HEMIBALLISMUS: /ETIOLOGY AND SURGICAL TREATMENT

J. Neurol. Neurosurg. Psychiat., 1950, 13, 115.

HEMIBALLISMUS: /ETIOLOGY AND SURGICAL TREATMENTBY

RUSSELL MEYERS, DONALD B. SWEENEY, and JESS T. SCHWIDDEFrom the Division of Neurosurgery, State University of Iowa, College of Medicine,

Iowa City, Iowa

Hemiballismus is a relatively uncommon hyper-kinesia characterized by vigorous, extensive, andrapidly executed, non-patterned, seemingly pur-poseless movements involving one side of the body.The movements are almost unceasing during thewaking state and, as with other hyperkinesias con-sidered to be of extrapyramidal origin, they ceaseduring sleep.

Clinical AspectsCases are on record (Whittier, 1947) in which the

abnormal movements have been confined to a singlelimb (" monoballismus ") or to both limbs of bothsides (" biballismus ") (Martin and Alcock, 1934;von Santha, 1932). In a majority of recordedinstances, however, the face, neck, and trunk as wellas the limbs appear to have been involved.The patient usually retains a measure of voluntary

control over the affected member so that acts suchas walking, putting on slippers, and touching thehand to the face can be performed even if withdifficulty. The ballistic movements cannot howeverbe voluntarily checked for more than a few momentsat a time. The muscle tonus of the involved limb(s)may be normal but in many instances is slightlydiminished (Martin and Alcock), and in consequenceof the reduced initial tension of the muscles the deepreflexes of the affected limbs are often moderatelydiminished.The vast majority of patients are at or beyond

middle life at the time of onset of hemiballismus.However, reports of patients as young as seven(Bonhoeffer, 1930) and eighteen (Bianchi, 1909)years have been recorded. There appears to be nosignificant difference in sex incidence.

Spontaneous recovery is not to be expected andby far the majority of patients succumb within afew days to several months of progressive exhaustion,cardiac failure, or pneumonia (Grinker and Bucy,

*Case I of this communication was reported with motion pictureillustrations at the Sixteenth Annual Meeting of the Harvey CushingSociety on August 19, 1948, at San Francisco, Cal. The paper inits present form was given with motion picture illustrations at theSeventy-fourth Annual Meeting of the American NeurologicalAssociation on June 13, 1949, at Atlantic City, N.J.

1949; Whittier). A few instances are on record inwhich the disorder has run an extended chroniccourse (Touche, 1901 ; Marcus and Sjogren, 1938),while in one case reported by Lea-Plaza and Uiberall(1945) the abnormal movements are said to haveceased spontaneously after seven weeks. Hemi-ballismus has also been known to cease followingthe supervention of a haemorrhagic ictus.

Terminology.-There appears to be among writerson this subject no agreement regarding the precisefeatures of the clinical phenomena to which theterm hemiballismus may properly be applied.Various authors have credited Kussmaul and Fischer(1911) with introducing the term hemiballismus tosignify the flinging or flipping character of the limbmovements, but in general each observer's conceptof the features of hemiballismus represents but anabstraction of his own limited experiences. Theconsequence is that at most clinical demonstrationsof a case alleged to be one of hemiballismus one ormore members of the audience are moved to assertthat the case in question is not " really " an exampleof hemiballismus, but one of hemichorea, athetosis,dystonia, or perhaps hysteria. If, on the otherhand, the case is presented initially as an instanceof hemiathetosis or chorea, similar protests may beraised.The terms athetosis, chorea, and ballismus refer

to involuntary, non-patterned movements of bodilymembers, which are unpredictable in respect of timeand form. They may be distinguished from themyorhythmias, myoclonus, the coarse, alternatingtremors of Parkinsonism, and the fine tremors ofthyrotoxicosis. They are present in the waking stateand absent during sleep. They appear both " at rest'"and during the execution and maintenance of"voluntary" movements. In our view, the threevarieties may be distinguished from one another onthe following grounds. In athetosis the movementsare relatively slow, vermiform and writhing, and inthe limbs are chiefly due to the action of distalmuscles. The proximal limb muscles may alsoparticipate. Athetotic movements are character-

115

116 RUSSELL MEYERS, DONALD B. SWEENEY, AND JESS T. SCHWIDDE

istically intermittent so that periods of severalminutes may elapse without the exhibition ofinvoluntary activity. In chorea the movements are

much more sudden and, insofar as the limbs are

concerned, are subtended chiefly by action of themore distal rather than the proximal muscles. Asin athetosis, choreic movements are frequentlyintermittent. Ballistic movements resemble choreicmovements in that they also are rapidly executed.The muscles of the face, neck and/or trunk may or

may not be involved. The proximal muscles of thelimbs are regularly involved, with the consequence

that the limbs exhibit large excursions. Closeinspection however usually reveals involvementalso of the more distal limb muscles. In contrastto the intermittent activity characteristic of choreaand athetosis, that of hemiballismus is almostceaselessly at play during the wakeful state.

Additional Clinical Manifestations.-The intellec-tual, emotional, and sensory functions are seldomderanged in hemiballismus. However, the coexis-tence of psychological aberrations, vegetative dis-turbances, hemiparesis, dysphasia and/or oculo-

motor disorders with hemiballismus has not in-frequently been reported. In most such instancesnecropsy has demonstrated multiple or extendedlesions involving neural structures other than thoseconsidered responsible for the ballistic movements.

Aetiology and PathologyThe pathological process most frequently en-

countered in hemiballismus is a circumscribedencephalomalacia which is usually the result ofcerebral thrombosis. Diabetes mellitus not infre-quently underlies the thrombotic process. Somewhatless frequently small circumscribed hemorrhagesand emboli appear to be the responsible agents(Martin and Alcock, 1934). Other lesions, in-cluding tuberculoma, gumma, metastatic abscess,metastatic neoplasm, encephalitis and the co-calledprimary degenerative processes of the brain, havebeen identified. In two instances trauma has beenconsidered the causative agent (Bucy, 1944;Schob, 1920).The anatomical sites at which lesions have been

encountered may be conveniently described undertwo main headings.

Lesions Involving the Corpus Subthalamicum.-The structure most frequently involved in hemi-ballismus is the corpus subthalamicum contralateralto that side of the body exhibiting abnormal move-

ments. Greiff (1883) is credited with having beenthe first to observe this correlation. In a carefullyexecuted study forty years later this observationwas firmly substantiated by Jakob (1923). Subse-

quent post-mortem studies have provided overfifty instances in which the corpus Luysi has beendamaged, though often along with other neuralstructures. Indeed, the number of studies in which,like those of Matzdorff (1927) and Thurel andGrenier (1947), the only demonstrable brain lesionis limited strictly to the corpus subthalamicum isvery small.

In a review of the literature up to 1934, Martinand Alcock (1934) asserted that no case of hemi-ballismus had been fully described in which thecorpus subthalamicum was not found damaged.This extreme position no longer appears tenable,but Whittier's (1947) observation that hemi-ballism is the " apparently inevitable symptom inman of destruction of the subthalamic nucleus" isprobably quite correct.

Lesions Involving Other Structures.-Several casesof hemiballismus have been reported in which thecorpus subthalamicum is asserted to have been foundintact at necropsy. In these the lesion consideredresponsible for the hyperkinesia was disclosed inother parts of the brain.

Lesions of the Afferent or Efferent Fibres of theCorpus Subthalamicum.-In two of the earliest caseson record, those of Bianchi (1909) and Bonhoeffer(1897), the abnormal movements were imputed(Martin, 1928) to lesions at a level ")lower than"the corpus Luysi, presumably situated so as todestroy its efferent connexions to the midbrain.That damage to either the afferent or efferent fibresof the nucleus might produce hemiballismus waspostulated in 1939 by Moersch and Kernohan.Subsequently, Papez, Bennett, and Cash (1942)described a case which they considered illustrativeof the point. The presence of multiple lesions intheir case unfortunately equivocates their interpre-tation. Kelman (1945), however, recorded thenecropsy findings of a hemiballistic subject in whicha small metastatic carcinoma from the lung involvedthe afferent subthalamic tracts.

Lesions of the Corpus Striatum.-In 1926 Fragnitoand Scarpini described the necropsy findings in an80-year-old male who had suffered from hemi-ballismus for three years before his death. Anencephalomalacic process had markedly damagedthe medial portion of the putamen and, to a lesserdegree, the head of the caudate nucleus. Thecorpus Luysi, medial lemniscus, and fields of Forelwere reported to be undamaged.* Several otherinvestigators (Vogt and Vogt, 1920; Austregesiol

*This is the only case acknowledged by Martin and Alcock toconstitute a valid exception to the doctrine concerning the necessaryinvolvement of the corpus subthalamicum in the pathogenesis ofhemiballismus. Its importance theoretically and practically isobvious.

SURGICAL TREATMENT OF HEMIBALLISMUS

and Gallotti, 1924) have reported cases in whichlesions of the corpus striatum were consideredresponsible for hemiballismus or for hyperkinesiassimulating the latter. Certain of these cases havebeen objected to by Martin and Alcock (1934) onthe ground that conclusive evidence clearing thecorpus subthalamicum of damage has not beenproffered. However, Austregesilo and Borges-Fortes (1937) and Davison and Goodhart (1938)have each reported well-studied cases in which theneostriatum alone has been the seat of damage.Although their cases were published under the titlesof " hemichorea " and " monochorea ", the des-criptions of the abnormal movements conform soclosely to those of hemiballismus that differentiationappears difficult if not impossible.

In Case 2 of the present communication, thelesion apparently responsible for the patient'shyperkinesia was disclosed in and adjacent to theright lenticular nucleus, i.e., involving both theneo- and paleostriatum. Histological examinationrevealed no implication of the corpus Luysi.

Lesions in the Thalamus.-Lewandowsky andStadelman (1912) reported a case of hemiballismusin which a large lesion was disclosed in the lateralthalamic nuclei extending inferiorly into Forel'sHaubenfelde. Unfortunately, however, the authorsmade no specific report regarding the corpus sub-thalamicum. In a case described by Nikitin (quotedby von Santha, 1932) a relatively large lesion wasencountered in the ventrolateral thalamus. Itsventral border is stated to have extended to within2 mm. of the dorsal border of the corpus Luysi, thelatter structure being evidently uninvolved. Atleast two other instances have been recorded inwhich the apparently responsible lesions weredemonstrated in the thalamus at a sufficient distancefrom the corpus Luysi to vitiate the frequentlyraised objection that the blood supply of the lattermust have been compromised (Malan and Civalleri,1921; Zontoni, 1908).Lesions in the Postcentral Gyrus.-In 1929 Wilson

reported a case in which degenerative, atrophicchanges of the contralateral postcentral gyrus werethe only lesions demonstrable at necropsy. Althoughthis report is unique, it must be retained for con-sideration in view of Wilson's reputation forthoroughness and objectivity.

In brief recapitulation, then, apparently valid casesof hemiballismus have resulted from lesions in theafferent and efferent fibres of the corpus Luysi, theneo- and paleostriatum, the ventral and lateralthalamus and the postcentral gyrus. Although fewin number compared with those in which theresponsible lesion involves the corpus Luysi, such

cases compel reservation concerning the doctrinethat hemiballismus is invariably subtended by alesion of the corpus Luysi.Thus far animal experimentation has shed but

little light on the anatomico-clinical problemsrelated to hemiballismus. Morgan (1927) reportedfailure to produce hyperkinesia in a series of dogs inwhich lesions were placed in the corpus subthala-micum and varying adjacent structures. Thesefindings contrast sharply with those in the humanin whom damage of the corpus subthalamicum bydisease is regularly followed by ballistic movements.Wilson (1914) and Delmas-Marsalet (1925)

reported inability to produce choreic or ballisticmovements in laboratory animals by placingexperimental lesions in the caudate nucleus. Similarnegative evidence from human material was adducedby one of us (Meyers, 1941; 1942a ; 1942b).Extirpation of all available portions of the head ofthe caudate nucleus in nine patients with Parkinson-ism failed to produce hyperkinesia of any variety.In certain instances the rostral third of the putamenand in one instance the oral pole of the globuspallidus were removed in addition to the caudatehead without producing abnormal movements.Browder (1948) in an independent study on thesurgery of Parkinsonism, has reported ablation ofthe superior half of the head of the caudate nucleusin a series of humans without the development ofhyperkinesia.

Mettler and Carpenter (1949) have recentlysucceeded in producing choreo-athetoid movementsin rhesus monkeys by electrocoagulation of thesubthalamus. These investigators are at presentendeavouring to assay the effects of " analytic"surgery in these animals. Their work constitutesthe first promising animal experimentation in thehyperkinetic area.

TherapyIn the endeavour to treat the distressing symptoms

of hemiballismus numerous physical and pharma-cological agents have been employed. As withmost other extrapyramidal disorders, the results ofsuch therapy have been largely disappointing(Meyers, 1942a). The treatment of hemiballismusmay be conveniently discussed under three generalheadings: physical, pharmacological, and surgical.

Physical Measures.-Bertrand and Garcin (1933)described a case in which it became necessary tolash the hemiballistic arm to the patient's side inorder to prevent self-injury. Wilson described asimilar case in which the hyperactive upper limbwas secured to an iron frame in order to prevent thepatient from raining blows upon herself.

117


Pharmacological Measures.-It has been a matterof general experience that sedatives, hypnotics, andnarcotics prove effective in hemiballismus onlv whenadministered in doses sufficiently large to inducesleep. The atropine derivatives have yielded uni-formly indifferent results. Stramonium has beensaid to exert a " somewhat quieting " effect andtobacco is stated to be " useful ", but in no instancehas it been claimed that these agents have abolishedthe hyperkinesia.

FIG. 1.-Diagram illustrating the cortical area reshemiballismus. (Reprinted through the courtthe University of Illinois Press from " The I1944, p. 362, Urbana, Ill.)

Ballistic movements were temporarily arrested inonecase (Papez and others, 1942) by the intramuscularuse of curare. Unfortunately, the dose requiredto achieve this effect (80 units) produced also a

generalized paresis. The unpleasant side-effects ofcurare, erythroidin hydrochloride, and related agents,their short-lived action, and the necessity of often-repeated parenteral administration impose seriouslimitations on their continued use.

Surgical Measures.7-Amputation, peripheral, andcentral operations have been reported, withvarying results.

Peripheral Operations.-Amputation has beenresorted to as a measure of desperation in at leastone case of hemiballismus (Schaller, 1937), and inanother instance paralysis of the brachial plexusby multiple injections of alcohol was employed(Kulenkampff, 1938). In 1931 Jermutowicz re-ported a salutary result following mechanicalstretching of the brachial plexus. A severe paresisof the limb was produced, but the ballistic move-ments were sufficiently damped to permit the

patient to resumework "albeit withdifficulty ".Central Operations.-

Jermutowicz quiteproperly anticipateda return of ballisticmovements in hispatient. It is of con-siderable historicinterest that he pro-posed among otherpossibilities extirpa-tion of the contra-lateral " corticalcenter for theshoulder and arm ".Just such anoperationwas performed in1939 by Bucy (1944)for the relief of leftupper monoballismusin a 21-year-old manwho four years beforehad sustained a severehead injury. Pneumo-e ce phalographyrevealed a dilatedright lateral ventricle.

ected in Bucy's case of left At operation the righttesy of Dr. Paul Bucy and A prto h ihPrecentral Motor Cortex" precentral gyrus was

identified by electricalstimulation and that

portion from which movements of the upperlimb were elicited was extirpated together withportions of the frontal gyri lying rostral thereto(Fig. 1). The anterior wall of the central fissurewas included in the resection.The patient's abnormal movements were abolished

following this procedure. By the end of the firstyear they had returned to a slight degree but at theend of the third year they were described as " mini-mal and usually present only when he is consciousof being observed ". Following operation, con-vulsive seizures supervened for a time. These wereultimately brought under control by barbiturates.


The enduring neurological residua included a markedspastic paresis of upper motor neuron type, hyper-reflexia and dyspraxia of the previously ballisticlimb. Gross movements, however, could be per-formed and the patient was gainfully employed.

Linear Cortico-subcortical Section just Anterior toArea 4: A New Procedure

With the objects of ameliorating hemiballisticmovements and preserving all possible function,the writers recently employed a surgical procedurebased on a tentative theory of the pathogenesis ofhemiballismus.* In essence, the operation consistsof subpial section of the cortex and subcorticalwhite matter immediately rostral to Area 4 in suchfashion as to interrupt the U-fibres coursing betweenAreas 4, 3, 1, and 2 posteriorly, and 4-s and 6anteriorly. Abstracts of two cases treated in thismanner follow.

Case ReportsCase I.-F.B., a 50-year-old woman, was admitted to

the University Hospital on Feb. 14, 1948, complaining ofceaseless agitation of the limbs of the left side. Thedisorder had begun in mid-December, 1947, as irregulartwitchings of the left half of the face. The movementsincreased steadily in frequency and amplitude so that atthe end of the second week the left upper and lowerlimbs exhibited violent and for the most part uncon-trollable movements. Toward the end of the first monthsome slowing of speech and a slight weakness of the leftlower limb were noted. The patient considered hercondition intolerable.The previous medical history was non-contributory,

except that for the past seven years she had been knownto have diabetes mellitus. This condition had been wellcontrolled by insulin and diet.

Examination disclosed an intelligent, affable, whitewoman in good general condition. The blood pressurewas 180/100 mm. Hg. The optic fundi disclosed amoderate degree of diabetic retinopathy. The out-standing finding consisted of vigorous, irregular, andrapidly executed movements of flinging type implicatingthe proximal and distal muscles of the left-sidedappendages. The head jerked irregularly towards theleft, the jaws champed and the left side of the faceexhibited short-lived grimaces. These movements werealmost constantly in play during repose and duringvoluntary activity they often became violent. Theywere absent only in sleep. Quick-succession movementsand the finger-to-nose tests were awkwardly performedon the left side. The right biceps jerk was sluggish, theleft absent. The knee jerks were sluggish but equal, andthe achilles reflexes were absent. The remainder ofthe examination was non-contributory.Laboratory Studies.-The blood counts, hemoglobin

level, blood Wassermann test and radiographs of the skulland chest revealed no remarkable findings. The fasting

*This pathogenetic theory will be made the subject of a separatecommunication.

blood sugar was 100 mg. % and the urine showed a3 plus sugar reaction. The cerebrospinal fluid was clearand under an initial pressure of 125 mm. H2O. Its totalprotein measured 37 mg. %. The cell count was normaland the Wassermann test negative.

Special Studies.-Pre-operative electrographic studiesof the right corpus striatum and neighbouring structures(deep leads) were carried out simultaneously with electro-encephalographic recordings from standard placementsof scalp leads. The technique employed has beenpreviously described (Meyers and others, 1949). The fre-quency, amplitude, wave-form, and polarity character-istics of the patient's recordings were compared withdata similarly derived from 11 control subjects. Theydisclosed no distinguishing features (Figs. 2 and 3).

Operation, March 2, 1948.-A right fronto-parietalbone flap was reflected under local procaine anxsthesiaand the excitable motor cortex identified by means ofbipolar stimulation with a 60 cps. sine-wave. Thethreshold was disclosed at 3V and *25 ma. (approx.).The extent of the motor cortex was mapped out with a3 5V. stimulus and Area 4 was thus envisioned. Towardthe end of this procedure a generalized convulsion wasinadvertently produced and during the ensuing periodof stupor the ballistic movements were absent. When,after some twenty minutes, the patient again becamealert, the hyperkinesia reappeared. A subpial incisionwas then made through the cortex and subcorticalU-fibres along a line just anterior to the envisionedArea 4. It was carried to a depth of approximately2-5 cm. and ran through those portions of the precentralgyrus described by von Bonin as Areas 4a and 4y (Fig. 4).The suppressor area, 4-s, and the premotor cortex,Area 6, were presumed to lie anterior to the line ofincision.During execution of this incision the ballistic move-

ments ceased abruptly and a flaccid, areflexic monoplegiaof the left upper limb was demonstrable. A slight degreeof voluntary control over the proximal muscles of theextremity was retained. The left half of the face and theleft lower limb proved to be only moderately paretic.

Early Postoperative Course.-No hyperkinesia wasobserved at examination eight hours after operation.The left upper limb remained paralyzed but the patientexecuted fairly vigorous movements of the left lowerlimb, both spontaneously and to command. The supra-nuclear left facial paresis remained unaltered. Plantarstimulation evoked a flexor response bilaterally. Theleft upper limb was both analgesic and anxsthetic.The patient's subsequent course was generally satis-

factory. On the fourth postoperative day she suddenlybecame unresponsive for a period of one minute. Theeyes deviated to the left but no tonic or clonic phenomenaappeared. Upon recovery from this episode the patientwas immediately able to engage in conversation. Shewas neither confused nor distressed. A similar pheno-menon occurred on the fifth postoperative day. Afterthe institution of phenobarbital and sodium diphenylhydantoinate therapy there was no recurrence. Theanticonvulsant drugs were continued until the end of thethird month.

119

.S'J''#r4\Ai.A $SOA * -. .,i i P . VtW :

K F -t

A A.. .^

4.,

FIG. 2.-EEG of Case I. The scalp electrodes in the standard positions were connected to reference electrodeson the ears (e). The prevailing alpha wave frequency is 8/s., the amplitude ranging between 25 and 50,uv. Thefrequency, amplitude, wave-form and polarity characteristics of the recordings were similar to those obtainedin 11 normal controls. LF and RF=left and right frontal LP and RP=left and right motor; LO andRO=left and right occipital; RT=right temporal zones.

- C

>- tE .

3 ¢

3.-szO;-

7 -..

RRP - e So

FIG. 3.-EEGs of Case I derived from ring-type pick-up electrodes placed in the right corpus striatum and neigh-bouring structures. Electrode No. 1 was the deepest of a series of eight, and was located just ventral to the headof the right caudate nucleus. l-e, 3-e, 5-e, and 7-e indicate that monopolar recording technique was used,the numbered ring being one of a pair of pick-up leads the second member of which was attached to the rightear (-e). The mean frequency of waves derived from such sources was 8k/s., with a range of 61-11. Theamplitude ranged between 10 and 6O0,v. 1-3, 3-5, and 5-7 indicate that bipolar recording technique was used,the rings being connected to each other as numbered. The mean frequency of waves thus recorded was 11/s.,with a range of 8-16. The amplitude ranged between 5-25,uv. RP-e=scalp electrode placed over right motorarea and connected to a reference electrode on the ear (cf. RP-e, Fig. 2). These electrostriatograms are indis-tinguishable from those obtained in 11 normal controls.


The paresis of the left lower extremity increasedsomewhat during the third, fourth, and fifth postoperativedays, but repeated examinations failed to discloseevidence of spasticity, hyperreflexia, clonus, or Babinski'stoe phenomenon. Early in the second week the paresisof this limb improved perceptibly. The paralysis of theleft upper limb, however, continued unaltered. Thepatient was discharged to her home on March 19, 1948.Later Postoperative Course.-At the end of the third

month the patient was still free of hyperkinesia. Theleft facial paresis had completely receded and thefunctions of the left upper extremity had returnedsufficiently to permit of the cutting offood, tying of shoe-laces, and similar practical skills. The motor power ofthe left upper limb was considered approximately halfnormal and quick-succession movements and the finger-to-nose test could be executed, although awkwardly.The left biceps and triceps jerks were more brisk thanthose on the right but Hoffman's sign was absent. Therewas a slight stereognostic defect. Otherwise, thesensory functions on the left side were unimpaired. Inwalking, the patient exhibited a slight to moderate degreeof circumduction. She could, however, walk on herheels and there was no Babinski sign. A residue of theprevious hyperkinesia was discernible as short-lived,irregular, incoordinate movements of small amplitude,present chiefly when the patient was walking or executing" voluntary " movements with the left upper limb. Inrepose, she exhibited no hyperkinesia.The patient was re-examined at the end of the fifth

month and at intervals of three months thereafter. Todate, improvement has been progressive and she is ingood health, gainfully employed as a housekeeper. Hergait is entirely normal and the left lower limb is entirelyfree ofhyperkinesia. The upper limb exhibits occasionaltwitches of small amplitude, particularly when thepatient is excited. She is able to use both hands forwashing, cooking, ironing, sewing, fastening buttons,tying shoelaces, etc. Dyspraxia is apparent only in suchcomplex skills as crocheting. Stereognosis has returnedto normal. The tendon jerks continue somewhat over-active, but there is no spasticity and the Hoffman signcannot be elicited.Comment.-Without such information as would be

afforded by necropsy study it is manifestly impossible toassert with confidence where the pathological lesionresponsible for this patient's hyperkinesia was located.Irrespective of its locus, however, it seems reasonableto infer that the salutary effect of operation was achievedby interrupting some neural circuit upon the integrityof which the mediation of ballistic movements depended.Case II.-J.S., a 71-year-old housewife was admitted

to the University Hospital on January 19, 1949, com-plaining of severe jerking movements of the limbs ofthe left side of three months' duration. Early inSeptember, 1948, she had experienced unsteadiness and" dizziness " for a week. In mid-October uncontrollablejerking movements appeared, first on the left side of theface and neck and shortly thereafter in the left shoulderand arm. After a few days the movements abatedsomewhat, but they soon returned with renewed vigour,D

and within a month possessed also the left lower limb.During the three weeks before being admitted to hospitalthey became increasingly more vigorous and were almostceaselessly at play during the waking state. The patientcould not voluntarily suppress the movements. Theinvolved muscles felt sore and fatigued but were notpainful. Sedatives were required for sleep.The patient became very irritable and weary, and

during the third and fourth months of her illness lostover 25 pounds. For the ten days preceding her admis-sion she had been unable to walk, even with assistance.Anorexia and nausea were prominent symptoms duringthe week before admission.Past History.-Dyspncea and palpitation had been

present " for many years ", and the patient was knownto have,had arterial hypertension for about a decade.Digitalis therapy had been administered for one andone-half years before the present illness and there hadbeen no symptoms of cardiac failure. She had bornefive healthy children. Except for " nervousness " andintermittent attacks of gall bladder colic over the courseof 25 years, the past history was unremarkable.

Physical Examination.-The patient appeared ill andharassed. The blood pressure measured 220/80 mm.Hg. The retinal arteries were somewhat narrow, andowing to opacities of the ocular media, the optic discmargins appeared hazy. The cranial nerves wereclinically intact. A small nodule was palpable in theleft lobe of the thyroid gland. The heart was slightlyenlarged to the left and exhibited an irregular irregularityof beat. The chest was emphysematous and the radialarteries felt sclerotic.Almost continuous ballistic movements possessed the

left-sided limbs, the head, tongue and left half of the face.Owing to their presence the tendon reflexes of the leftside could not be reliably evaluated; however, theyseemed in general less brisk than those of the oppositeside. In the same manner, the muscle tone of thehyperkinetic limbs seemed diminished. The ballisticmovements (as analyzed by slow-motion cinematography)were manifestly produced by activities of the musclesof the shoulder and hip girdles as well as of the proximaland distal muscles of the extremities themselves. Thestrength of the left grip was approximately half thatexhibited on the right. Tests for gait and coordinationcould not be carried out.

Operation, Jan. 25, 1949.-Under rectal avertin andlight intravenous pentobarbital anmsthesia the motorcortex was outlined and the threshold of that portioncorresponding to the upper extremity determined at2 25 V. and 0 25 ma., using a 60 cps. sine-wave.Stimulation along the anterior margin of the precentralgyrus and the posterior margins of the frontal gyri justanterior thereto with 4 0-50 V. evoked no motor dis-charges but following the application of such stimuli by8-15 seconds the previously ascertained threshold valueof stimulus failed on several trials to evoke its usualresponse when applied to the motor cortex. The sub-sequent return of excitability of the latter at 2-25 Vappeared to indicate that no spontaneous heightening ofits threshold had taken place. Other portions of the

121


exposed cortex failed to show this phenomenon.Although not without reservation, the operators con-sidered that the forward region probably correspondedto the suppressor strip, 4-s. A subpial cortico-subcorticalincision was then carried out as described in Case I (Fig. 4).

4.MOTOR dORTEX4a (FA)4\ 4- s, SUPPR ESSI

- .., 6,, PREMO'

FIG. 4.-Diagram of lateral aspect of right hemisphere shtopographic relations of the line of subpial section cEand II. The motor cortex was identified by electricaldisposition of Brodmann's Area 4 then estimatedThe subpial incision followed a line just anterior to Arethrough 4y and 4a of von Bonin. Area 4-s, the sArea 6, the premotor cortex, are represented rostral t(The pyramidal and parapyramidal corticofugal fibresare represented in their course through the depth ofthe inset, the depth of the incision through the suladjoining and near-neighbouring gyri is illustrated. I

capsule; Crus=right crus cerebri; G.Pre-C. =precent=postcentral gyrus.

Early Postoperative Course.-Ballistic movements were

absent throughout the remainder of the patient's course.

A left hemiplegia of flaccid, areflexic type was demon-strable following recovery from anesthesia. The vitalsigns reached normal levels on the third postoperativeday. At this time a well-patterned Jacksonian seizureof slight to moderate severity and 30-40 seconds' durationoccurred. It began in the left thumb and spread toimplicate the entire left upper limb and the left side ofthe face. Several such episodes occurred on the thirdand fourth days. They then regressed and were not

seen again. The patient was moderately stuporous forthe better part of the first week; after this she becamemore responsive and by the tenth day was fairly rational.Muscle tonus was first detectable by clinical tests in theleft lower limb. The tendon reflexes were at this time

absent on the left side andnormal on the right. The heartrate was 88 per minute and theblood pressure 130/60 mm. Hg.

OR AREAX During the third postoperativeweek the patient at times seemed

rOD, .ORTEX bewildered although most of thetime her speech was rational.On the seventeenth day sheexecuted weak movements withthe left lower extremity andinquired when she might gohome. Although examinationwas considered not completelyreliable, her sensory functionsappeared essentially undisturbed.The left hemiplegia persisted.

Terminal Course.-On thetwenty-sixth postoperative daythe patient's general conditionappeared satisfactory. Whilesitting up eating supper sheaspirated a quantity of milk soup.Dyspnoea and cyanosis super-vened at once. The skin became

C. .PRE-C. cool, mottled, and profusely wet<-/> with perspiration. The heart

rate increased to 140 per minuteand the blood pressure rose to

4 WG/ I' 220/80 mm. Hg. Auscultation(y l,I:{.# revealed no unusual heart

findings, but coarse rales wereheard anteriorly over both sides

iowing the presumed of the chest. Speech was slowarried out in Cases I and measured. The patientstimulation, and the remained conscious throughoutby visual inspection. and followed verbal directions,a 4, probably cutting well. A moderate quantity of;uppressor strip, and thin mucoid q yo the line of incision thm bchsomaterial was aspir-arising from Area 4 ated by bronchoscopy from thethe hemisphere. In trachea and left upper andbcortical U-fibres of lower bronchi. Oxygen therapyInt.C. =right internal was instituted. The acute episode,ral gyrus; G.Post-C. subsided within 45 minutes.

Electrocardiography failed toprovide evidence of a coronary

occlusion or pulmonary infarct. A few hours laterfine and coarse rales were audible over the left lowerlobe. Despite the prophylactic administration ofpenicillin, pneumonia developed and the patient diedthree days later, on the twenty-ninth postoperative day.

Necropsy.-An acute bronchopneumonic consolidationwas revealed throughout the right lung. There wasevidence of a generalized arteriosclerosis. All othersignificant post-mortem findings were confined to thebrain. Some residual subpial hvmorrhages and localized


swelling were noted in the gyri adjacent to the line ofcortico-subcortical incision. The i-nterface of theincision was well agglutinated. In depth, it closelyapproximated 2-5 cm. throughout its course. Theoperative site was not otherwise remarkable.

microglia were present but on the whole the glial elements,like the neurocytic, appeared devitalized (Fig. 7, a and b).

All portions of the right corpus subthalamicumexhibited well-preserved ganglion cells, normal glialelements, and granular ground substance (Fig. 8). The

ansa and fasciculus lenticularesdisclosed moderate degrees: of

-' j'¾ demyelination traceable into theventromedial nucleus of thethalamus, the reticular formationof the anterior midbrain and the

4%_\ substantia nigra....

%. _.

\ ...\ }

-4:s i: :.

b. :F s

F: #l. ..

:;

F w:

.A_,S,,f._$rt'.

F ....;...,.:;,.,..i

{.. .,% ,,|.stv _ e,.s

o.lit

'Fr

FIG. 5.-Gross coronal section of brain of Case II at level ofmammillary bodies(seen from behind), showing encephalomalacic shrinking and obscurationof structures in and neighbouring upon the right lenticular nucleus, externalcapsule, and claustrum. The corpora Luysi appear normal.

The brain was suspended in formalin for a week. suggests that the inteCoronal sections were then cut at intervals of approxi- a particular structuimately 0-5 cm. The ventricular system was moderately centripetal and centidilated. The gross sections revealed only one lesion: to suppress hemiballia brownish-purple, granular-appearing discolorationwithin an obviously shrunken lenticular nucleus (Fig.5).This finding was most intense in those portions of the Bucy's case and tputamen and both crura of the globus pallidus strate that by theappearing in the coronal section through the mid-portion of the mammillary bodies. The internaland external medullary lamine were almost wipedout. The external capsule and claustrum werelikewise obscured and the internal capsule appearedto be but slightly involved. More rostrally andcaudally, the lenticular nucleus exhibited less intensepathological changes. The corpus subthalamicumexhibited no gross changes in colour, size, shape,or texture (Fig. 6).

Microscopic study of serial sections alternatingthrough Nissl, Weil, and Masson stains confirmedthe gross pathological findings. The right putamenand globus pallidus were the seats of well-establishedencephalomalacic changes. Their neurocytes werein various stages of disintegration. The less severelyaffected cells exhibited chromatolysis, swelling, andnuclei in eccentric positions. The more severelyaffected cells appeared as mere smudges. Manycells had evidently completely disappeared. No in- FIG. 6.-Gross coronalflammatory cells or vascular proliferative processes in front at level juwere discernible. The heterogeneous appearance of motor nerves show]the ground substance betrayed its general dis- and substantia nigrorganization. A few gemistocytic astrocytes and damaged right lentil

Comment.-The absence ofpathological changes in thecorpus subthalamicum and thecircumscribed character of thepathological findings in thelenticular nucleus opposite theside of the hyperkinesia rendersthis a critical case. It may ofcourse be denied that this was agenuine example of hemiballis-mus; if so, criteria other thanthose set forth above are in needof definition. If, however, thecase be accepted as one of hemi-ballismus, it follows that lesionslying outside the corpus Luysimay subtend the disorder, acircumstance which in its turn

Zgrity of a neural circuit, rather thanre (e.g., the corpus Luysi and its:rifugal fibres) is normally requiredismus.

Discussionthe two cases cited above demon-use of relatively simple surgical

section of brain of Case II seen fromist anterior to emergence of oculo-ring apparently intact corpora Luysira. Arrow points to the side of theicular nucleus.

123

.::.....

------------


A

0.. :, 16W

.. ;i~~.,~~~~~~~~~~~~~'. I?f ~ ~~~~~~~~~~~~,-l*# b:~~~~~~~~~~~~~~~~.

9

4

a 9.

J X

-W.

4 .E4ml 4 '.

*~~~~~~~

.'aI%1 4*b4-:

9.s *4 9

*9

44.~~ ~ ~ ..4* <

449. i4 9

. 4

V

FIG. 7.-Photomicrographsof putamen (A) andcrus II ofglobus pallidus(B). The neurocyticand glial elements ex-hibit various stages ofdevitalization in con-sequence of encephalo-malacia presumablysecondary to athero-sclerosis. A con-siderable outfall ofneurocytes is apparentand the ground sub-stance of the globuspallidus reveals patchesof early coagulationnecrosis. (Nissl stain.)

or' 0:

R

procedures ballistic move-ments can be promptlyarrested, or so signifi-cantly suppressed as toafford gratifying relief and.thus avert exhaustion. In

view of the refractorinessof hemiballismus to con-servative therapy, the e.negligible prospect ofspontaneous recovery, theseverity of the disability,and the frequency withwhich the illness endsfatally, it seems desirableto recommend earlysurgical intervention.

It must be acknowledgedthat by no means all oreven a large majority ofcases of hemiballismuscan be saved by surgicalintervention. Pathologicalinquiry clearly reveals that in some instances theresponsible lesions are progressive in nature, comingultimately to involve those mesencephalic anddiencephalic structures upon which life itselfdepends. In yet other cases multiple lesionsdevelop. Though surgery appears to offer mostpromise in those cases where there is a well-circum-scribed, non-progressive lesion, it seems reasonable

J 0

P~~~~~~~~W_ Y4

... k.,Gk . 'k :: 1t :A

* .-O.' 4:.:

4Jv . .l

I

w :s-

r

to urge operation for all cases on the ground thatthis provides the only effective therapy at presentavailable.

Summary1. The excessive, almost ceaseless movements of

hemiballismus commonly lead to exhaustion,cardiac failure, and/or pneumonia. The disorder

X Ss:.

* > X \ S

?. :

*.] qs ..... t S . @

w

fs b #

ag glg | 11?

ak !v. et a.f jja sz 4,

¢ :.: .: . # :st:

x J:* # -.e. .. ;, - * S JS ::

:. % ro.

.499 '.4

944

.4.94

4. .. '9

.Isom iml mr.. Kum' 'As

: qI*o '11

W4, .I

u

A

1. 0 .1.

..A..

v

T..:bi 1

t.4:

J,

-P

- -low .,-s 4

..' ik5t, I


I * ' * t*9* ,~ i * 0* 9 ***

E~~~ ~ 9q L .,, . 9s 'o *;X & '* *"X

skt''9*0-^# ' -' * 's ';

* 999m9' ''S 8 t* , .. - ' ' , .S'

4,~~ ~ ~ ~ ~ ~ *

a*oo ' i i #** * t = ' * _

d.~~~~~~~~~~~~~- 9..~S

AS~~ ~ ~ ~

lb* ~ 9 4

4~~P~ 4' 9,, '*~70 Is9'p~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4

;NL~ ~ ' , 4 . 9' *

s

44~~~~v.i

%.

1% _'.I9

thus complicated almost always terminates fatallywithin a few weeks to several months of onset.

2. The lesions responsible for hemiballismus maybe located within the contralateral neostriatum,paleostriatum, thalamus or postcentral gyrus as

well as in the corpus subthalamicum and/or itsafferent and efferent connexions.

3. Spontaneous cessation of hemiballismus israre and treatment by physical measures, drugs or

peripheral surgical procedures has proved to bedisappointing. On the other hand, "central"neurosurgical procedures have afforded relief in afew instances. Two cases subjected to linearcortico-subcortical section just anterior to Area 4are reported. This procedure appears to produceless neurological deficit than does cortical ablation.

4. The " central " neurosurgical procedures (cor-tical extirpation, linear cortico-subcortical section,and midbrain pyramidotomy) constitute the onlydefinitive treatment upon which reliance may beplaced for arresting ballistic movements. In viewof the grave prognosis of the disorder, surgeryshould be regarded as a semi-emergency measure.

REFERENCES

Austregesilo, A., and Borges-Fortes, A. (1937). Rev.neurol., 67, 477.

and Gallotti, 0. (1924). Ibid., 1, 41.Bertrand. I., and Garcin, R. (1933). ibid., 2, 820.Bianchi, L. (1909). Ann. Nevrol., 27, 1.Bonhoeffer, K. (1897). Mschr. Psychiat. Neurol., 1, 6.

(1930). Ibid., 77, 127.Browder, J. (1948). Amer. J. Surg., 75, 264.

*r* St < &

.i

I, ~.. --7

,b A *& 5o

Bucy, P. C. (1944). "The Precentral Motor Cortex"(case 2). Pp. 361 and 404. Univ. Illinois Press.

Civalleri, A. (1921). Policlinico, sez. med., 28, 242.Davison, C., and Goodhart, S. P. (1938). Arch. Neur.

Psychiat., Chicago, 39, 939.Delmas-Marsalet, P. (1925). " L'etude des fonctionsdu noyau caude." These de Bordeaux, p. 159.

and Pachon, V. (1924). C. R. Soc. Biol., Paris,91, 558.

Fischer, 0. (1911). Z. ges. Neurol. Psychiat, 7, 463.Fragnito, O., and Scarpini, V. (1926). Riv. pat. nerv.

ment., 31, 524.Greiff(1883). Quoted by Lea-Plaza, H., and Uiberall, E.

(1945).Grinker, R. R., and Bucy, P. C. (1949). " Neurology"

(Ed. 4), p. 315. C. C. Thomas, Springfield, Ill.

Hayne, R., Meyers, R., and Knott, J. R. (1949).J. Neurophysiol., 12, 185.

Jakob, A. (1923). " Die extrapyramidalen Erkran-kungen." Berlin, J. Springer, p. 183.

Jermutowicz, W. (1931). Rev. neurol., 55, 374.Kelman, H. (1945). J. nerv. ment. Dis., 101, 363.Kulenkampff, D. (1938). Zbl. Chir., 65, 2466.Lea-Plaza, H., and Uiberall, E. (1945). Rev. meid. Chile,

73, 938.Lewandowsky, M., and Stadelmann, E. (1912). Z. ges.

Neurol. Psychiat., 12, 530.Marcus, H., and Sjogren, H. (1938). Rev. neurol., 70, 1.Martin, J. P. (1927). Brain, 50, 637.

and Alcock, N. S. (1934). Ibid., 57, 504.Matzdorff, P. (1927). Z. ges. Neurol. Psychiat., 109, 538.Mettler, F. A., and Carpenter, M. B. (1949). " The

modification of subthalamic hyperkinesia in primates"(to be published in Trans. Amer. neur. Assn.).

Meyers, R. (1941). Res. Publ. Ass. Nerv. Ment. Dis.,21, 602.

(1942a). N. Y. St. J. Med., 42, 535.(1942b). Ibid., 42, 317., Hayne, R., and Knott, J. (1949). Journal of

Neurology, Neurosurgery and Psychiatry, 12, 111.

125

FIG. 8.-Photomicro-graph of right corpusLuysi showing theintegrity of its neuro-cytic and glial com-ponents. All partsof the nucleus wereexamined and noneshowed evidence ofpathological change.(Nissl stain.)

1:


Moersch, F. P., and Kernohan, J. W. (1939). Arch. Vogt, C., and Vogt, 0. (1920). J. Psychol. Neurol.,NeuroL. Psychiat., Chicago, 41, 365. Lpz., 25, 627.

Morgan, L. 0. (1927). J. Comp. Neurol., 44, 379. von SAntha, K. (1932). Z. ges. Neurol. Psychiat., 141,Papez., J. W., Bennett, A. E., and Cash, P. T. (1942). 321.

Arch. Neurol. Psychiat., Chicago, 47, 667. Whittier, J. R. (1947). Arch. Neurol. Psychiat., Chicago,Schaller, W. F. (1937). Ibid., 37, 983. 58, 672.Schob, F. (1920). Dtsch. Z., Nervenheilk., 65, 210. Wilson, S. A. K. (1914). Brain, 36, 427.Thurel, R., and Grenier, J. (1947). Rev. neurol., 79, 502. (1929). Dtsch. Z. Nervenheilk, 108, 4.Touche, M. (1901). Ibid., 9, 1080. Zonnoni, U. (1908). Gazz. Osped., 29(2), 1156.

Documents

HEMIBALLISMUS: /ETIOLOGY AND SURGICAL TREATMENT