Surgical treatment of epilepsy

SURGICAL

TREATMENT

OF EPILEPSY *

A. EARL WALKER,? M.D. AND HERBERT C. JOHNSON,~ M.D. Baltimore,

Maryland

M

AN‘L’ different pathologic condiwith tions may be associated convulsive seizures, so that an accurate diagnosis must be made in every case before any type of therapy, medical or surgical, can be rationally prescribed for the attacks. The etiologic factor may be suspected from the history- but often cannot be demonstrated even after compIete physical, electroencephaloneurologic, graphic, pneumoencephalographic and arteriographic examinations. In childhood, idiopathic epilepsy associated with a cerebral dysrhythmia is common but in adult life a convulsion is much more Iikely to be a symptom of a foca1 brain lesion.63 Even if al1 the examinations mentioned fail to disclose a cause for the attack, the patient must still be suspected of harboring a small cerebra1 lesion which may later give more definite evidence of its presence. With the exception of t.hose cases of brain tumor, which require immediate surgery, the therapy of both idiopathic and symptomatic epilepsy is primariIy medical. Only when anticonvulsive drugs in doses approaching the toxic levels have faiIed to control the attacks is surgica1 intervention justified. IDIOPATHIC EPILEPSY Trephination. Historical writers have produced evidence that the earliest operation for the relief of epilepsy was simple trephination. Numerous trephined skulls have been found in the ruins of ancient civilizations. This procedure was presumably done to reIease deviIs considered to be irritating the brain. AIthough the demonological conception has passed, the use of trephination in the treatment of epilepsy has not been abandoned entireIy. Cutter’ * From the Division of Neurological j Now at Johns Hopkins Hospital,

discusses five cases reported by Dudley in I 828 in which trephination and decompression was done. But in general, neurosurgeons concur with Penfield3’ who states that the resuIts of subtemporal decompression do not justify the procedure in essential epilepsy. Pneumoencephalograph~. Therapeutic effects of pneumoencephalography have been repeatedIy reported since the introduction of this technic. In idiopathic epilepsy the procedure has not changed the course of the disease. 60 It is true that some patients have fewer attacks after the air injection but in most cases this improvement may be attributed to better medical management of the case. Thus Penfield”i reports five of ninety-six patients to be seizure-free for an average of 2.8 years after pneumoencephalography. These patients were under sixteen years of age and had had attacks Iess than four years. Freeing

of Corticodural

Adhesions.

Ney:’

ascribes some cases of idiopathic epilepsy to local cortical tension due to traction by corticodural attachments at the vertex. Two hundred twenty-five patients with idiopathic epiIepsy were operated upon and the adhesions relieved. The results were said to be good but the follow-up is not adequate to evamate this procedure. Scar-IV believes that there is a specific epileptic syndrome caused by anomalous pacchionian granulations extending into the motor area. Division of these granulations thus freeing the corticodural adhesions relieves many of the patients to some extent. Miscellaneous Procedures. On the assumption that the sympathetic supply to the cerebral vessels is unstable in idiopathic epilepsy, sympathectomy has been performed. However, Penfield states, “Sym-

Surgery, Baltimore,

200

the University Md.

of Chicago,

Chicago,

III.

\

01

I .S .XV,

NC,. ,

W’alker, Johnson--Epilepsy

pathectomy, howe\-er complete, does not make seizures impossible and is hardly justifiable as a treatment for epilepsy except, in the presence of unusual e\-idence of abnormality of the s?;mpathetic nerl’ous s\.-stem.“‘li Removal of the carotid beds and denervation of the carotid sinus have also been used in the treatment of epilepsy but Penfield:li reports that the procedure IS of no \Talue unless specific abnormal hyperactivity of the carotid sinus exists. With the purpose of pre\renting the con\,ulsive seizure becoming generahzed and thus preventing loss of consciousness in epileptics having a uniIatera1 beginning of their attack, Van Wagenen”; divided the corpus callosum. Six of ten patients so treated have had no generaIized seizures since operation. Van Wagenen”” has aIso made a hypothaIamic puncture at the upper end of the pituitary staIk to induce dehydration by diabetes msipidus. In one case no convulsions had occurred at the time of the report. In psJ-chomotor epilepsy-, negative spike foci ha\re been found at the tip of the right or left or both tempora1 lobes.” AbIation of the tip of the tempora1 lobe and the focus has been carried out but the results are not yet kno\vn. SYMPTO\IATIC

EPILEPSl

Manv disorders of the brain are associated \
American

Journal

of !hur~:r,~ \

20

I

In cerebral has been 2g,35*41 List gIioma to tive of the

neopIasms and scars this factor studied in some detaiI.2.‘“.1”‘2”’ considers the histoIogic type of pIay an important rGIe irrespecIocation of the tumor.‘!’ II. Severit_y ?f Cortical Damage. Factors which tend to cause extensive cortical damage have been correIated with a high incidence of epilepsy. Thus penetratink dural wounds, infected wounds and extensive wounds are associated with con\rulsions about twice as often as a general, unselected group of head injuries. Retained foreign bodies, however, appear to have no reIationship to the development ot con~uIsions.“~~,l1.1”.~0..“.“8~“!’~fil III. Heredity. Is there an inherent predisposition to convuIsions? The problem has been more adequatery studied in posttraumatic epilepsy than other types of symptomatic epilepsy. Cobb gilyes statistics bearing on this question.” In 250 norma1 control subjects epiIepsy occurred in 2.6 per thousand of 1,896 parents, siblings In 1,086 non-traumatic and offsprings. epilepsy occurred in 21 per epiIeptics, thousand of 9,139 reIatives. In 235 patients with epiIepsy who had suffered head trauma, there were 14 per thousand epileptic reIatives. This figure is considerabI\higher than that occurring in the group ;I norms1 controls. The figures obtained in the group of post-traumatic epileptic patients studied at Cushing General Hospital differ from those of Cobb.” In this series onIy 4.3 per cent of the patient’s reIatives had convuIsions compared to an inciclence of 3.4 per cent in normal families and 1’ per cent in epiIeptic famiIies.?’ Ziskind6” doubts that there is an inherited predisposition to epilepsy since it would have to be present in from 20 to 40 per cent of patients with head injuries and tumors. The epileptogenic factors are obviously not yet compIeteIy understood. T_ype of Seizure in Symptomatic Epilepsy. There is nothing peculiar in the seizure ot symptomatic epiIepsy to distinguish it from that associated with idiopathic convuIsions. In a series of cases, focal attacks are

much more common in symptomatic than in the idiopathic epilepsy.“~‘“~“~“’ The mode of onset of the seizure is usually an indication of the site of the lesion within the brain. Thus lesions in or near the motor area usually induce attacks beginning in the contralateral arm or hand, Iess commonly in the face and rarely in the leg. Sensorv beginnings of the attack indicate the position of the lesion in the parietal, temporal or occipita1 lobes. The aura or onset of the attack is usually an excitatory phenomenon but paralytic manifestations are seen. Thus the most frequent aurae are twitching of the face, eyes, fingers, etc., but occasionally the onset is indicated by an inability to talk or a weakness or heaviness of the arm. Similarly in the sensory sphere, the aura may be excitatory or paralytic. For example, \;isual aurae may consist of a blurring of vision in one or both visual fields. This is frequently described as a black ball, dots, circular curtain moving across the visual field in an unduIating or pulsating fashion. Frequently this scotoma begins in the periphery of a hemianoptic sector, proceeds across the vertical meridian to involve the norma field causing a compIete loss of vision for a few minutes, With the onset of blindness a generarized attack sometimes develops. The visual aura may also consist of white or colored “lights” in the affected visua1 fields. When IateraIized the hallucination occurs in the visual field contraIatera1 to the cerebra1 wound. Vertiginous aura manifested bv dizziness just before a generalized attack‘ is not an uncommon complaint and in some cases a true vertigo is present. Auditory aura and epigastric aura may occur at times. The epigastric aura of a feeling of uneasiness in the pit of the stomach, nausea, a choking sensation, cramping abdominal discomfort and vomiting are not so frequently encountered as
stitute the entire attack or they may progress to a generalized convulsive seizure. A true Jacksonian seizure is not often seen in symptomatic epilepsy nor are petit ma1 attacks as common as in the idiopathic

seizures.l~,“7,“1,““.fil

ELECTKOENCEPHALOGKAPIIIC SYMPTOMATIC

FINDINGS

IN

EPILEPSY

The electroencephalogram usually distinguishes the symptomatic from idiopathic epilepsies. It is based upon evidence of foca1 brain damage such as slow delta waves with superimposed sharp, spiky waves. This question has been most extensively studied in the post-traumatic epilepsies which will be discussed in some detail. In a study of eighty-one cases of posttraumatic erlilepsy Jasper and PenfieldZR found that Ain 90 per cent the electroencephalogram revealed a local area in one hemisphere from which random spikes or sharp waves were most prominent, usually upon a background of random delta waves. Slow waves aIone, they believed, were not a reliable guide and positive electroencephalographic evidence of a focal Iesion was present only when spikes or sharp waves were repeatedl)recorded from the same discrete area. In an earl\- report LZ’ilIiams states that the character of the changes seen in the electroencephalographic record in trnumatic epilepsv appear to be the same as in idiopathic epilepsy. 64 In a later report, however, he states that focal abnormality is found in a large proportion of cases of traumatic epileps)! in contrast to the group of idiopathic epileptics.“: An abnormal electroencephalogram persisting after a head injury does not necessarily increase the IikeIihood of traumatic epilepsy; but if paroxysmal outbursts are present, the likelihood is increased by four times. If a patient’s electroencephalogram shows larvaI epileptic outbursts, epilepsy is almost certain to supervene. Gibbs, Wegner and GibbslG found that focal electroencephalographic abnormality is four times as

c~~ninion in post-traumatic epileptics as in unselected epileptics and that the foc:~l electroencephalographic abnormalities correlate with focal seizures. Focal seiz.urr--discharges are twenty-one times ~1s common among post-traumatic epileptics as among head injur! patients without seizures. Roseman and Woodha I I ‘!I reported paroxvsmal electroencephalogrnphic changes ‘in fifteen cases of posttraumatic epilepsy; in twelve cases with focal seizures localized electroencephalographic discharges were present. In four of the fifteen patients with convulsions abnormal electroencephaIographic foci were recognized f;rst. HoM;ever, there were another fifteen cases in the series with similar paros~smal abnormal electroencephalopraphic changes but without clinical c>on\-&ions. liaufman and Walker” studied the electrc,encephnlo~ran~s of 24 I cases of post-traumatic epilepsy and of eight>three cases of severe head injur?; but without seizures. The records were abnormal i II 9 I .; per cent of the epileptic and 7;. I per cent of‘ the non-epileptic group. A focal :tbnormnlit,v was present in 83.8 per cent ot‘ the epileptic cases and in 67.5 per cent of the control group. The most common focal abnormality- in both groups \~as the presence of slon wa\ves, but this ~~3s much more frequent in the epileptic t-3 per cent I than in the non-epileptic group (48.2 per abnormalities consistcent I. Paroxh,smnl ing of sharp waves or spikes \vhich are u5unll,~ considered indicative of epileps?, occurred about equally (approximate]) 20 per cent) in the epiIeptic and control srries. (Fig. I.‘) The epileptic character of a focus ma! bc brought out by metrazol :tcti\xtion”:’ 01. in sleep records.‘” Following a rapid intravenous injection of 200 tng. (2 cc. i 01‘ metrazol, within thirty to sixty seconds focal epileptic: abnormalities usualIy apThis pear in the electroencephalogram. focal discharge consists of slou- wax’es or humps hn\.ing :I frequency less than the aIph:l rhythm or single or multiple

spikes, at times occurring I-h) thn~icalI~, giving the pattern of a localized electroencephalographic seizure. (Figs. 2 and ;. 1 This type of focal discharge occurY itI about 67 per cent of a series of cxse’s 01 post-traumatic epilepsy. If the procedurca is repeated on another occasion, :IpproximateI,v one-half of the cases ha\ ing no change on the lirst examination \vi[l ha\ c’ electroencephslogrnphic alterations on the second trial. Abnormalities present Ijet’orc activation are usunlI,v aggra\-ated b\ thr metrazol. In about 88 per cent of tht* positive cases the focal cle~trot~nc,eph~tlographic alterations induced 11,~ metrazol are paroxysmal. In IO per cent of c’ases generalized electroencephalographic alterntions consisting of single or multiple SIOM wxves or spikes are present simultaneousl\ in tracings made from se\,eral parts 01‘ both sides of the head, giving :I pattern of a generalized electroenceph:Llr)~r~~f:,hi(, seizure. If the metrazol activation is made aftel the patient has had no nntic.onvulsarit medication for severa davs, c.linical seizures similar to the patient’s usual attacl\ will occur in about 13 per cent of the cases. Clinica seizures have not occurred Lvhen anticonvulsant medication \v;ts not interrupted, but in a small percentage of cases the patient may experience ;I srnv~r’,~ 01 motor aura. The epileptic manifestations III:I>. lx brought out by spontaneous or induced sleep while the patient is ha\,& an clewtroencephalographic record made.”

C0nvulsix.e attacks are ;I comnlon s>-mptom of brain tumor. Sat-gent”’ reported an incidence of con\xtIsions of 30.9 pet- cent in ;I series of’ 270 patients. Parker”” in :I series of ;I 3 p atients with celcbral tumor found that sixty-seven (2 I .1 per c’ent I had major epileptic con\*uIsions. Dowman and Smith”’ found that seizures occurred in 39 per cent of I 00 cases of intracranial vowth. In a statistical sur\72\- of Bailey’s ?of Chicago c:lSW . L from the Cinivcrsit!

Walker, Johnson--Epilepsy

T-IO’-

c\Y

~--~~~~~h~-~n-~-~Y1^_~~--~-~~~~~~~~,__.

FIG. 1. EIectroencephalogram of a patient suffering from post-traumatic epilepsy. The sIow wave focus at the site of the small crania1 wound just above the ear is evident. The Ieads arc indicated on the traces. The horizontal Iine at the base rcprescnts an interval of one second and the vertical line a calibration of microvolts.

Fro. .z. Records showing the effect of metrnzo1 activation in post-traumatic epiIepsy. In the upper record the spiking at point 14 is apparent. In the lower record the marked slowing of the cortical rhythm with particular irrcgularitirs in the second trace is evident. An interval of one second is indicated by the horizontal line at the base.

\o,

I.XNV,

No

Walker,

I

Johnson-EpiIepsy

lx?- and M’aIkeF found that 25.7 per cent of the patients suffering from tumor of the brain had generalized convulsions. Penfield et al.‘” report an incidence of 51 per cent in intracerebral and 62 per cent in cxtracerebrnl neoplasms.

Amrrican

Jor~rnal 01 !hirp.r.b

20 5

the greater is the Iikehhood of conv&iv-c seizures developing.“g*4’ Gibbs’j and Groff’” have found that a tumor of the temporal, parieta1 or fronta Iobe products seizures more frequentIy than one in another region. Infratentorial and pituitary tumor3

12*-4

A---

4- 14 14-10 . IO - 12

~b++++w@WN,“x.,” 570

mgm.

met razol

i.m. (6 mgm.1

,-*-W,U.-~,-~~-“. Starting T

mti

**VW+

kg.)

B<

Inh,n#fl 4lOsec.

after

metrazol

CC

450sec.after metrazol FIG. 3. Electroencephalogram to show the activating

effect of intramuscular mrtrazol. The leads are as indicated. The spiking from point IO is well shown in tht lower two sets of tracings. An interval of one second is indicated by tflc horizontal line at the base.

The incidence of convulsive attacks in patients with cerebra1 tumor varies with the t) pe and location of the neoplasm. The nearer the tumor is to the central sulcus

rarely are associated with con\xlsions:” Tumors with a long life history arc likel! to be associated with convulsive seizures. In Penfield’s series of cases” 3:’ per cent

206

A mericnnJourd

of Surgery

Walker,

.IANUAHY, 10.$x

Johnson-EpiIepsy cases

patient

medication is IN)L given has developed attacks.

EPILEPSY

ASSOCIATED

WITH

until

t hc

CONGENITAI.

LESIONS

FI(;. 1. Angionxc of left inferior ing convulsive scizurrs.

frontal

region

induc-

the glioblastomas had seizures, 70 per cent of the astrocytomns, 68 per cent of the meningeal tumors and 92 per cent of the oligodendrogliomas. List found a simibetween epilepsy and lar relationship histologic type of tumor.‘” RemovsI of the tumor frequently abolishes the attacks. This result may be enhanced bv care at the time of craniotomy that no de&alized or macerated cerebral cortex is left at the site of the tumor. Penlield et al. ” report that of twenty--nine epileptic patients with meningiomas, nineteen continued to have seizures after removal of the tumor. In a series of twentyone non-epileptic patients with meningiomas nine had developed seizures after operation. In a group of twenty-seven epileptic patients with astrocytoma, twentythree continued to have seizures; of six seizures, patients without preoperativre three later developed seizures. a patient who Following operation, has had convulsive attacks preoperatively should continue to take anticonvuIsant medication. If he remains attack-free for a period of two to three y-ears, the anticonvulsant medication can then be gradually discontinued. The question arises as to whether al1 patients with brain tumor should receive postoperative anticonvulsant medication since a certain percentage of them will develop seizures. This is a matter of individual choice but in most

Of

Congenital lesions of the cerebrum are numerous in type and degree. Epileps? is an infrequent symptom of certain of the congenital lesions although in the Iocalized aplasias and especially the vascular abnormalities it is particularly common. (Fig 3.) UsuaIIy incIuded in the congenital abnormahties are cerebral Iesions incident to delivery such as cerebra1 hemorrhage, contusion or laceration and vascular occlusion. These injuries cause the formation of atrophic gyri and cysts. The treatment of epileptic seizures secondary to congenital lesions is primarily medical. With the use of appropriate doses of anticonvulsant medication many of the cases can be controlled. Roentgen therapy is of value in some cases of vascuIar abnormality. The indications for surgery in the treatment of epiIeptogenic congenita1 lesions are few even in many of the cases not controlled Frequently the damage by medication. to the cerebrum is diffuse and widespread so that little can be accomplished by surgical intervention. Surgical therapy is indicated in those cases not controlled by adequate anticonvulsant medication in which there is neurologic, roentgenoIogic or electroencephalographic evidence of a focal lesion. The congenita1 vascular abnormalities constitute a specia1 problem. The former high mortality involved in their surgical treatment has been reduced and by present method s of hemostasis adequate blood repIacement.“6 EPILEPSY

ASSOCIATED

\VlTH

INFLAMMATORY

DISEASES

There are a number of types of inflammatory disease of the centra1 nervous system which may result in convulsive seizures. The most common of these are the pyogenic infections. This includes the acute pyogenic meningitides, chronic arach-

noiditis secondar! to a meningitis and pyogenic brain abscess. Seizures occurring in an acute meningitis are, like tever and delirium, onI! temporary symptoms of the disease. Only rarely will a chronic con\ ulsivr state develop secondary to pyogenie meningitic involvement. Seizures are common symptoms of an acute brain abscess and not infrequently continue ii-hen the abscess has been aspirated, drained or ablated. The convulsions ma) begin a number of years after drainage of the 3~~SCeSS~l.‘i,21,:‘:i Occasionally a chronic encapsulated abscess which vas ne\,er surgically treated may remain for years unrecognized with symptomatic consulsi\‘e seizures. (Fig. 5.) The convulsive seizures of neurosyphilis ;cre IveIl known. Tuberculomas are frequently diagnosed as cerebral tumors and ma? be associated with convulsive seizures. The yeasts and fungi may invade the central nervous system producing a chronic meningitis with occasional epileptic attacks. An amebic abscess is sometimes found within the cerebrum. Certain parasitic infections such as echinococcus, cj-sticercus and the shistosomes involve the cerebrum vith con\uIsivc seizures as a common symptom. These infections, becoming more common in this country, may be diagnosed in some cases by calcification in the multiple lesions.y,!’ The treatment of the epileptic seizures in these various inflammator> diseases consists of specific therapy directed against the causative organism and anticonvulsive therapy. If medication does not control the seiZLlTt'S, ;I surgical exploration and ablation of the epileptogenic zone is indicated. EI’IL EPS’I’

ASSOCIATED

WITH

CEKEBKAL

SCAKS

Following it cerebral wound, a firm
tion for post-traumatic epilepsy this subdural membrane is found firmly adherent to the dura mater. It may be 2 to IO mm. in thickness and quite extensive o\,erIying as much as one-half of a hemisphere in some cases. The brain adjoining tht meningocerebral scar in some cases appears normal, but frequentI> it is yellow or brown, softened and cystic. The arachnoid is often thickened and opaque, occasionally containing white cahified plaques. Because of atrophic g)‘ri, arteries and \,eins usually concealed in sulci ma\ present on the surface of’ the brain.;’ (Figs. 6 and -.I Histologically a meningocerebral cicatris consists of connective tissue \vith collagen libers and blood vessels. The cicatrix ma! contain neuroglia cells of the piloid variet?but usuallv no ganglion cells.““~4” Penlield has described the anastomosis of the vascular channels in the connective tissue with the vase-astral network of the brain. Because of this anastomosis, as the cicatris contracts there is a resultant pull and traction on t.he brain.R” Between the scan and normal gray matter is an intermediate zone of partially destroyed gray matter in which there is evidence of progressive neuronal destruction in small patches surrounding blood vessels. The presence of compound granular corpuscles and changes in the neuroglia are said to bc evidence of the progressive nature of the This histologic picture is the lesion.4:’ same whether the patient does or does not develop epilepsy.a~‘fi~4’1 The distortion of the brain clue to the cerebral scarring is well illustrated in pneumoencephalograms. The most common pneumoencephalogrnphic linding in the cases of penetrating head injuries is ;L bilateral ventricular en1argemen.t with an outpouching at the site of the skull defect. In a series of 109 patients \vith postwho had pneumoentraumatic epileps) cephalograms made from three months to three years after their head injury, this type of pneumoencephalogrnphic appearance was present in sixt>--three cases. In

208

A ,ncllCnn ..

Jourd

or Surgery

Walker,

Johnson-Epilepsy

JANUARY, 190

FIG. 3. A calcified capsuIe of a chronic :tlxccss srcondary to otitis media. A, fatcral roentgenogram of the skull; H, cstirpntetl ;I~WX~S.

twelve cases there was unilateral ventricular enhrrgement with an outpouching and in eight cases a simple unilateral ventricular dihrtation on the side of the Iesion.61 (Fig. 8.) The subarachnoid space, usuaIIy on the side of the lesion, is poorIy demonstrated in many cases, but occasionally cysts or ditated s&i are visualized. The pneumoencephaIograms of patients suffer-

ing from closed head injuries were either normal or had symmetrical ventricular dilatation. Of the entire group of 109 patients, only fourteen pneumoencephalograms were reported as normaI. These VentricuIographic changes are not peculiar to cases of post-traumatic epilepsy. Practicahy the same findings were present in a non-seIected group of head injuries re-

ported by Troland et al.“” although the incidence of severe ventricular distortion was higher in the epileptic series than in the unselected one. Convulsive attacks Ina) occur immediateI!, after a head injury or they ma!; make their appearance rnan!~ years later. For convenience most authors divide post-traumatic epiIepq into immediate (attacks in the first twelve to twent\--four hours), delayed, (attacks in lirst iour weeks) and late epiIepsy. PenlieldL4 believes that the “early ,fit” khoulcl not be incIuded in the discussion of lxjst-traumatic epilepsy. It is true that a large percentage of patients with immecliate attacks do not continue to have recurring seizures. In Penfield’s group of 40: cases of civilian head injur?; there were fourteen patients with early seizures out of which only 4 or 28.3 per cent develt tped trauma tic epiIepsy.4’ In Wagstaffe’s 7-cases of gunshot wounds of the head > ii~-t‘ of twelve patients with early seizures became chronic epileptics.“g Ascroft” states that in about one-third of the patients haying fits, the attacks are more or less character. >Mtby:“’ reof :I transient ported thirt>--four cases of con\xlsive dis-

FIG.

7.

kleningocrrebral

cicatrix

with

FIG.

6.

I argc

mcningocerebral cicatrix c:lusin:: wsscls tub :lrd 1 II{, \C:I I_,

traction of the Sylvian

order following penetrating craniocerbral of these ~YISCYthe injur)-. In one-third attacks occurred within a week or t\vo of the injuly; three of these patients continued to ha\~e seizures. In the majority of cases of post-traumatic epilepsy the attacks begin within the first few years. In a series ot joo cases of gunshot wounds of the head, Gliddon’! states that ~3 per cent of t.he cases appeared in the first Tear folIo\ving injur\. The average latent Interval in tlxenty-six cases of post-traumatic epilepsy reported

adjacent

thin,

subdural

membrane.

2 IO

Atncl-icar

Journal

of Surgery

\$TaIker, Johnson-Epilepsy

FIG. 8. Lateral pncumocncephaIogran1 to shoa and out-uouchinp in the direction of a skuII,dcfcct traumatic epilepsy.

by Symonds”” was two years, nine months. In seventeen cases it was three to eighteen months and in nine cases two to sixteen years. In another group of cases reported by EIvidge11~12 the average time of appearance of attacks of forty cases of grand mal was 5.3 years. Twelve appeared within the first vear, tweIve more in one to five years and ten more in five to ten years. One case occurred twenty-five to thirty years after injury. The average time in sixteen cases of petit ma1 was 5.82 years. Ascroft states that at the end of the second year the rate of onset reaches a very low figure and that onI\- about one patient in twenty-seven deveIoped epiIepsy between the fifth and twentieth years. In a series of 246 cases of post-traumatic epilepsy treated at Cushing Genera1 Hospital, 27 per cent of the cases had the first seizure within three months and 30.7 per cent between three and six months after injury.l’ The reported incidence of post-traumatic epilepsy varies considerably. This is probably due to several factors. There are differences in the type of head injury

the mxrkrd in a patient

trigonal suffering

dibntion from post-

occurring in the v-arious groups of cases. Criteria vary as to what to include as a ConvuIsive seizure. Some figures include immediate attacks while others omit these. In a series of 500 cases of gunshot wound of the head including all cases having had one incident Inbelled epilepsy at an\- time folIowing the head injury, GIiddon recases of epilepsy, an ports forty-nine incidence of 9.8 per cent.‘!’ In a similar series of 377 cases with the time of onset varying from two months to seven ant1 one-half years, 1V’agstaffej” givres the same incidence, 9.8 per cent. In other reports of series of cases of gunshot wounds of the head, the incidence of convulsive seizures is as foIIows: Rawling,‘” ~52 cases, 22 per cent; Stevenson,“’ 17,300 cases, 1.5 per cent; Sargent,“’ 18,000 cases, 4.5 per cent; Gushing,” 34 cases, 32 per cent; Maltby,“” 200 cases, 17 per cent. In a series of 40; cases of civilian head injury-, Pentieldli found an incidence of epiIeps>- of 2.’ Per cent. From these figures it can be seen that the incidence of post-traumatic epilepsy including a11 tvPes of head injur) varies from I.5 to 34 per cent.

SUKGICAL

nuixTMENT

SYXIPTOMATIC

procedures because of the small number of cases reported. The usual procedure used for the treat%‘hen medical management fails to cont,rol con\-ukive seizures due to a cerement of post-traumatic epilepsy is the surgical removal of the epileptogenic, tocus. bral lesion, surgical procedures may offer it \vas thought that tht relief if no contraindications exist. It is Until recently lesion, scar, tumor, softening, not wise to subject indil-iduals who are etc., 1V’i.S c :Iphasic, paraplegic the focus responsible for the seizures, but or quadriplegic and lvith the use of electrocorticogrnphv it ha5 cvnfined to bed, to a serious operation for now been clearly shown that the marginal the relief of convulsive seizures which cortex is the source of epileptic discharges. play onI? a minor r6le in their disability. In n recent discussion of post-tr~lumati~. If :I pattent is mentally disturbed or imepilepsy Penfieldla stated : “ \4’c must paired to such a degree as to render him identify more definitelv, not the gross or her incompetent of folIowing medical :kd\-ice, the ad\-isabilit\of operation is lesion but the actual focus of neuronal hyperirritability. doubtful. Because a co&al resection conIt is the focus in M-hi& electrographic spikes and sharp stitutes a major surgical procedure, the \vaves general condition of the patient should be originate and which causes the Lcttacks.” The question arises if the ep~leptogenic as good as possible prior to operation. focus is stable and if it does not xvith the If the primary wound was infected or did passage of time set up other epileptogenic not heal b)- primary intention, it is well foci? These questions cannot be answered to dela\r cortical exploration for at least with assurance at this time although the six months to a year after the wound has available evidence favors the \t:rbilitJ healed. U’hen the operative procedure is and constancyof an epileptogenic focus. then done, chemotherapeutic agents, peniLVhether secondary foci u-ill form if the cillin and sulfadiazine should br used preprimary lesion is left, remains for turther :~nd postoperatively. research. Perhaps that i< the explanation .Just. as in the cases of idiopathic epifor certain failures. leps,v, a number of surgical procedures Cases suitable for this type of surgical ha\-e been tried in the treatment of svmpprocedure must have some e\~iclence of tomntic epilepsy. Meredith3’ has descked localization of the origin of the seizure. ;ln operation for internal communicating These may be apparent from the histor! h>drocephnlus following certain traumatic of the attack, observation of the sci1)r i nilnmmator?; Icsions associated with seizures. The operation consists of ~-1 zure, neurologic findings, electrc,~llcephalographic examination either spontxneous transcortical incision through a relati\-el) or nctixxted, or roentgenogr:~ms of the unimportant portion of the less dominant skull with or without air injection. The cerebral hemisphere le:r\.iny a large stoma localization so determined M-illindicat,e the into the lateral krentricle. At the same time proper placement of the osteoplnstic Ilap. the choroid plexus is remo\recl from the Anesthesiu. 120~al infiltration of’ the opened ventricle. Siri? reported five cases scalp along the proposed in&ion is the of focal epilepsv treated b\- encephsloanesthesia of choice, because it does not m\opesJ-. Thre; patients had not had interfere w+th the cortical acti\,it\ nor the seizures for five to seven months poststate of consciousness of the patient who operatively and the other t\vo had dimican report subjective experiences during nished frequent)and intensity. Marsh:” the corticn1 exploration. klorphine ma> stated impro\-ement occurretl in some be used to alla>- restlessness or anxiety. cases of traumatic epilepsy in 3 series which had been given radiation therap).. Operative Technic. A large t jsteoplast ic It is difticult to evaluate the$e different flap is turned down to expose the epileptoEPILEPSY

c)F;

212

American

Jounml

WaIker,

of Surgery

Johnson-Epilepsy

.laXuaRv. 1948

placed over the longitudinal sinus. This will decrease the likelihood of serious bleeding shouId the patient have a seizure as the result of cortical stimulation. The dura mater must be refIected carefully to avoid tearing subdural adhesions. Sharp dissection is advisable to cut meningocerebral scars. If there is an extensive adhesion, it may be better to open the dura mater only at the margin of the lesion where the epileptogenic focus is thought to lie. When the lesion and surrounding cortex have been exposed, the cortex is systematically explored by electrocorticograph) and electrical stimuIation. A special electrode holder* is cIamped to the margin of the bone. Ten electrodes are placed on the exposed cortex to record the spontaneous and induced activity.

Fig.9.)

Flc. 0. Sketch to show cortical ~lcctrotlcs.

the

method

of application

of

genie focus. Occasionally the type of skin incision may be dictated by previous scaIp wounds. A tantalum plate may have to be removed. In most cases the entire procedure can be done in one stage. In certain instances, however, where the lesion Iies close to the midIine necessitating exposure of the IongitudinaI sinus it rnax be wise to perform the operation in two stages. In the first stage the bone ffap is turned down and tantalum foi1 or pIate

Penfield and Jasper”” describe the Iocalized electrocorticographic alterations in focal epilepsy as consisting of large spikes or sharp waves appearing in random sequence against a background of Iower amplitude activity arising from the border zone between the Iesion a n d normal cortex. In fourteen of thirty-nine cases examined by electrocorticography Walker et al.“’ found such focal spontaneous epileptic activity confined to a portion of one gyrus. (Fig. IO.) Occasionally such a *The signed

dcctrode and made

holder used in this by Dr. C. AlarshaH.

Clinic

\+:e tic

FIG. LO. El~ctrocorticogralII from the right inferior frontal region of n patient sufl;>ring from post-t raum:ltic epilepsy, the result of L( severe penetrating wound of the head. The spiking wztivity at both points o and c) is WCII shown. The horizontal liw at the lx~sr indicates an intcrvd of one second.

-

_~--..

-_-,

-.

-

. .,

_

.-

,

Walker,

.I*NI?,lo.. ,0(X

Johnson-EpiIepsy _~___.

conclttm I

cortex

! D“7-a

of on removal of bon, flap, and AcfccJ In dura mater,

+--

/

mater

of FIG. I 2. Sketch to illustrate the surgical rcmovd

focus was found in norms1 appearing convolutions some distance from the scar. The cerebral cortex is then explored using a sine wave current from I.; to 2.5 volts for five seconds. The electrical activity of the cerebra1 cortex, the motor responses and the patient’s subjective sensation at the time of stimulation are noted. If no change is observed in the electrocorticogram after a minute interval, another point is stimuIated and so on unti1 the entire cortex has been mapped. An after-discharge Insting for a few seconds at the point stimuInted is considered within physioIogic limits. However, generalIy an area of cortex adjacent to the scar can be found which on stimulation induces a Iong Iasting spiky after-discharge. This epileptic discharge has been seen to persist for as Iong as twenty-eight minutes. (Fig. I I.) In many cases the sensory aura which

of a meningocercbrd

“,;I

\CIIl1.ICIcI

scar to the ventricle.

precede the patient’s spontaneous attacks occur with this after-discharge. OccasionaII\; the electrocorticographic attack progresses to a cIinica1 seizure. But in some cases the individual does not have any motor or sensory concomitants during the eIectrocorticographic discharge. With this technic the likelihood of the patient developing a severe seizure is much less than with the use of uncontrolIed electrica stimuIation. There is evidence to suggest that the focus found by electro-activated electrocorticographj and that localized by electrical stimulation aIone are not quite identical. On man)occasions the after-discharge is present for severa minutes and changing in character before the patient notices the aura of his attack. FrequentIy the latter is associated with decreased electrocorticographic activity. This suggests that the

P??OTl

OCCl

Cross-

Suby shad

Indicales

subpial

cortev

ddrk Jlne lndlcates cortex removed.

YEmoved.

~le~t.rocorticohr~~~hic focus is ;I primary t’ocus which fires ;I secondary focus initiating the xura of the attack. Ry rearrangement of the cortical elect rodes and push-pull recording the epileptogenic focus can be delimited. It is usually :~tlixent to the cerebral scar but occasionnll\in ;I p?-rus slightI>. removed. \lulti$le foci are occasionaII\~ present and :I focus may- migrate rareI>, from one C’OItical area to another. In most instances, ho\n,e\,er, the focus is constant after-discharge can be and the typical reproduced time after time. In those cases in which the clinical x111-a and electrcxwrticographic manifestations of an attack are induced it seems probable that the stimulated focus is responsible for the patient’s spontaneous :ttf:Lcks. In those cases in \vhic,h c)nJ,~ the

se&on

resectIon

t FL!4

showlnq Heavy Fart

of

electrocortico~rltI>hic epileptic manifest:\tions can be elicited, the evidence that the point stimulated is the origin ()I’ the attacks is not conclusive, even though there is no doubt that such areas are functioning abnormally and xre pot:enti:lll y i I-‘ not actually epileptogenic. The epileptic focus hsxing been Iocated

and its extent determined, one of sex-era1 technics,

it is r-mm-ccl b>

es&ion of the focus and the scar to the ventricle, rc’mo\xl of the focus 2nd the scar to normal white matter or subpial resection of the SubpiaJ foc>us 1:1,‘?2,;>0 resection of the focus is done by making an incision in an a~ascuIar porti;n of the pia arachnoid and :tbIating the cortex by subpial dissection and suction. The involved gyrus is removed completely from trough to trough. I Figs. 12,

13

~LIKJ

I Jr. )

216

A IIIC~~C:LII

Journal

of Surgery

WaI ker,

area

Johnso n-EpiIepsy

iblated. FIG. ‘4.

The decision as to which of the several technics to use wilI depend on evaIuation of the individual case. It is not desirable to Ieave the patient with a more severe neurologic deficit than he had previously. If the scar is Iarge and adjacent to the motor area, extensive resection may aggravate a paraI?sis or aphasia. The danger of compIicatlons is Iessened if the ventricIe is not opened. In the occasiona case in which a clear cut focus cannot be demonstrated, it is probabIy we11 to excise the scar and entire surrounding epiIeptogenic zone. Irrespective of which procedure is carried out, the blood suppIy to the remaining cerebra1 cortex is disturbed as IittIe as possibIe. FolIowing remova of the focus the eIectrodes may be placed upon the adjacent cortex to determine any remaining epiIeptogenic tissue. In some cases before all convuIsive activity can be eliminated a

rather extensive subpial dissection of severa1 gyri running into the scar must be carried out. During the ablation, speech and motor power are checked frequently. Postoperative Medical Therapy. After any operative procedure for the relief of epiIepsy the medical management, especiaIIy anticonvuIsant drug therapy, must be continued for a period of several years. If no further attacks occur, the medication may be decreased at the end of two years and graduaIly eliminated over another twoyear period. If seizures occur, the medication may be augmented or changed. The patient’s socia1 and economic rehabiIitation wiI1 be greatIy faciIitated if, through vocationa guidance, an occupation, suitabIe to his or her physica1, mentaI and can be found. Conpsychologic status, tinued guidance may be necessary to effect a satisfactory adjustment to a11 of these factors.

The Iargest series of patients treated by surgical extirpations of focal epileptogenic lesions has been reported by Penfield and Erickson.“g During a ten-year period 165 cases were operated upon with ;I mortality of 4.2 per cent. In the group of patients \vith excision of a meningocerebrsl cicatrlx, 22.5 p er cent were completely free from attacks, 6q.j per cent were at Ieast 50 per cent impro\red, I I per cent unchanged and 2 per cent worse. In the group of patients with a cerebral cicatrixi 19 per c~rnt \+.cre completeIy free from attacks, 49 per cent \yere improved, 3o per cent unchanged and 2 per cent worse. III :t group of 2; patients having negative expIorations, 2. j per cent were compIetely free from att:\cl\-s, 22.5 per cent improved, 69 per cent unchanged and 6 per cent worse. At the present time it would seem that cortical excisions may offer help to a certain group 01‘ patients not benefitted b! medical management. fCt?Sllll.S.

I 2.

I 3. 14.

I 5.

10.

17.

18.

I ‘).

20.

21.

REFERENCES I. 41so\,

4.

\\~. :mtl GAIG,

man:rgemcnt

(If brain

If.

\lcK.

abscess.

,4nn.

22.

The surgicn1

S1crg.,

loI

:

-, 1035. 2. AXIWI: r, P. IS. Traumatic 1.

epilepsy after gunshot \vounds of the head. nrir. M. J., I : 730-744, I94 I. 1341~~1~~4, S. I<., tio~xr.ot~, 1.. hl. and KOPELOFI-, lu. Brain lesions associated with cspcrimental “cl,iIeptiform” seizures in the monkc~y. .4m. .J.

I’.s~Tbiur., 100: -27- -37, 4. (.~IK\s, I I. I Icacl injuries

I().$.+.

in war. with especial rcti~rcnc~ to gunshot wounds. Il’ar ,jled., 2: 7p +i j, I l&Z. ;. (.ot%R, S. Causes of epilrpsy. .4wb. .Yeuro/. I” I’v~~biu~., L-: I 215 I 263, 1932. 0. (:0111 \, I. Idatc. complications of brain abscess. I.cll:vn,~o.~c~)pe, 50: I 13()--II4$, *()J.O. -. (:I -1I‘ER, I. S. t.:mdmarks in surgical progress. Benjamin \V. DudIcy and the surgicn1 relief ol 1r:ium:ttic cpilcpsy. Internal. Ahst. Surg., ;(I: X. DIXOU, I I. B. I;. and SIIITIIEI~ D. \I’. EpiIepsyin cysticercosis [Taenia Solium‘i. Quur!. .I. Med., 3: 603, 1)34. 4,. DIXOY, Ii. B. 1~. and HARCKCAVES, CV. H. Cysticercosis (Taenia Soliuml. A further ten years’ clinical study, covering 281 CBSCS. Quart. j. Med., I 3:

23.

EWIDGE, A. R. The posttraumatic convuisi~c :md aIIied states. In Rrock, S. Injuries trf Skull. Brain and Spinal Cord. Pp. 223 261, H:lltimorc. 1943. Williams and \ViIkins. FURLOW, 1.. T. SubpiaI resection of 111e co~-tr\ fog focal epilepsy. Further observai ions. J. ,\. \I. ‘4.. I I I: 2092-2oc)5, 1938. GARLAND, I I. S. Discussion on traumatic cpilcpsy. Proc. Ko,~.. Sot. :Ited., 25: 7-3. -78, 1042. GIBBS, F. A. Frequency with \~hicl~ tumor’s in various parts of the brain produc(~ curl ain symptoms. Arch. Nwro/. I’- ~‘SI chiut., 28: (J&), I932. GIBBS, I;. A., \VI;GULK, IV. II. :tntl Glnt~s, I,.. I_. The electroencephalogram in post-t raurnatic epilepsy. ,4m. J. Ps~chiar., I c~0:-3X- -49, I()++. G~nns, F. A., GIBBS, i. I>. and FI STEH, B. Prculial low temporal localization of sleep-induced scizur< discharges of psychomotor type, prcs<.ntcd at. the. Am. Ncurol. Ass. meeting, Atlantic ( ity. .June 18. ,947. G~nss, F. A. and GIBBS, E. I.. t’rcsc~ntt~cl ;rt th<, meeting of the Ass. Res. Nrr\-. h \Ic,nt. I&s.. New York City, December, 1940. GLIDDOY, \v. 0. Gunshot wounds of t 111.h~;~(l. 1A review of the after-effects in SW) ( ::I n~dian pensioncrs from the Great U’nr, 1014 I<)IX . ~~~~nutl. M. A. J., 49: 373-377, 1943. GROFF, R. A. The meningioma as :I C;~IIW 01 epilepsy. Ann. Surg., IOL: 16-~-1-i, 1!)3j. ImfOLDEN, \v. B. EpiIcpsy occurring twenty yc.nrh after operation for brain abbess. S. (_l~n. .V
365-377, ‘943. 24. hAl-FM.4U, I. C. and &'ALKER, A. 1:. 1%~ elc.ctroencephalograms after head injury. c 1 II ~IWSS.) 2;. I(AC:FhlAV, 1. c., ~~AKSHALL, c. :rnd \~ALKI:K, A. I!. Activated electro-cnccphaIogral,hS in post-t r:tt1matic epilepsy. (In press.) LO. KOPELOFF, 1.. M., BARRERA, S. H. and KWI:LOPF, N. Recurrent convulsive seirures in animals proand them ical m(‘ans. duced by immunoIogic Am. J. l’sycbiat., 98: 881.-902, ,942. 2:. l.Eiiuox, I\‘. G. Personal communication. 28. IXY, A. and \V.~I.KER, A. E. Statistical [cvicw 01 two hundred and thirty intr:lcranial t untorn. lie{,. de cir. de Barcehna, IO. L.IST, C. F. Epileptiform

to:

107,

[I))<.

attacks in c:LL;~‘srbf gliom:l of the cerebral ht~misphcrcs. Relation I o th location and histologic tj-pc‘ of the glioma. -4x% Neural. f? Psycbiat., 35: 323.-350. l(930. 30. RIALTBT, G. I,. Penetrating crnniocerebral injuries. Evaluation of the late results in a group ()I’ ZOG~ consecutive penetrating cranial \j’.ar \\ounds. J. Neurosurg., 3: 239-249, 1946. 3 f \lhns~, C. Post-traumatic epilepsy. SCII~~Cobservations as to its pathogenesis and treatment. Bull.

944. 32. hkR!zDIw, epilepsy

Los

Angeles

Neural.

SK.,

IJ:

-0

4Kl.

’

J. h1. An operation for the rrIicf 01’ following certain traumatic :lml ill-

218

Ame&an Jor~rnalof Surgery

Walker,

Johnson-EpiIepsy it).

33.

and management of brain abscess. Am. J. Sur,q., 63: 10-15, ,944. 34. NEY, K. \I’. Pathologic factors found in the surgical investigation of epilepq-. Am. J. Surg., 17: 573mI\~EREDLTH,

585,

.I. hl.

&cognition

PARL
I I. L.. IlpiIeptiform

1939. 39. PENFIELD, W. and EKI~KSO~, T. C. Epilepsy and Cerebra1 Localization. Vol. 12, p. 623, Springlield, III., ,941. Charles C. Thomas. 40. PENFIELD, M’. and JASPER, H. Elcctrocncel,haIography in foca1 cpilcpsy. Tr. Am. Neural. A., 66: 209-z I I, I 94”. 4r. PENFIELD, iv., EMCKSO~, T. C. and TAI~LOY, I. Relation of intracranial tumors and symptomatic epilepsy. A&. Neurol. P+ f’sycbiut., 41: 300-3’5, 194”. 42. PENFIELD, u’. and HUMPIIREX, 1. Epilcptogcnic lesions of the brain. A histologic study. Arch. Neurol. (5~ Psycbia~., ~3: 240-26 I, 1940. 43. PENFIEI.D, W. and BRIDC.ERS, \V. II. Progressive tissue destruction in epileptogenic lesions of the flrain. Tr. Am. Neural. A., 68: 15X-163, ,942. 44. PEEFIELD, \I’. and SHAVER, XI. The incidence of traumatic epilepsy and headache after head injury in civil practice. Proc. A. liesearcb New. w hfent. Dis., 24: 620-634, ,943. 45. PENFIELD, W. Post-traumatic cpilcpsy. Am. J. Psycbiat., LOO: 750-75 I, ,944. 46. PILCHER, C. Angiomatous malformations of brain; successful extirpation in 3 cases. ,4nn. Surq., 123: 766-784, 1946. 47. QLIADFASEL, F. A. and WALKER, A. E. Problems in post-traumatic epilepsy in an Army General Research I Iospital. (In press.) 48. BAWLING, I>. B. The remote efl‘ccts of gunshot woundsofthe head.Brit.J.Surg., ro:93-126, ~922. 49. ROSEMAN, E. and WOODHALL, B. The electroencephalogram in war w-ounds of the brain; with particuIar reference to post-traumatic epilepsy. Proc. A. Kesectrcb Nerv. cI* ,Llent. Dis., 25 : 200-2 I(),

~915. Acknowledgment:

treat mcnt of Jacksonian rpiIepsy (Horsley open-ation) with ohservntions on areas 4 and 6. Brain, j8: 492~503, 1935. S.m;ENT, P. Some observations on epilepsy. Bruin, 44: 312-328, 1921.

$2. SCARPI;, J. E. A specific

1<)@.

convulsions. Arch. Neural. P Z’sycbitrt., 23: 1032~1o4r, 1932. 36. PBKI;IELD, \V. The mechanism of cicatricial contraction in the brain. Bruin, 50: 491)-5 17. 1927. 37. PENFIELD, IV. EpiIepsy and surgical thcrqy. Arch. Neural. U’BI’s~cbitrt., 36: 449-484, ~936. 38. PENFIELD, W. The epilepsies: with a note on radial therapy. New Englcrnd J. Med., 22 I : zoc)-2 18, 35.

5 I.

S4ctis. E. The subpial Ichcction of t hc cork\ in t II<

Grateful

ncknowlcdgmcnt

epileptic

s~ndronw

tavor-

ably clfcctcd by Iysis of Pacchionian granulations. Prcscntcd at the Am. Nrurol. A. hlrcting, AtIantic City-, Junr 16, ,947. 53. SIMS, J. f1. The ut[lization ofcnccph:lIolnyol,cs~ in srlcctcd casts of Post-tr:lumatic focal epilepsy. Surger),, L3 : 653-060, r943. 54. S-IX\XNSOW, \I’. E. Epilepsy and gunshot wounds 01 thr head. Bruin, $1: 2!1--22~, ,931. 55. SVMOUIX, C. P. Traumatic cpilcpsy. Lance/, 2: 121Tp1220,

1)3j.

TROUPE, C. IX., BAXTEH, D. II. and SCIIA.I.ZKI, R. Observations on cncel,h:lIogr:lphic lindings in ccrcbral trauma. J. Neurosur,q., 3: 390-398, Ic)+h. j'. VAN \i'Ac;EbJLCK, \f'. P. antf f ISI
2118, ,941. W. \V. The incidcncc of traumatic epilepsy after gunshot wound of the head. Luncel,

59. WAGSTAFFE,

2: 861-862,

1928.

60. \V*L KT,H, A. IX. Iinccl,h;llogr:tpl~~ in ci~ildrcn. J. Roenlgenol., 32: 337.-45h, I q3+

/tm.

01. \1:ht_xEt<,A. E. Pelt-traumatic

cpilcpsy. American I.ecturcs in Surgery (in press). 62. \Ir~r.w.l~, A. E., I\~AKSIIAI_I.,C. and BEICESFOICD, !Z~~r~rrr~tc N. EIcctrocorticogr;\I~~~ic char:lcteristics of the ccrcbrum in posttraumatic c’pilcpsv. (In press.)

66. ZISKI~\D, 1:. and Zlsrtrl\n,

E. S. Focal ccrcbral Irsions in rclxtion to hereditary predisposition in cpilcpsy. BuU. Los Aqeles Neural. SK.. 3: 2 I -27, 193x.

is made of a grant by the Simms Fund in aid of tfw work d~~scribctl.