a myoclonic jerk y. Acute intravenous administration of lisuride or clonazepam reduces the severity of the myoclonic jerks without affecting (or even enhancing) the amplitude of the SEP. The explanation for this phenomenon is not known. One possibility is that uncoupling could occur between a hyperexcitable region of sensory cortex and motor cortex. Additional evidence that it is the sensory rather than the motor cortex that is hyperexcitable comes from one case in which the site of the giant SEP was found by corticography to be post-central 7. Reticular reflex m y o c l o n u s

Reticular reflex myoclonus is thought to be a fragment of a type of generalized epilepsy 11. Myoclonic jerks typically affect the whole body; proximal muscles are affected more than distal ones 71

perspeotives and flexors are more active than extensors. In some patients, only a segment of the body is affected, such as both legs, and all the muscles of the segment are involved with each jerk. Myoclonus can be spontaneous or induced by action or somatosensory stimulation (reflex). As with cortical reflex myoclonus, the critical somatosensory stimulus can be modality specific. The genesis of reticular reflex myoclonus is thought to be hyperexcitability of a portion of the caudal brainstem reticular formation. Analogy to animal models, as will be discussed later, suggests that the site of origin is the nucleus reticularis gigantocellularis 12. The data supporting this conclusion are as follows: (1) Many muscles on both sides of the body participate simultaneously. There is significant jitter of body parts with respect to each other, so that in different jerks the arm might precede the leg or the leg precede the arm. The jitter may be consistent with the multiple synapses in the reticular formation itself and between the terminals of the reticulospinal tract and the anterior horn cells. The widespread influence and the proximal and flexor preponderance are appropriate characteristics. (2) When muscles innervated by cranial nerves are involved in a muscle jerk, the precise timing of onset of activation suggests that the set of impulses that generate the myoclonus travels in a caudorostral direction through the brainstem. Specifically, the trapezius (eleventh cranial nerve) is active before the orbicularis oris (seventh cranial nerve), which is itself active before the masseter (fifth cranial nerve). Activation of eleventh cranial nerve muscles is the earliest manifestation of a myoclonic jerk and suggests an origin near the level of the medulla. (3) A spike in the EEG is often associated with myoclonic jerk. The spike is generalized in distribution with highest amplitude at the vertex and usually follows the first electromyographic sign of the myoclonus. Additionally, the spike is not time-locked to the myoclonus. These facts suggest that the spike does not originate in cortex, but is a result of a subcortical event and is not directly responsible for the myoclonus. Moreover, the SEP is not enhanced, so there is no evidence for cortical hyperexcitability. The plausibility of brainstem reticular origin of myoclonus is strongly supported by animal studies. Systemic urea infusion in cats induces myoclonic jerking as a prelude to generalized seizures, and electrical recording has revealed that these discharges originate from the nucleus reticularis gigantocellularis 12. Microelectrode recordings in this area revealed cellular activity similar to paroxysmal depolarization shifts (PDS). Cobalt powder implanted in this region will also cause myoclonus. The anatomical pathway for reflex myoclonus in reticular reflex myoclonus is not yet clear. The spinobulbo-spinal pathway does not seem to be the candidate since for that reflex the afferent spinal pathway is rapidly conducting and the efferent pathway is slowly conducting; this situation does not hold for reticular reflex myoclonus 11. 72

on d

sease

Primarygeneralizedepileptic myoclonus Primary generalized epileptic myoclonus, which we have only recently described, is thought to be another fragment of generalized epilepsy 13. Specifically, it is a fragment of the type of epilepsy called primary generalized epilepsy. Clinically, the myoclonus takes several forms. Most common in our experience are small, focal jerks often involving only finger movements. The appearance is that of minor finger twitching or tremulousness, which might be described as minipolymyoclonus. The term minipolymyoclonus was originally coined to refer to small jerks seen in patients with motor neuron disease. Minipolymyoclonus of central origin and minipolymyoclonus of peripheral origin have a similar clinical appearance and are probably most easily separated by their association with epilepsy and muscle denervation respectively. A second clinical presentation of primary generalized epileptic myoclonus is generalized, synchronized whole body jerks, not unlike those seen with the reticular reflex myoclonus. This type of myoclonus is thought to arise from firing of a hyperexcitable cortex driven synchronously by ascending subcortical impulses. The evidence for this is: (1) Myoclonic jerks involve muscles on both sides of the body with exquisite synchrony. This is clinically, as well as electrophysiologically evident, in the big jerks. This is also true for the small jerks, but it can be demonstrated only with electromyographic recording. The small jerks are a fragmentary manifestation of a generalized central nervous system event. (2) When cranial muscles are involved, activation through the brainstem is in a rostrocaudal direction. This differs from the other type of generalized myoclonus, reticular reflex myoclonus. (3) The electroencephalographic correlate to the myoclonus is a generalized event, a slow, bilateral fronto-centrally predominant negativity that pre~ cedes the myoclonus. For small jerks this event lasts 100-250 ms and jitters in time with respect to the jerk (Fig. 2). For large jerks the event lasts 30-100 ms and is more time-locked with respect to the jerk. Such a distribution is consistent with subcortically driven central EEG rhythms such as sleep spindles and 3 Hz spike-and-wave (which is the EEG correlate of the petit mal type of primary generalized epilepsy). (4) Eight of the eleven patients described in our report had primary generalized epilepsy, including one with typical genetic petit mal epilepsy. The physiology of primary generalized epilepsy remains obscure and controversial. The most popular current view is Gloor's corticoreticular theory 14. The first feature is that the whole cortex receives input from reticular formation and non-specific thalamic nuclei (the 'centrencephalon'). Frequent activity in this path causesthe 'recruiting rhythm' in the normal awake brain and sleep spindles in the sleeping brain. The second feature of the theory is that the cortex is generally hyperexcitable and responds to the subcortical input with a paroxysmal event. In animals it has been shown that if the cortex is bathed in dilute TINS- February 1987 [10]

perspectives

o n disease

penicillin, subcortical input will induce 3 Hz spikeand-wave activity. One of our patients showed a remarkable B Hz bilateral fronto-central rhythm (time-locked with the myoclonic jerks), which had the appearance of the recruiting response. Our hypothesis about primary generalized epileptic myoclonus is that the jerks are caused by the influence on the motor system of the waves of excitability in the hyper-responsive cortex. Ballistic movementoverflow myoclonus Essential myoclonus is myoclonus occurring in the absence of any other disturbance of the CNS including epilepsy, dementia, cerebellar disturbance or EEG abnormality. Such disorders can be sporadic or familial, and the etiology is unknown. Ballistic movement overflow myoclonus is one type of familial essential myoclonus 15. Its genetics are autosomal dominant, and it is related to essential tremor in some families. Myoclonic jerks that can be generalized in distribution are frequently associated with voluntary movement. The pathophysiology is thought to be hyperactivity of the mechanism that generates the initiation of voluntary movement. When normal subjects make voluntary rapid movements (ballistic movements) of a single joint, the EMG activity pattern is characteristic. The agonist muscle begins with a burst of activity that lasts approximately 100 ms and then silences. At about the time of the silence, the antagonist produces a burst of activity of approximately 100 ms. Then the agonist resumes activity. This pattern is sometimes called the 'triphasic pattern'. Ordinarily some muscles distant from the moving joint may also show some low levels of activation, and some of this activity may be functional, for example, to fixate the more proximal part of the limb. The myoclonus in these patients is characterized by the appearance of large amplitude triphasic patterns in widespread muscles with the same timing of the activation of appropriate muscles. It is as if the command to generate a ballistic movement has overflowed into an excessive number of muscles with excessive amplitude. The steps taken by the brain to generate movement are still only vaguely understood. One step must be the specification of which muscles should participate. Cases like this demonstrate the failure to limit participation. It may well be that one of the fundamental deficits in Parkinson's disease is just the opposite 16, i.e. the failure to generate sufficient participation.

Exaggeratedstartle The startle reflex produces a rapid generalized body activation in response to an unexpected stimulus. It may be useful in bracing the body to counteract any perturbation and may prime the motor system for subsequent specific action. There are patients who 'startle' even when the stimulus loses its novelty or who have a grossly exaggerated response; they fall into the category of the

TINS- February 1987 [10]

..................

L..

exaggerated startle syndromes and their disorder can be confused with other types of myoclonus. Hyperekplexia, or 'startle disease', is in fact clearly a unique type of myoclonus rather than exaggeration of the startle reflex 17. This disorder is frequently familial and is characterized by onset, early in life, of jerking with an exaggerated response to sensory stimulation, including falling to the ground. Physiological investigations have shown some similarities to both cortical reflex myoclonus and reticular reflex myoclonus. True exaggerated startle does exist and has been seen in a number of patients with a variety of CNS lesions 17. Electrophysiological features of the startle reflex to auditory stimuli are well-defined and can be used to recognize this disorder. At least in animals, the neural pathway for the audiogenic startle reflex is known, and follows an oligosynaptic pathway in the brainstem through the nucleus reticularis pontis caudalis TM. This is a different brainstem circuit than that used by reticular reflex myoclonus. Perhaps it is differential loss of inhibition that leads to the hyperactivity of one reflex rather than another.

Selectedreferences 1 Halliday,A. M. (1967) Brain 90, 241-284 2 Marsden, C. D., Hallett, M. and Fahn, S. (1982) in Butterworth's International Medical Reviews (Neurology, VoL 2: Movement Disorders) (Marsden, C. D. and Fahn, S., eds),

pp. 196-248, Butterworth Scientific 3 Hallett, M. (1985) Epilepsia 26 (Suppl. 1), $67-$77 4 Hallett, M., Marsden, C. D. and Fahn, S. in Handbook of Clinical Neurology (Bruyn, G.W. and Klawans, H. L., eds), Elsevier(in press) 5 Muskens,L. J. J. (1928) Epilepsy: Comparative Patho&enesis, Symptoms and Treatment, William Wood 6 Hallett, M., Chadwick, D. and Marsden, C.D. (1979) Neurolo&y 29, 1107-1125 70beso, J. A., Rothwell,J. C. and Marsden,C. D. (1985) Brain 108, 193-224 8 Shibasaki,H., Yamashita,Y., Neshige,R., Tobimatsu, S. and Fukui, R. (1985) Brain 108, 225-240 9 Rothwell, J. C., Obeso, J. A. and Marsden, C. D. (1984) J. NeuroL Neurosurg. Psychiatr. 47, 33-42 10 Marsden,C. D., Rothwell,J. C. and Day,B. L. (1983) in Motor Control Mechanisms in Health and Disease (Desmedt, J. E., ed.), pp. 509-539, RavenPress 11 Hallett,M., Chadwick,D., Adam,J. and Marsden,C. D. (1977) J. NeuroL Neurosurg. Psychiatr. 40, 253-264 12 Zuckerman, E. G. and Glaser,G. H. (1972) Arch. Neurol. 27, 14-28 13 Wilkins, D. E., Hallett, M. and Erba, G. (1985) J. NeuroL Neurosurg. Psychiatr. 48, 506-516 14 Gloor, P. (1979) Epilepsia 20, 571-588 15 Hallett, M., Chadwick, D. and Marsden, C. D. (1977) Brain 100, 299-312 16 Hallett, M. and Khoshbin,S. (1980) Brain 103,301-314 17 Wilkins, D. E., Hallett, M. and Wess,M. M. (1986) Brain 109, 561-573 18 Davis,M., Gendelman,D. S.,Tischler,M. D. and Gendelmen, P. M. (1982) J. Neurosci. 2,791-805 19 Wilkins, D. E., Hallett, M. and Berardelli,A., Walshe, T. and Alvarez, N.

Please mention TINS when responding to advertisements

73