3rournal ofNeurology, Neurosurgery, and Psychiatry 1994;57:1195-1202

Subependymal heterotopia: a distinct neuronalmigration disorder associated with epilepsyA A Raymond, D R Fish, J M Stevens, S M Sisodiya, N Alsanjari, S D Shorvon

The National Hospitalfor Neurology andNeurosurgery, QueenSquare, London, UKA A RaymondD R FishJM StevensS M SisodiyaN AlsanjariS D ShorvonCorrespondence to:Dr A A Raymond, EEGDepartment, NationalHospital for Neurology andNeurosurgery, QueenSquare, London WC1N3BG, UK.

Received 30 December 1993and in revised form21 March 1994.Accepted 27 April 1994

AbstractSubependymal heterotopia has recentlybeen recognised as a cause of epilepsy,but the clinical and investigational fea-tures have not been fully described. Theclinical, psychometric, imaging, andelectroencephalographic features of 13adult patients with subependymal hetero-topia and epilepsy have been reviewed.Age at seizure onset ranged from 18months to 20 years (median 13 years).There were significantly more female(12) than male (1) patients (p < 0.01).Diagnosis of subependymal heterotopiawas made by MRI in 11 patients and CTin two. The heterotopic grey matter wasnodular in 11 patients and diffuse in two;bilateral in eight and unilateral in five.There were significantly more patientswith predominant right than left cerebralhemisphere involvement (p < 0.01). Themost commonly involved site was theoccipital horn of the lateral ventricles (10of 13 patients). Eleven patients presentedwith partial epilepsy, 10 ofwhom also hadsecondarily generalised seizures. Theclinical description of the seizures oftensuggested either an occipital (fourpatients) or temporal (five patients)onset. Two patients presented withabsence attacks without clear focal fea-tures. Patients demonstrated normalearly milestones (12 of 13 patients),including normal motor development (allpatients) and average or above averageintelligence (10 of 13 patients). An EEGexamination showed normal backgroundactivity in all but two patients, one ofwhom had large intracranialhaematomas. Epileptiform activity wasusually widespread (10 of 13 patients) andin three patients, there was generalised 3-Hz spike and wave activity that hadpreviously led to an erroneous diagnosisof concomitant primary generalisedepilepsy. Onset of epilepsy in the seconddecade of life, normal developmentalmilestones and intelligence, and the find-ing of an overwhelming female prepon-derance differentiates subependymalheterotopia from other cortical dysgene-ses. The female preponderance supportsthe importance ofthe X chromosome andsex steroids in the maturation and devel-opment of the cerebral cortex.

(J Neurol Neurosurg Psychiatry 1994;57: 1195-1202)

Migration of neuroblasts from the periventric-ular germinal centres occurs maximallybetween the 7th and 16th gestational weeks,although it is known to continue for severalmonths postnatally.' Neuroblasts migrate cen-trifugally along radial glial fibres towards thepial surface. They come to rest in the corticalplate after detaching themselves from thefibres, in an inside out fashion, with morerecent arrivals lying more superficial to earlierones:2' More recently, there is evidence thatthere is also lateral dispersion of neuroblastsduring migration.4The causes of neuronal migration disorders

may be both genetic and maternal/environ-mental.5 A defect in the radial glial fibres hasbeen proposed to be the underlying mecha-nism for migrational arrest in many ofthe neuronal migration disorders.6 Morpho-logically, the disorders appear either ascollections of neurons in abnormal locations(grey matter heterotopia) or disordered corticallamination and gyration (for example, macro-gyria, polymicrogyria).

Subependymal heterotopia seems to be awell defined and discrete entity-namely,heterotopic grey matter situated adjacent tothe wall of the lateral ventricles,6 just beneathand abutting the ependyma. Sometimes alsotermed "periventricular heterotopia", sub-ependymal heterotopia may represent eitherclusters of neurons that have been arrested inmigration, or failure of programmed celldeath of collections of neuroblasts within theperiventricular germinal matrix. Owing to itslocation and association with normal overly-ing cortex, the defect in the radial glial fibresis likely to be proximal and likely to haveoccurred after most neurons have migrated tothe cortical plate.7 There is a dearth of litera-ture on "pure" cases of subependymal hetero-topia.7-" Many reports either lump all theheterotopias together, or the subependymalheterotopias with other cortical dysgeneses.Certain features of subependymal heterotopianot previously described have become appar-ent, however, and are in our opinion, suffi-cient to make it an entity distinct from otherneuronal migration disorders. The clinical,EEG, psychometric, and neuroimaging fea-tures of 13 adult patients with epilepsy aredescribed and the English medical literatureon similar cases is reviewed.

Patients and methodsOf 69 adult patients with epilepsy and corticaldysgenesis diagnosed on CT or MRI at our

1195 on F

ebruary 14, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 October 1994. D

ownloaded from

Raymond, Fish, Stevens, Sisodiya, Alsanjari, Shorvon

Table 1 Clinicalfeatures and summary of neuroimagingfindings in 13 adults with epilepsy and subependymal heterotopia

Age at Diagnosisl No of miscarriagesseizure onset (Mo:P)/

Patient (y)/sex siblings(S:B) IQ Location ofsubependymal heterotopia Seizure description

1 27/12/F 2:0/0:0 VIQ 112; BL; large nodules; inferolateral margins Becomes vacant and unresponsive; orofacial andPIQ 107 trigones, extending to temporal horn on R. manual automatisms; ± clonic movements R hand;

dysphasic and amnesic afterwards.2 19/11/F NA NP BL; small nodules; superolateral margins Feels as if looking through a dark tunnel with

frontal horns (R > L). approaching flashing lights/vague feeling in head;head turns to L; always progresses to SGTCS.

3 29/20/F 0:0/2:1* FSIQ 73 BL; large nodules; lateral margins body/ (i) Brief absences. (ii) SGTCS with predominantanterior horns LV (CT only). L sided twitching and transient post-ictal

L hemiparesis.t4 29/16/F 0:1/2:0 NP UL; 1 large nodule; medial margin R trigone. Jerking of L hand without LOC; ± SGTCS

(mainly nocturnal).5 26/19/F NA NP UL; diffuse; lateral margin R trigone/posterior Feels as if "on other side" and looking through

body LV. frosted glass; remains responsive and capable ofnormal speech; ± SGTCS.

6 22/15/F 0:0/2:1 FSIQ 119 BL; large nodules; lateral margins body/ Deja vu feeling; becomes unresponsive; oral/manualtrigone LV. automatisms; ± SGTCS.

7 34/13/F 0:0/0:0 NP BL; large nodules; inferolateral margins (i) Brief absences; ± SGTCS, (ii) deja vu feeling;trigones (R > L). abrupt mood changes and altered perception of

surroundings.8 31/17/F 0:0/0:2 NP BL; diffuse; lateral/inferolateral margins Objects become distorted or change colour; becomes

trigones/temporal horns (R > L). unresponsive; ± motor automatisms; ± SGTCS.9 31/3/F 0:3/0:1 VIQ 93; BL; large nodules; lateral/superior margins (i) Sees light flashing in a distance in R eye; stares

PIQ 78; trigones, lateral/inferior margins temporal and becomes unresponsive; manual automatisms;FSIQ 85 horns and lateral margins frontal horns (L > R). ± SGTCS. (ii) becomes unresponsive; violent

shaking ofR limbs, then all four limbs.-10 21/7/M NA FSIQ 69 UL; 1 small nodule; inferolateral margin R (i) Brief absences; ± SGTCS. (ii) becomes quiet;

trigone. shouts for help; arms raised and moves about;shakes all 4 limbs.tt

11 25/1 1/2/F NA VIQ 97 BL; small nodules; superior margins + Aura of an urge to urinate; becomes unresponsive;PIQ 86 trigones/occipital horns (R > L). dystonic posturing ofR hand; orofacial automatisms;FSIQ 92 dysphasic afterwards; ± SGTCS.

12 31/20/F NA/1:0§ NP UL; large nodules; lateral margin R trigone.11 Becomes unresponsive, R side becomes weak;manual automatisms; incontinent of urine; confusedafterwards.

13 28/1 1/2/F 1:0/0:0 NP UL; large nodules; lateral margin R trigone ± Aura of "fizziness" in head; ringing in ears; white(CT only). spots cross visual fields; becomes unresponsive and

stares into space; confused afterwards; ± SGTCS.

Mo = mother; P = patient; B = brothers; S = sisters; VIQ = verbal IQ, PIQ = performance IQ; FSIQ = full-scale IQ, SGTCS = secondarily generalised tonic-clonic seizure; BL = bilateral; UL = unilateral; LV = lateral ventricle; * a twin brother and two older sisters who are also twins; tassessed to be pseudoseizures onvideo-EEG telemetry; tonset after head injury at age 16; §twin sister died at birth, cause unknown; Iladditional subcortical white matter nodular heterotopia;NA = not available; NP = not performed, but IQ assessed to be average or above average from profession/interaction with medical staff.

centre, 13 had subependymal heterotopia andwere included in the study. Age at diagnosisranged from 19 to 34 years (median 27 years).All patients attended either the NationalHospital for Neurology and Neurosurgery orthe Chalfont Centre for Epilepsy. Clinicaldetails, including age at seizure onset, seizuredescription, antenatal and birth history, devel-opmental milestones, family history, neuro-psychological assessment, and physical signswere obtained retrospectively by reviewingtheir medical records (table 1). Two patientsproceeded to surgery and their pathologicalslides were reviewed by one of us (N A).

Diagnosis of subependymal heterotopiawas based on neuroimaging. All patients hadCT and 11 also had MRI. In nine patients,MRI was performed on a 1-5 Tesla GE Signaunit (GE Medical Systems, Milwaukee,USA). Five of these patients also had 1P5 mmcontiguous coronal slices using a 35/5/1(TR/TE/NX) pulse sequence, flip angle 350,and matrix size 256 x 128, which permittedmultiplanar reformatting with an independentconsole. The hippocampal volumes of thesefive patients were also measured by themethod described by Cook et al.16 Of theremaining patients, one was scanned on a 1-5Tesla Siemens Magnetom unit, and the otheron a 0-26 Tesla Picker International system.The EEGs were recorded on a minimum of

16 channels with the 10-20 international sys-tem of electrode placement, and incorporatedhyperventilation and photic stimulation.

Statistical analysis was performed by thetwo tailed binomial test and a p value <0 05was deemed significant.

ResultsCLINICAL FEATURES (TABLE 1)There were significantly more female (n=12) than male (n = 1) patients (p < 0-01).Age at seizure onset ranged from 18 monthsto 20 years (median 13 years). Nine of 13patients developed seizures between the agesof 11 and 20 years. Family histories wereunremarkable for epilepsy and tuberous scle-rosis. Further analysis in nine patients showeda total of 12 siblings (seven females and fivemales), none of whom have epilepsy, althoughone patient (4) has two sisters with undiag-nosed early morning jerks. Three patients hadno siblings. To date, none of our patients'relatives have had brain scans.A personal or maternal history of sponta-

neous abortions was present in four of eightpatients in whom this information was avail-able (table 1); the mother of one of thesepatients (1) had also had an ectopic preg-nancy. A history of prenatal or perinatal prob-lems was present in eight patients-namely,preterm delivery (<37 weeks) in four (patients2, 3, 9, 12), post-term (>40 weeks) delivery intwo (10, 13), pre-eclampsia in two (9, 13),forceps delivery in two (1, 13), twin pregnan-cies in two (3, 12), and pyrexia shortly afterbirth in one (6). Patient 6 also had a persistent

1196

on February 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 O

ctober 1994. Dow

nloaded from

Subependymal heterotopia: a distinct neuronal migration disorder associated with epilepsy

ductus arteriosus that was surgically repairedat age 30 months.

Delayed early developmental milestoneswere reported in only one patient (10). Theproduct of a prolonged labour induced at 42weeks of gestation, he sat and walked at 12and 30 months respectively, and also mani-fested aggressive and compulsive behaviourfrom the age of 13. At the age of 16 he had aserious head injury from a fall, resulting inhaematomas in the right frontal and parietallobes and alteration to his seizure pattern.

Childhood febrile convulsions werereported in two patients (2, 11). One of thesepatients"1 had a prolonged febrile convulsion(> 20 minutes) at 10 months and showedMRI and subsequently pathological evidence,of concomitant hippocampal sclerosis (seelater).

In all patients, the seizures were medicallyrefractory, in as much as they either interferedwith employment or lifestyles or occurred atleast once a month and were resistant to atleast two consecutive antiepileptic drugs.Seven patients (1, 3, 7, 9, 10, 11, 12) hadadditional descriptions of their habitualseizures from video-EEG telemetry. The clini-cal features of the seizures were hetero-geneous, suggesting an extratemporal onset insix patients and a temporal onset in five, andwere generalised at the onset in the remainingtwo (table 1). Ten of the 11 patients with par-tial epilepsy had secondarily generalisedseizures as well. Of the extratemporal cases,four complained of visual auras suggesting anoccipital onset: flashing lights in two, "whitespots" in one, and distortion of shape andcolour in another. In one patient (2), theseizures always became secondarily gener-alised, and in two patients (3, 13), there wasrapid loss of awareness, but progression tosecondary generalisation was rare (< oneevery two years). In the fourth patient (9),prominent manual automatisms such as fid-dling with clothing ensued, suggesting spreadto temporal structures. Localisation of seizureonset was less clear in the other two patientswith extratemporal epilepsy but the clinicalfeatures suggested possible onset in the motorcortex, namely, clonic movements of the lefthand in one patient (4) and rapid weakness ofthe right side in the other (12). Features sug-gesting a temporal onset in the other fivepatients with partial epilepsy included orofa-cial/manual automatisms, psychic/experientialphenomena, or both, and in one patient (11),clear dystonic posturing of the contralateralarm. The two patients (2, 10) without focalonsets had only brief absences which at times,progressed to generalised tonic-clonic convul-sions. Three patients (3, 9, 10) had coexistentpseudoseizures documented on video-teleme-try. In one of these patients (10), pseudo-seizures were the dominant seizure type. Ahistory of convulsive status epilepticus waspresent in three of 12 patients (3, 7, 9),although pseudostatus was suspected in one(3).

Physical examination did not show any per-manent neurological deficits. Examination of

the skin showed two cafe au lait spots inpatient 8 and one in patient 5, but no otherstigmata of tuberous sclerosis, although not allwere examined with UV light. Concomitantcongenital abnormalities were present in twopatients: a persistent ductus arteriosus(patient 6) and shortened terminal phalangesof the thumbs and great toes (patient 9).Patient 10 was a normal male phenotypically,although we have not been able to performchromosomal analyses on him.

Results of formal psychometric tests wereavailable for five patients. Only two patientshad an IQ <80, one of whom (10) wasassessed only after he had had extensiveintracerebral haematomas from a fall. Theintellectual capacity of the other patients wasjudged by their occupations, educationalachievements, and interaction with medicalstaff, and was generally average or aboveaverage.Two patients (1, 11) have had anterior

temporal lobectomies after a diagnosis of con-comitant hippocampal sclerosis based on clin-ical, EEG, and neuroimaging data (seeearlier). In both patients, the resection did notinclude the heterotopic tissue, this being con-firmed by postoperative MRI. Pathologicalexamination confirmed the diagnosis of hip-pocampal sclerosis, and in patient 11, therewas neuronal ectopia in the subcortical whitematter (or "microdysgenesis", as defined byMeencke and Veith'6) and splitting of thedentate fascia. Only patient 11 had a definitehistory of childhood febrile convulsions.Patient 1 has been followed up for eightmonths after operation and has had completefreedom from seizures and auras. In the 10months of follow up, patient 11 has had onlyone seizure, similar to her preoperativeseizures, representing a greater than 90%reduction in seizure frequency.

NEUROIMAGING FINDINGS (TABLE 1)The brain CT scans of four patients wereavailable for review. The heterotopic tissueappeared either as broad bands lying along thelateral walls of the lateral ventricles (patient 5)or multiple nodules protruding into the lateralventricles (patients 3, 6, 13). These lesionshad densities similar to that of the cortical rib-bon and calcification was not seen (fig 1).Reports for CT were available for eight of thenine remaining patients; only one reportedsubependymal heterotopia.The anatomical delineation of the hetero-

topic tissue by MRI was excellent: only thelateral ventricles were involved, and the het-erotopic tissue was isointense to grey matteron all sequences. The distribution of thesubependymal heterotopia was bilateral ineight and unilateral in five patients. Whenunilateral, the heterotopic tissue was alwayson the right. When bilateral, it was eithersymmetric or more extensive on the right,except for patient 9, where it was moreextensive on the left. The right compared toleft sided predominance was statistically sig-nificant (p < 0.01). Patient 12 had evidence ofadditional large nodules of heterotopic grey

1197

on February 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 O

ctober 1994. Dow

nloaded from

Raymond, Fish, Stevens, Sisodiya, Alsanjari, Shorvon

Figure 1 Axial CTimage showing multiplesueedmal heterotopic

nodules (arrows).

Figure 2 Coronal imagefrom a volumetricacquisition with strong Tlweighting (see text fordetails), showing bilateralnodular subependymalheterotopia projectingslighdy into each occipitalhorn from the inferolateralaspect.

Figure 3 Coronal imagefrom a volumetricacquisition with strong Tlweighting (see textfordetails), showing bilateraldiffuse subependymalheterotopia around thelateral and inferiormargins of the atrium ofeach lateral ventricle.

matter in the subcortical white matter adja-cent to the subependymal heterotopia as wellas expansion of the ipsilateral temporal lobe.

Like CT, on MRI, the subependymal het-erotopia appeared either as solitary nodules(nodular subependymal heterotopia, fig 2)projecting into the ventricles (nine patients)or diffuse broad bands (diffuse subependymalheterotopia, fig 3) lining the ventricles (twopatients). The nodules were either rounded orovoid. We have arbitrarily classified thenodules as small, when they were <5 mm insize, or large (table 1). The diffuse bands hadan irregular and "lumpy" outline and seemed

5 to contain multiple nodules which had| "coalesced". Whether the heterotopic grey

matter was nodular or diffuse did not seem toaffect its unilaterality or bilaterality or seizurefrequency.The trigones and occipital horns were the

most commonly involved sites (10 patients).' In one patient (2), MRI showed only one

small nodule lying adjacent to the superolat-eral margins of each frontal horn. The otherpatient (3) without evidence of peritrigonalinvolvement had CT only (see earlier), and

P thus could have had more extensivesubependymal heterotopia. In two patients,the subependymal heterotopia extended for-wards from the trigones to the body of the lat-eral ventricles, and in four patients, to thetemporal horns. The frontal horns wereinvolved in four patients. When the temporalhorns were involved, the subependymal het-erotopia sometimes abutted or displaced thehippocampal formation (patient 9, fig 4).None of the patients exhibited additional

gyral abnormalities or more extensive areas ofcortical dysgenesis. Hippocampal volumemeasurements in five patients showed sym-metric hippocampi in three (4, 7, 8) andasymmetric hippocampi in the remaining two(1, 11). Patient 1 had a smaller right hip-pocampus with a right to left ratio of 0-84 andpatient 11 a smaller left hippocampus with aleft to right ratio of 0-65. Both patients hadbilateral subependymal heterotopia, withmore extensive involvement on the right. Inpatient 1, the subependymal heterotopiaextended to the temporal horn adjacent to thesmaller hippocampus but did not abut theipsilateral hippocampus.EEG FINDINGSUnilateral subependymal heterotopia (n = 5)Only two patients showed focal and later-alised EEG abnormalities. These were runs ofhigh amplitude, 2-Hz to 3-Hz activity, andfrequent spike and slow wave activity in the

j hemisphere containing the subependymal het-erotopia, maximal in amplitude and frequencyanterior to the abnormality (patient 12), andfrequent focal sharp waves in the right occipitalregion in addition to a mild generalised back-ground abnormality (patient 13, fig 5).Of the remaining three patients, one

showed a generalised background abnormalityonly, with loss of the alpha rhythm and anexcess of slow activity (patient 10), oneshowed generalised 3-Hz spike and wave

1198

on February 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 O

ctober 1994. Dow

nloaded from

Subependymal heterotopia: a distinct neuronal migration disorder associated with epilepsy

Figure 4 A pair of nearlycontiguous images from amultislice acquisition usingan inversion recoverysequence (TR 20001TI7001TE 18) showingnodular subependymalheterotopia, involving thelateral aspect andfloor ofeach temporal horn (farlarger on the left), and thelateral margin of the rightfrontal horn (arrows).Note the unusualconfiguration of eachhippocampus.

activity mimicking primary generalisedepilepsy (patient 4) and finally, one showedan unusual burst of bilateral theta activityduring overbreathing, which was unaccompa-nied by a change in clinical behaviour. Photo-sensitivity was not seen in any case.

Bilateral subependymal heterotopia (n = 8)The alpha rhythm was preserved in allpatients, but in two patients (6, 7), there wasalso a generalised excess of slow activity.

Figure 5 A monopolarEEG recording (usingaverage reference) showingfocal sharp waves in theright occipital region (02)and a mild generalisedbackground abnormality.This patient's CT (patient13) showed nodularsubependymal heterotopiaprojecting into the trigoneof the right lateralventricle. Channel 16(01), not displayed here,did not show similar sharpwaves.

There were unequivocal epileptiform abnor-malities in all patients. Two patients (2, 7)showed generalised 3-Hz to 4-Hz spike andwave activity mimicking primary generalisedepilepsy (fig 6); in one (2), this was the onlyabnormality and in the other, there was addi-tional focal right temporal spike and slowwave activity. In the remaining patients, therewere focal spikes, sharp waves, or sharp andslow wave activity, distributed either indepen-dently or synchronously in both temporallobes (1, 11), or widely in one hemisphere, in

FP2

F4

C4

P4

FP1

F3

C3

P3

F8

T4

T6'

02

F7

T3

T5

HFF70 Hz TC0.3s 100 vi S

1199

on February 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 O

ctober 1994. Dow

nloaded from

Raymond, Fish, Stevens, Sisodiya, Alsanjari, Shorvon

Figure 6 Bursts ofgeneralised spike and waveactivity mimickingprimaty generalisedepilepsy in a 34 year oldwoman with bilateralnodular subependymalheterotopia (patient 7).

FP2-F8

F8-T4

T4-T6

T6-02

FP1-F7

F7-T3

T3-T5

T5-01

FP2-F4

F4-C4

C4-P4

P4-02

FP1-F3

F3-C3

C3-P3

HFF35Hz TC0-3s 100pV1 s

the frontocentrotemporal region (3, 6, 9). Inthe remaining patient (8), there were bilateralindependent spike and wave complexes with-out any clear localisation. Photosensitivity wasnot seen in any case.

DiscussionGrey matter heterotopias are clusters of neu-rons thought to have been arrested duringmigration. They appear as nodules or bands,and may be subpial, subcortical (in the cen-

trum semiovale), or subependymal in loca-tion.6 These subdivisions are to a certainextent artificial, in that they may be found inassociation with each other or coexist withother cortical dysgeneses (for example, macro-

gyria, polymicrogyria, and schizencephaly).When in a subependymal location, the nod-ules may either be discrete or may be closetogether to form irregular or "lumpy" bands

adjacent to the lateral ventricles, often bilater-ally and symmetrically. We consider, how-ever, that the term "band heterotopia" as usedby some authors9 to describe the latter, is mis-leading and should be reserved only for sub-cortical bands of heterotopic grey matter inthe centrum semiovale, as described previ-ously as a specific entity.'8 '9 We prefer theterms nodular and diffuse subependymal het-erotopia. Overall, subependymal heterotopiais more likely to be bilateral than unilateral(tables 1 and 2). Also, it is our experience as

well as the experience of some other authors'3that subependymal heterotopia is the com-

monest form of grey matter heterotopia, atleast in the adult population, based on imag-ing alone.A defect involving the radial glial fibres,

which act as guides for the migration of neu-

rons has been proposed to be the mechanismfor subependymal heterotopia. This defect

Table 2 Reported age at seizure onset and anatomical and sex distribution in subependymal heterotopia

Number of cases

Bilaterall No known Seizure onsetAuthors Total Femalelmale unilateral seizures > 10 years Comments

Barkovich and Kjos' 8 6/2 7/1 2 5/6 Normal intelligence in six of eight patientsDiMario et al9 2 2/0 2/0 0 2/2 Report of a mother and daughterHayden et al'° 1 1/0 0/1 0 0 CT only; seizure onset at age 9 yearsHutterliocher et al I" 7 6/1 5+2/Ot 2 3/5 Includes six patients from one family of four

generationsKernohan" 1 1/0 1/0 0 0 Post-mortem diagnosis; patient died from

status epilepticus in infancy; additionalaqueductal stenosis

Palmini et al 1 1/0 1/0 0 ? Patient aged 36 years at evaluationPalmini et al7 1 1/0 1/0 0 0 Seizure description suggests occipital onsetSmith et al ' 7 ? 6/1 1 6/6 Sex of only two patients (IF, IM) givenZisch and Artmann 14 2 ? 1/1 0 2/2 A third patient had associated pachygyriaCurrent series 13 12/1 8/5 0 9/13Total 43 30/4*** 34/9 5 27/34

*** < 0-0001.tReported as bilateral in five, and presumed bilateral in two cases.

1200

on February 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 O

ctober 1994. Dow

nloaded from

Subependymal heterotopia: a distinct neuronal migration disorder associated with epilepsy

may be due to direct damage to the fibres,deficiency of adhesion molecules necessary forthe migration of neuroblasts along radial glialfibres, or premature transformation of thefibres into astrocytes.'720 The timing of thisdefect is probably late during the migrationperiod-that is, nearer the 16th than the 7thgestational week, when most of the neuronshave migrated to the cortical plate. Althoughmost of our patients (eight of 13) had a his-tory of prenatal or perinatal problems, thesewere generally mild, and the numbers are toosmall to make any inferences regarding aetiol-ogy. Palmini et al7 also postulated that inaddition to the timing of events, the site(proximal) of the defect in the radial glialfibres and number of fibres involved may alsobe important factors in the pathogenesis ofsubependymal heterotopia. How prenatal andperinatal factors cause selective and not dif-fuse defects in the radial glial fibres remainsunknown although focal subependymalinfarcts or haemorrhages have been pro-posed.'

Subependymal heterotopia can be confi-dently diagnosed by neuroimaging.6 8 9 13 Itsanatomical delineation by CT is inferior toMRI.13 On CT, irregularities along the lateralwalls of the lateral ventricles may be the onlyclue to the diagnosis of subependymal hetero-topia.'4 High resolution MRI can preciselydelineate the morphology, distribution, andextent of subependymal heterotopia, and wehave shown that subependymal heterotopiaaffects the trigones most often and frontalhorns of the lateral ventricles least often. Wehave no explanation for this finding at pre-sent. In our experience, small nodules mayrequire the use of thin section volumetricMRI to be adequately visualised (for example,patient 2).The main differential diagnosis of

subependymal heterotopia on neuroimaging istuberous sclerosis. The salient features thatdifferentiate the nodules of subependymalheterotopia from the hamartomas of tuberoussclerosis have been outlined by Barkovich etal6 and include their ovoid and smooth ratherthan elongated and irregular shape, isointen-sity to grey matter rather than isointensity orhypointensity to white matter, and lack ofenhancement after gadolinium injection. Also,the lack of calcification of the nodules on CTwould favour subependymal heterotopiarather than tuberous sclerosis. 13 Moreover,none of our patients had a family history orstigmata of tuberous sclerosis.The predilection of subependymal hetero-

topia for the right cerebral hemisphere isintriguing and has not been previously high-lighted. One explanation may be that the rightsided neuroblasts complete migration slightlylater than those on the left.

Seizure onset in our series was typically inthe second decade (nine of 13 patients) andthis is the finding of other authors as well(table 2). The seizure pattern associated withsubependymal heterotopia is variable.8'4 Bothpartial as well as generalised epilepsy havebeen reported, and in our series, do not seem

to be dictated by the distribution and size ofthe subependymal heterotopias. In fact,whether subependymal heterotopias are aetio-logical or just a marker of a more widespreaddevelopmental abnormality has been raised bysome authors.'3 Moreover, it has been diag-nosed in patients without epilepsy (table 2).Indeed, the fact that the two patients withconcomitant hippocampal sclerosis in ourseries have become seizure free or consider-ably improved after standard temporal lobec-tomy questions the relevance of some, if notall subependymal heterotopias, in epilepsy.To date, we have not been able to find reportsof surgical resection of subependymal hetero-topia for the control of epilepsy in the Englishliterature. Morrell et al 21 have demonstrated,however, with the use of intracerebral record-ings, that subcortical laminar heterotopic greymatter can generate its own epileptic activity,thus supporting intrinsic epileptogenicity ofheterotopic grey matter, although we are notaware of the use of a similar technique inpatients with subependymal heterotopia.Although most of our patients have had severeepilepsy from childhood or adolescence, thismay reflect patient selection bias, and the fre-quency of subependymal heterotopia inpatients with mild or later onset epilepsy isnot known.The most likely explanation for the variabil-

ity in the epileptogenicity of subependymalheterotopia, which probably also applies toother cortical dysgeneses, is that some otherfactor or insult, probably in postnatal life, issometimes necessary to trigger the epilepsy.The relation of hippocampal sclerosis to corti-cal dysgenesis is beyond the scope of thispaper but it will be interesting to see if ourtwo patients with concomitant hippocampalsclerosis remain seizure free during subse-quent follow up. Despite these controversies,it is our belief that subependymal heterotopia isepileptogenic in all our patients.

Three patients in our series showed 3 or 4per second generalised spike and wave activ-ity. Whether this is concomitant primary gen-eralised epilepsy or due to the subependymalheterotopia is unresolved. Clinically, only oneof these patients (patient 7) had absences sug-gestive of primary generalised epilepsy but herattacks have been medically refractory despiteadequate doses of sodium valproate, ethosux-imide, and phenytoin. The onset of epilepsyin the second decade, the negative family his-tory for epilepsy, the absence of early morningmyoclonic jerks, and lack of photosensitivity,are all further points against the diagnosis ofprimary generalised epilepsy in any case.22The possibility then exists, that there may besome patients, particularly females, who havean EEG diagnosis of, and masquerade asprimary generalised epilepsy, who havesubependymal heterotopia. The possible rela-tion of this finding to the suggestion thatmicrodysgenesis may be the pathological basisof primary generalised epilepsy in somecases'3 is also intriguing.A female predominance has been reported

in bilateral band heterotopia ("double cortex

1201

on February 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 O

ctober 1994. Dow

nloaded from

Raymond, Fish, Stevens, Sisodiya, Alsanjari, Shorvon

syndrome")'819 and Aicardi syndrome.24Subependymal heterotopia would be the thirdneuronal migration disorder associated with afemale predominance. A review of the Englishmedical literature of cases of "pure"subependymal heterotopia also shows afemale predominance (table 2). When thesecases are combined with our cases, there is atotal of 30 female and four male patients. Thisfemale predominance was statistically signifi-cant (p < 0 0001). Subependymal heterotopiathat is associated with other cerebral abnor-malities (for example, agenesis of the corpuscallosum and other cortical dysgeneses) doesnot seem to be associated with a sex predomi-nance, but seems to be associated with mentalretardation and a poor prognosis.25-28 Theterm "pure subependymal heterotopia" hasbeen used with caution because the diagnosisin all previous reports except one," includedin this analysis were based on neuroimaging.As illustrated by our patient with microdysge-nesis and splitting of the dentate fascia, not allcortical dysgeneses will be detected by neu-roimaging. And, it is possible that some of thepatients in our and previous reports mighthave abnormalities beyond the resolution ofneuroimaging.The reason for the female predominance is

unclear, but an X linked dominant inheri-tance with prenatal lethality in hemizygousmales, a mechanism proposed for Aicardi syn-drome,29 30 has been suggested for some cases:Huttenlocher et al 15 recently reported sixmembers of a family (all of whom werefemales) with subependymal heterotopia;multiple spontaneous abortions and a defi-ciency of male offsprings strongly suggested Xlinked dominant inheritance with prenatalmale lethality. Four out of nine of ourprobands had male siblings. This does notexclude an X linked dominant inheritancebecause there is still a 25% chance of ahealthy male being born. Because well docu-mented male cases exist, however, there mustbe sporadic forms of the condition or alterna-tive as yet unrecognised factors.The female predominance in subependy-

mal heterotopia and bilateral band heterotopiais striking and would lend support to the role ofthe X chromosome in cerebral development.

In conclusion, subependymal heterotopia isclearly distinct from other neuronal migrationdisorders: females are predominantly affected,age at seizure onset is usually in the seconddecade, and intelligence is often normal oronly mildly compromised. And, in the pres-ence of preserved intellectual function, somepatients whose EEGs show normal back-ground activity and 3-Hz or 4-Hz spike andwave activity may masquerade as primarygeneralised epilepsy. Subependymal hetero-topia is more'often bilateral than unilateral,most commonly affects the occipital horns ofthe lateral ventricles, and affects the rightmore than the left cerebral hemisphere.Further studies are required to investigate theinteraction of genetic and perinatal factors inthe pathogenesis of subependymal hetero-

topia, and the prevalence of asymptomaticand familial cases.

1 Sarnat HB. Cerebral dysplasias as expressions of alteredmaturational processes. Can J Neurol Sci 1991;18:196-204.

2 Angevine JB Jr, Sidman RL. Autoradiographic study of cellmigration during histogenesis of cerebral cortex in themouse. Nature 1961;192:766-8.

3 Rakic P. Neuronal migration and contact guidance in theprimate telencephalon. Postgrad Med J 1978 (suppl);54:25-40.

4 Walsh C, Cepko CL. Clonal dispersion in proliferativelayers of developing cerebral cortex. Nature 1993;362:632-5.

5 Barth PG. Disorders of neuronal migration. Can J NeurolSci 1987;14:1-16.

6 Barkovich AJ, Gressens P, Evrard P. Formation, matura-tion, and disorders of brain neocortex. AJNR Am JNeuroradiol 1992;13:423-46.

7 Palmini A, Andermann F, de Grissac H, et al. Stagesand patterns of centrifugal arrest of diffuse neuronalmigration disorders. Dev Med Child Neurol 1993;35:331-9.

8 Barkovich AJ, Kjos BO. Gray matter heterotopias: MRcharacteristics and correlation with developmental andneurologic manifestations. Radiology 1992;182:493-9.

9 DiMario FJ, Cobb RJ, Ramsby GR, Leicher C. Familialband heterotopias simulating tuberous sclerosis.Neurology 1993;43:1424-6.

10 Hayden SA, Davis KA, Stears JC, Cole M. MR imaging ofheterotopic grey matter. J7ournal of Computer AssistedTomography 1987;11:878-9.

11 Kernohan JW. Cortical anomalies, ventricular heterotopiasand occlusion of the aqueduct of Sylvius. Archives ofNeurology and Psychiatry 1930;23:460-80.

12 Palmini A, Andermann F, Olivier A, et al. Neuronal migra-tion disorders: a contribution of modern neuroimagingto the etiologic diagnosis of epilepsy. Can J Neurol Sci1991;18:580-7.

13 Smith AS, Weinstein MA, Quencer RM. Association ofheterotopic grey matter with seizures: MR imaging.Radiology 1988;168:195-8.

14 Zisch R, Artmann W. MR in the diagnosis of heterotopicgrey matter. Neuroradiology 1991;33:527-8.

15 Huttenlocher PR, Taravath S, Mojtahedi S. Peri-ventricular heterotopia and epilepsy. Neurology 1994;44:51-5.

16 Cook MJ, Fish DR, Shorvon SD, Straughan K, StevensJM. Hippocainpal volumetric and morphometric studiesin frontal and temporal lobe epilepsy. Brain 1992;15:1001-15.

17 Meencke HJ, Veith G. Migration disturbances in epilepsy.In: Engel J Jr, Wasterlain C, Heinemann U, Avanzini G,eds. Molecular neurobiology of epilepsy. Amsterdam:Elsevier, 1992:31-40.

18 Barkovich AJ, Jackson DE Jr, Boyer RS. Band hetero-topias: a newly recognised neuronal migration anomaly.Radiology 1989;171:455-8.

19 Palmini A, Andermann F, Aicardi J, et al. Diffuse corticaldysplasia, or the 'double cortex' syndrome: the clinicaland epileptic spectrum in 10 patients. Neurology 1991;41:1656-62.

20 Barkovich AJ, Chuang SH, Norman D. MR of neuronalmigration anomalies. A3rNR Am J Neuroradiol 1987;8:1007-17.

21 Morrell F, Whistler WW, Hoeppner TJ, et al.Electrophysiology of heterotopic gray matter in the 'dou-ble cortex' syndrome. Epilepsia 1992;33(suppl 3):76.

22 Niedermeyer E. Epileptic seizure disorders. In:Niedermeyer E, Da Silva FL, eds. Electro-encephalography: basic principles, clinical applications, andrelated fields. Baltimore: Williams and Wilkins, 1993:461-564.

23 Meencke HJ, Janz D. The significance of microdysgenesiain primary generalized epilepsy: an answer to theconsiderations of Lyon and Gastaut. Epilepsia 1985;26:368-71.

24 Aicardi J, Lefebvre J, Lerique-Koechlin A. A new syn-drome: spasm in flexion, callosal agenesis, ocular abnor-malities. Electroencephalogr Clin Neurophysiol 1965;19:609-10.

25 Bergeron RT. Radiographic demonstration of corticalheterotopia. Acta Radiol 1969;9: 135-9.

26 Mueller CF. Heterotopic grey matter. Radiology 1970;94:357-8.

27 Bairamian D, Di Chiro G, Theodore WH, Holmes MD,Dorwart RH, Larson SM. MR imaging and positronemission tomography of cortical heterotopia. Journal ofComputer Assisted Tomography 1985;9: 1137-9.

28 Yasumori K, Hasuo K, Nagata S, Masuda K, Fukui M.Neuronal migration anomalies causing extensive ventric-ular indentation. Neurosurgery 1990;26:504-7.

29 Pembrey ME. Genetic factors in disease. In: WeatherallDJ, Ledingham JGG, Warrell DA, eds. Oxford textbook ofmedicine. Oxford: Oxford University Press, 1987:4.1-440.

30 Carter CO. Sex-linkage and sex-limitation. In: Ounsted C,Taylor DC, eds. Gender differences: their ontogeny and sig-nificance. Edinburgh: Churchill Livingstone, 1972:1-12.

1202

on February 14, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1195 on 1 O

ctober 1994. Dow

nloaded from