Hippo Cam Pal Sclerosis in Children With Intractable Temporal Lobe Epilepsy- Detection With MR Imaging

8/3/2019 Hippo Cam Pal Sclerosis in Children With Intractable Temporal Lobe Epilepsy- Detection With MR Imaging

http://slidepdf.com/reader/full/hippo-cam-pal-sclerosis-in-children-with-intractable-temporal-lobe-epilepsy- 1/4

1045

0361-803X/93 /1615-1045

ci Am erican Roen tge n Ra y S oc ie ty

H ippocam pa l S c le ro s is in

C h ild ren w ith In trac tab le

Tem po ra l Lobe Ep ilep sy :Detection with MR Imaging

J . 0 . G ra tt an -S mi th 1

A. S. Harvey2

P . M. D es mo nd 3

C . W . Chow4

Rece ived M ay 6 , 19 93 ; ac cep te d a fte r re v is ion

June 28, 1993.

1 D epartm en t o f R ad io log y , R oya l C hild ren ’s

Hosp ita l, F iem in g to n Rd ., P arkv i lle , v ic to ria , 30 52 ,

A us tra lia . A dd res s co rre spondence to J . D . G na t-

tan -Smith .

2D epan tm en t o f P ae d ia tn cs , U n iv e rs ity o f M e l-

b o ume , M e lb o um e , A u st ra li a.

3D epa rtm en t o f R ad io log y , R oya l M elb oum e

H osp ita l, M elbo um e, A us tra lia .

4Depa rtm en t o f A na tom ic a l P a tho lo g y , R o ya l

C h ild re n ’s H o sp ita l, P a rkv ille , v ic to ria , 30 52 , Au s-

tral ia.

OB JECT IVE . H ippocam pal s c le ro s is is th e m os t com m on les ion assoc ia ted w ith

tem po ra l lo b e ep ile p sy . T em po ra l lo b ec tom y Is e ffe c tiv e fo r th e con tro l o f m ed ic a lly

re frac to ry se izu re s In the se pa tien ts . B e fo re the adven t o f M R im ag in g , h ippocam pa l

sc le ro s is w as rare ly d iagnosed p reope ra tive ly . T h e pu rpose o f th is s tudy w as to

de te rm in e th e frequency o f h ippocam pa l s c le ros is In ch ild ren w ith In tra c tab le tem po -

ra l lo b e ep ile p sy and th e accu ra cy and re lia b ility w ith w h ich h ippocam pa l s c le ro s is

c an be d iagnosed on th e bas is o f M R find ings in ch ild ren .

M A TER IALS AND ME THODS . W e rev iew ed the M R im ages o f 5 3 ch ild ren (m ean age ,

10 yea rs ) w ith m ed ic a lly re frac to ry tem po ra l lo b e ep ilep sy . The M R im ages w e re

re vie we d b lin dly and Independen tly b y tw o rad io lo g is ts o n tw o oc ca sio ns , and w e re

c las s ifie d as show ing h ippocam pa l sc le ro s is , o th e r le s ions , o r no abno rm a lity . H ippoc-

am pa l s c le ro s is w as d iagnosed w hen ev id ence o f h lppocam pa l a trophy w as p resen t o r

h ippocam pa l s igna l in tens ity w as abno rm a l w ithou t e v id ence o f a m ass les ion .

R ESU LTS . M R im ages show ed h ippocam pa l sc le ro s is in 3 0 ch Ild ren (57% ), o th er

les ions in 10 (1 9% ) (tum o rs in e igh t, c a ve rnous ang lom a in one , and ec top ic gra y

m atte r in one) , a nd no abnorm ality in 1 3 (24% ) (ln traobse rv e r ag re em en t: kappa =

0.77 and 0 .84 , In te robse rv er agre em ent: k appa = 0 .7 6 ). M R la te ra liza t lo n w as conco r-

d an t w ith ic ta l EEG in 36 (9 2% ) o f 39 ch ild ren . H ippocam pa l s c le ro s is w as b ila te ra l in

one ch ild and assoc ia ted w ith e x trah ippocam pal le s ions In n in e . H ippocam pa l sc le ro -

s is w as dete c ted on MR im ages o f 1 1 (8 5% ) o f 13 ch ild ren w ith patho log ic con firm a -

t io n o f h ippocam pa l sc le ro s is . B ene fic ia l resu lts w ere seen in 2 6 (9 0% ) o f 29 ch ild ren

w ho had tem po ra l lo b ec tom y .

C ONCLU S IO N . H ippocam pa l s c le ro s is is th e m ost com m on le s ion in ch ild ren w ith

in tra c tab le tem po ra l lo b e ep ile p sy , and it c an be dete c ted re liab ly and accu ra te ly on

MR im ages . W e suspec t th a t h ippocam pa l s c le ro s is is underd iagnosed in ch ild ren ,

poss Ib ly le ad ing to postponem en t o f su rg ery In ch ild ren w ith re frac to ry s e izu re s .

AJ R 1 99 3;1 61 :1 04 5-1 04 8

Tem po ra l lobe ep ilepsy is the m ost comm on fo rm o f in tra c tab le ep ilepsy in ch il-

d ren [1 ]. Tem po ra l lobec tom y is e ffe c tiv e in con tro llin g se izu res in app rox im ate ly

80% o f ch ild ren w ith in tra c tab le tem po ra l lobe ep ilepsy [2 -5 ]. A lthough h ippocam -

pa l s c le ros is is seen in 60-7 0% o f au top sy and su rg ic a l s e ries in a du lts w ith

re fra c to ry tem po ra l lobe ep ilepsy , it has been va riab ly repo rted in s tud ie s o f tem -

po ra l lobec tom y in ch ild ren [2 -7 ]. R ecen t repo rts [8 -14 ] ind ica te tha t h ippocam pa l

sc le ros is can be re liab ly de tec ted w ith M R im ag ing in 60 -100% o f pa tien ts ; how -

eve r, a num be r o f issues rem a in un reso lved . The re is no un ive rsa l ag reem en t on

the idea l m e thod o f M A im ag ing o f the h ippocam pus o r c r ite r ia fo r d iagnos ing h ip -

pocam pa l sc le ros is [1 3 , 1 4 ]. T he app lica tion to c lin ica l p rac tice o f quan tita tiv e MR

techn iques is unc lean [11 -14 ].

W e rev iew ed the M R im ages o f ch ild ren w ith w e ll-docum ented in tra c tab le tem -

pona l lobe ep ilepsy to de te rm ine the frequency o f h ippocam pa l sc le ros is and the

accu racy and re liab ility w ith w h ich h ippocam pa l sc le ros is cou ld be d iagnosed on

the bas is o f M R find ings .



M ateria ls and M ethods

F ifty -th ree co nse cu tive pa tie n ts o f the Ro ya l C h ild ren ’s H o sp ita l,

M elbou rne , w ith m ed ica lly re fra c to ry s e izu res and c lin ica l a nd ic ta l

EEG ev idence o f tem po ra l lobe ep ilepsy unde rwen t M R im ag ing

be tw een 1987 and 1 992 . The 27 boys and 26 g irls w e re 2 -1 7 yea rs

o ld (m ean , 10 years ) a tth e tim e o f im ag ing . The ch ild re n w ere id e n -

tif ied from a re v iew o f ic ta l EEG re co rd ing s and c lin ica l rec o rd s o f a ll

ch ild ren w ith p a r tia l se izu re s who underw en t v ide o -E EG m on ito ring

in th e depa rtm en t o f n e u ro lo g y . C h ild ren w e re in c lu d ed on ly if se i-

zune onse t w as c lea r ly lo ca lized to the tem po ra l reg ion and c lin ica ls e izu re cha rac te ris tic s we re cons is ten t w ith tem po ra l lobe onse t.

C h ild re n w ere exc lu ded if EEG abno rm a lit ie s w ere poor ly lo ca liz ed

o r e x tra tem pona l o r if c lin ica l s e izu re cha ra c te ris tics w ere inc on s is -

te n t w ith tem po ra l lo be on se t. Ic ta l s in g le -ph o to n em iss ion com -

p ute d to mo grap hy (S PE CT ) w ith 99 mT ch exam eth ylp ro py le ne am in e

ox im e was pe rfo rm ed du ring v ideo-EEG reco rd ing in 24 ch ild ren .

A t the end o f the s tudy pe riod , an te r io r tem po ra l lobec tom y had

been pe rfo rm ed in 28 ch ild ren and a les ion had been resec ted in

one . S e ven o f the se ch ild ren have been repo rted p re v io u s ly [2 ].

Mean age a t su rge ry w as 1 1 yea rs (range , 5- 1 8 yea rs ), a n d th e

m ean fo llow -up pe riod w as 22 m on ths (range , 1 m on th to 4 yea rs ).

R esu lts o f h is to log ic exam ina tions in 29 ch ild ren w e re rev iew ed by a

ped ia tric n e u ro pa th o log is t w ho d id n o t kn ow the pa tie n t’s id en tity o r

M R find ings . A fte r su rge ry , 21 ch ild ren w e re se izu re free on hav ing

aura s on ly , th ree ra re ly h ad se iz u re s , an d tw o had few er se izu re s

th an be fo re . S e izu re fre quen cy was un changed in o ne ch ild , a nd

one ch ild w ith R asm ussen ’s sy nd rom e d ie d .

M R im ages w e re ob ta ine d w ith a 0 .3 -T iro n co re re s is tive m agne t

(B 3000 , F onan , M e lv ille , N Y ) in 28 ch ild ren , w ith a 1 .5 -T s up er co n-

d u c tin g m agne t (M agne tom , S iem ens AG , En la ngen , G e rm any ) in

18 ch ild re n , a n d w ith b o th u n its in se ven . F iftee n ch ild re n had m u lti-

p ie s tu d ies . W ith th e 0 .3 -T un it, ge ne ra lly co rona l inv e rs io n -reco ve ry

(1500 /30 /500 [TR /T i/T E ]) and ax ia l and co rona l sp in -echo (2000 /

35 ,85 [TR ITE ]) im ages w e re ob ta ined . A 5 .1 -mm slice th ic kness ,

0 .5 -mm in te rs lice g ap , 1 9 2 x 256 m atr ix , and 24 .5 -cm fie ld o f v iew

w ere used . W ith the 1 .5 -T un it, u sua lly m agne tiz a tion -p repa red

ra p id g ra d ie n t-ech o im ages (M P RAG E , 1 0 /4 /1 0# {1 76 }TR/TE/f l ip

ang le ]) w e re ob ta ine d b y us in g a 180 x 256 m atr ix , 25 -cm nec ta ngu-

ar fie ld o f v iew , 2 .8 -m m s lice th ickn es s , an d no in tens lic e g ap . A x ia l

and co rona l sp in -e cho im ages (2 750 /20 ,80 ) w e re ob ta ine d b y us in g

a 7 5# {1 76 }lip ang ie , on e ex c ita tion , 4 -mm s lice th ickn e ss , 1 -mm in te r -

s lic e ga p , 1 92 x 256 m atrix , and 23 -cm fie ld o f v iew . C orona l im ages

we re pu lse ga ted . W ith bo th un its , a x ia l im ages w e re pa ra lle l and

co ro na l im age s w e re pe rp end icu la r to the lo ng a x is o f th e h ipp o c -

am pus , de fined on a sag itta l lo ca liz er.

B ecause o f rap id advances in MR techno logy du ring the cou rse

o f th e s tud y , im ages w ere o b ta ine d w ith tw o d iffe ren t m ach ine s , a n d

consequen tly the qua lity o f the im ages va ried . In add ition , im ag ing

p ro toco ls w e re m od ified as so ftw a re capab ilitie s changed . Tech-

n iq u es fo r q u an tita tive a na lys is w ere no t a va ilab le a t the tim e o f th e

s tu d y . Y oung ch ild re n w ere seda te d w hen neces sa ry , un de r su pen v i-

s io n o f a p ed ia tric an e s the tis t.

M R im age s we re re v iew ed ind ependen tly b y a ped ia tric na d io lo -

g is t and a neu ro rad io log is t. E ach M R study was rev iew ed tw ice by

each rad io log is t. Im ages w ere rev iew ed in random o rde r and w ithou t

know ledge o f the pa tien t’s iden tity , c lin ica l and EEG da ta , the nad io l-ogist’sfirstinterpretation,or the other radiologist’sinterpretation.MR

im age s o f ch ild re n w ho had m u ltip le exam in a tio n s w ere re v iew ed

toge the r. A ll sequences and p lanes o f im ag ing we re rev iew ed , bu t

pa rticu la r a tte n tion was d ire c ted to th e m ed ia l tem po ra l s tru c tu res .

Im ages w e re c la ss ifie d a s show in g h ip p ocam pa l sc le ro s is , o th e r

les io ns , o r no abno rm a lity . A bno rm a lit ies w e re c las s if ied as le ft

s id ed , rig h t s ide d , on b ila te ra l. H ipp ocam pa l s c le ro s is w a s d ia gnosed

when e v id en ce o f h ip p ocam pa l a trop hy w as p re sen t o r h ipp ocam pa l

s ig na l in ten s ity w a s abno rm a l w ith ou t e v ide nce o f a m ass les io n .

W hen d iffe re n t d iag nose s w e re m ade b y one ra d io log is t, the M R

im ages w e re rev iew ed once m o re and a fina l d iagnos is was m ade .

W hen d iffe ren t d iagnoses w e re m ade by tw o r ad io lo gis ts , th e im a ge s

we re rev iewed in con junc tion and a consensus d iagnos is was m ade .

E th ic a l c on s id e ra tio ns p rec lud ed th e use o f he a lthy ch ild re n as

con tro l su b je c ts , a nd p rac tica l p rob lem s w ith b lin d ing p re c lud ed th e

u se o f ch ild re n w ith e x tra tem pona l le s io ns a s con tro l su b je c ts .

C hance -ad ju s ted in tenobse rve r and in traobse rve r ag reem en t was

m easu red w ith C ohen ’s kappa s ta tis tic .

Resu l ts

The consensus M R d iagnoses fo r the 53 ch ild ren we re

h ippocam pa l sc le ros is in 30 (57% ), o the r le s ions in 10

(19% ), and no abno rm a lity in 13 (24% ). In traobse rve r and

in tenobse rven ag reem en t w as good (in traobse rven kappa ,

0 .77 and 0 .84 ; in te robse rver kappa , 0 .76 ). In tenobse rve r d is -

ag reem en t conce rned e igh t ch ild ren in whom the u ltim a te

d iagnos is was h ippocam pa l sc le ros is o r no abno rm a lity . S ix

o f the e igh t had M R s tud ie s w ith the 0.3 -T un it. U n ila te ra l

abno rm a litie s o f the tem po ra l co rte x w e re d iagnosed on M R

im ages in 39 ch ild ren , conco rdan t w ith the EEG and ic ta l

S PEC T la te ra liza tion in 36 ch ild ren (92% ).

H ippocam pal sc le ros is w as un ila te ra l in 29 ch ild ren and

b ila te ra l in one (F ig s . 1-3 ). Tw en ty -e igh t ch ild ren had h ip -

pocam pal a trophy and increased s igna l on T2 -w e igh ted

im ages; tw o ch ild ren had m a rked h ippocam pa l a trophy w ith -

ou t obse rvab le s igna l abno rm a lity . B ila te ra l h ippocam pa l

sc le ros is was d iagnosed on the bas is o f b ila te ra l h ippocam -

pa l a trophy and inc reased s igna l in tens ity (F ig . 3 ). In n ine

ch ild ren , h ippocam pa l sc le ros is w as assoc ia ted w ith le s ions

ou ts ide the tem po ra l co rte x : anachno id cys t in the m idd le

cran ia l fo ssa in tw o (one ips ila te ra l and one con tra la tena l to

h ippocam pa l sc le ros is ), pen iven tn icu lan h igh s igna l in tens ity

in tw o w ho had ven tn ic u lopen itonea l shun ts , ip s ila tena l fo ca l

o cc ip ita l a trophy in three , ip s ila te ra l pa raven tn icu la r tum o r

pro jec ting in to the tn igo ne in one , a n d gene ra lized ce reb ra l

a trophy in one . M R d iagnoses o f the o the r le s ions in the

tem po ra l co rte x w e re tum o r in e igh t, c ave rnous ang iom a in

one , and hem im acrencepha ly w ith ip s ila tena l e c top ic g ray

m a tte r in one .

O f the 17 ch ild re n w ho had su rge ry and in whom h ippo -cam pa l s c le ros is was d iagnosed on the bas is o f M R find ings ,

h is to log ic exam ina tion show ed h ippocam pa l s c le ros is in 11

and no abno rm a lity in s ix in w hom app rop ria te h ippocam pa l

tissue was no t p resen t in the su rg ica l spec im en . O f the s ix

ch ild ren w ho had su rge ry and in w hom a tum o r w as d iag -

nosed on the bas is o f MR find ings , h is to log ic exam ina tion o f

the ava ilab le spec im en show ed dysem b ryop las tic neu roep i-

the lia l tum o r in tw o , gang liog liom a in one , as tro cy tom a in

one , ch ron ic encepha litis in one , and no abno rm a lity in one .

T he cave rnous ang iom a and g ray m a tte r he teno top ia w e re

h is to log ica lly con firm ed . O f the fou r ch ild ren who had su rge ry

and in whom no abno rm a lity was d iagnosed on the bas is o f

M R find ings (a ll fou r im aged w ith the low -fie ld -s treng th un it) ,

h is to log ic exam ina tion showed h ippocam pa l s c le ros is in tw oand no abno rm a lity in tw o in whom app rop ria te h ippocam pa l

tissue w as no t p resen t in the su rg ica l spec im en . M R im ages

show ed ev idence o f h ippocam pa l s c le ros is in 11 (85% ) of 13

ch ild ren w ith h is to log ica lly p roved h ippocam pa l sc le ros is .

Discuss ion

Th is s tudy repo rts the M R find ings in a g roup o f ch ild ren

w ith we ll-docum en ted and in tra c tab le tem po ra l lobe ep i-

lepsy . T he d iagnos is o f tem po ra l lobe ep ilepsy w as based on

typ ica l c lin ica l se izu re cha rac te ris tic s and ic ta l EEG loca liz a-

tion to the tem po ra l lobe . A lthough in tra c ran ia l e le c trodes



A JR :161 , N ovem be r 1993 H IPPOCAM PAL SC LEROS IS IN CH ILDREN 1047

F ig . 1 . -1 2 -y e a r- ol d bo y w ith r igh t-s ided h ip po cam pal s c le ro s is .

A , Ang le d co ro na l M P RAG E (1 0 I4 l1 0 M R Im age show s m a rked h ip pocam pal a trop hy ( a r row ) on r igh t s id e . C om pa re w ith n orm a l h ipp ocam pu s on

le f t bo unded sup er io r ly by a lv eu s .

B an d C , C or on al p ro to n d en sity -w eig hte d (B ) an d T2 -w e lgh ted (C ) (2 50 0 /3 0 ,8 0 ) M R Im ages sh ow the a re a o f Inc re as ed s ign a l In te ns ity ( a r rows) is

local ized to h ipp ocam pu s on r igh t s ide .

F ig . 2 .-9 -ye ar-o ld g ir l w ith le ft-s id ed h ip -

po c ampa l sc le ros is .A an d B, A ng led ax ia l (A ) a nd c or on al (B ) T2 -

we igh ted ( 2500 /80 ) MR Im ages show poorly d e -fin ed are a o f in cre as ed s ign a l In ten s ity

( a r rows) e x ten d in g a lo n g h lpp ocam pus . H ip -

po c ampa l a tro ph y w as be tte r v isua liz ed on M P

R AG E c oro na l s eq ue nce (n ot s ho wn ).

F ig . 3 .-7 -ye ar-o ld g irl w ith b ila te ra l h ip po cam pal sc le ros is .

A, A ng led co ron a l M P RAG E (1 0 I4 I1 0 M R Im ag e show s b ila te ra l h ip po cam pal a tro ph y ( a r rows) , more ma rk ed on le f t s ide than on righ t s ide .

B an d C , P r ot on d e ns it y- we ig h te d (B ) an d T2-we igh ted (C ) ( 2500 /30 ,80 ) MR images show a rea s o f Inc rea sed s ign a l In ten s ity ( a r rows) loc a lize d to

h lp po cam pu s b ila te ra l ly . In de penden t b ila te ra l tem po ra l lob e hype rp erfus ion w as seen on Ic ta l S P EC T Im ages (n ot s ho wn ).



we re no t used , EEG loca liza tion w ith sca lp e le c trodes w as

suppo rted by the resu lts o f ic ta l SPEC T in 21 ch ild ren and

the bene fic ia l re su lts o f tem po ra l lobec tom y in 26 ch ild ren .

A bno rm a litie s on M R im ages w e re de tec ted in 75% o f ch il-

d ren , h ippocam pa l sc le ros is be ing the m ost comm on les ion .

EEG con firm a tion o f la tena liza tion w as p re sen t in 92% .

O ve ra ll ag reem en t be tween the two obse rve rs w as good ,

ind ica ting tha t h ippocam pa l s c le ros is and o the r le s ions can

be de tec ted re liab ly on M R im ages o f ch ild ren .

H ippocam pa l sc le ros is was d iagnosed on the bas is o f M Rfind ings in 57% o f ch ild ren in ou r s tudy and was con firm ed

pa tho lo g ica lly in a ll ch ild re n w ho had su rge ry in w hom

app rop r ia te h ippocam pa l tissue w as ava ilab le fo r exam ina -

tion . T he m e thod o f m es ia l tem po ra l re sec tion a t ou r in s titu -

tion (subp ia l a sp ira tion as opposed to en b loc resection )

m ean t tha t adequa te h ippocam pa l tis sue fo r pa tho log ic

exam ina tion was ava ilab le in on ly 11 o f 17 ch ild ren w ho had

M R evid ence o f h ippocam pa l sc le ros is . H ippo cam pa l sc le ro -

s is is de tec ted on M R im ages on at su rge ry in 60 -70% o f

adu lts w ith in tra c tab le tem po ra l lobe ep ilepsy [10 , 13 ]. T he

p reva lence o f h ippocam pa l sc le ros is in ch ild ren w ith in tra c-

tab le tem po ra l lobe ep ilepsy can be de te rm ined on ly from

su rg ica l se r ie s , w he re it has been va riab ly repo rted in 9 -

60% [2 -6 ]. The s tud ie s show ing few e r ch ild ren w ith h ippo -cam pa l sc le ros is m ay re fle c t the absence o f h ippocam pa l

tissue in th e ir su rg ica l spec im ens, a se lec tion b ias tow ard

ea rly su rge ry w hen im ag ing show s a m ass le s ion , on a d iffe r-

en t spec trum o f tem po ra l lobe abno rm a litie s in ch ild ren w ho

have in tra c tab le tem po ra l lobe ep ilepsy in ear ly ch ildhood .

The repo rted M R fea tu res o f h ippocam pa l sc le ro s is

in c lude h ippocam pa l a trophy , in c reased s igna l in tens ity on

T2-w e igh ted im ages, loss o f in te rna l a rch itec tu re o f the h ip -

pocam pus, and loss o f s igna l in tens ity on T i -w e igh ted

im ages [13 , 14 ]. In ou r s tudy , on ly the firs t tw o we re used as

c rite ria fo r a d iag nos is o f h ip p ocam pa l sc le ros is . In te re s t-

in g ly , in c reased s igna l in tens ity on T2 -we ighted im ages w as

seen in 28 o f 30 ch ild ren w ith h ippocam pa l sc le ros is ; th e firs t

echo o f the doub le -echo co rona l sequence was pa rticu la r ly

use fu l in m a rg ina l cases . Th is is in con tra s t to som e stud ie s

in adu lts , in w hom dem ons trab le s igna l abno rm a litie s w e re

ra re [1 5 ]. A recen t s tud y w ith T2 quan tita tive m easu rem en t

has show n in crea sed s igna l in tens ity in pa tien ts w ith in trac-

ta b le tem pora l lo be ep ile psy when no s igna l abnorm a lity

w as v is ib le (Jackson GD et a l., p re sen ted a t the Am erican

Ep ile psy Soc ie ty m ee ting , Decem be r 1 992). It is p oss ib le

tha t s igna l abno rm a litie s assoc ia ted w ith h ippocam pa l s c le -

ros is a re m o re p ronounced in ch ild re n o r tha t lo ss o f s ig na l

in tens ity occu rs w ith in c reas ing age .

In ou r s tudy , n ine ch ild ren w ith h ippocam pa l sc le ros is had

add itiona l le s ions ou ts ide the co rte x o f the tem pora l lobe ,

ra is ing doub t abou t lo ca liza tion o f the ep ilep togen ic fo cus .

A rachno id cys ts and pen iven tn icu la r w h ite m a tte r g lio s is a re

un lik e ly to cause se izu res by them se lves . H ow eve r, co rtica l

a trophy o r tum o r in the pos te r io r co rte x can m an ife s t e le c tn i-

ca lly and c lin ic a lly as tem po ra l lobe ep ilepsy . The ch ild ren

w ith d isco rdan t M R and EEG abnorm a lities and th e ch ild ren

w ith b ila te ra l se izu re on se t ha ve no t undergone su rge ry .

No les ion w as de tec tab le on MR im ages o f 13 ch ild ren ,

seven o f w hom we re exam ined w ith the h ighe r-fie ld -s treng th

un it. In fo u r c h ild ren w ith n o rm a l M R find ings , tem pora l

lo bec tom y w as pe rfo rm ed because o f re su lts o f e le c trophys -

io log ic s tud ie s . These fou r ch ild ren w e re exam ined on the

low -fie ld -s treng th m agne t ea rly in the s tudy period . T he qua l-

ity o f these im ages wou ld no t be accep tab le now . H ippocam -

pa l s c le ros is was found a t pa tho log ic exam ina tion in tw o

ch ild ren , and no abno rm a lity w as found in tw o ch ild ren in

w hom app rop ria te h ippocam pa l tissue w as no t ava ilab le fo r

exam ina tion . S om e ch ild ren w ith no rm a l MR find ings had

m ino r deg rees o f h ippocam pa l asymm e try , bu t in the

absence o f an app rec iab le in crease in h ippocam pa l s igna l

on T2 -we igh ted im ages , h ippocam pa l sc le ros is w as no t

d iagnosed . Q uan tita tive te chn iques m ay have led to a d iag -

nos is o f h ippo cam pa l sc le ros is in som e o f the se ch ild ren . Itis a lso poss ib le tha t som e o f the ch ild ren w ith no rm a l M R

fin d ings who have no t unde rgone su rge ry have sub tle m ig ra -

tiona l abno rm a litie s on m icrodysgenes is o f the tem po ra l con -

tex , w h ich is no t de te cta b le w ith MR .

A ll d iagnoses o f h ippocam pa l s c le ros is based on M R

im ages we re con firm ed pa tho log ica lly if app rop r ia te tis sue

was ava ilab le fo r exam ina tion . O u r resu lts sugges t tha t h ip -

pocam pa l sc le ros is is as p reva len t in ch ild ren , and as read ily

d iagnosed on the bas is o f M R find ings , as in adu lts . W e sus-

pect tha t h ippo cam pa l sc le ros is is underd iagnosed in ch il-

d ren , poss ib ly lead ing to pos tponem en t o f su rge ry in

ch ild ren w ith in tra c tab le se izu res .

A C KNOWLEDGMENTS

W e thank Pe te r B uchanan fo r adm in is tra tive ass is tance ; E ric G il-

fo rd , M ichae l K ean , and the M RI s ta ff a t S t. V in cen ts H osp ita l, M el-

bounne , fo r im ag ing ou r pa tien ts ; Ian J . H opk in s , L lo yd K . S h ie ld ,

K ev in J . C o llin s , and G eo ffe ry L . K Iug fo r con tribu tions to pa tien ts ’

tre a tm en t; an d B ria n M . T re ss , Ian J . H op k in s , a nd Sam F . B e rko v ic

fo r rev iew in g th e m anu sc rip t a n d p ro v id in g he lp fu l su gge s tio ns .

REFERENCES

1 . A ica rd i J . E pile ps y in ch ild re n. N ew Y or k: R av en , 1 98 6:1 40 -1 75

2 . H op k in s IJ , K Iu g G L. T em po ra l lob e c tom y fo r th e trea tm en t o f in trac tab le

com ple x pa r tia l s e izu res o f tem po ra l lo be o rig in in e a r ly ch ildh ood . 0e v

M ed C h ild N e uro l 199133 :26 - 31

3 . A de lso n P D , P eacoc k W J, C hugan i H T , e t a l. Tem pora l and e xte nd ed

tem po ra l resec tion s fo r th e tre a tm en t o f in trac ta b le se iz u res in e a riy ch ild -h o o d . P ed ia tr N eu ro su rg 1992 ;18 :169 -178

4 . D uchow ny M , L ev in B , J aya kan P , e t a l. T em pora l lo be c tom y in e a rl y c h il d-

hood . Ep i leps ia 1992 ;33 :298 -303

5 . D rake J , H o ffm an H J, Koba yash i J , Hw ang P , B ecke r LE . Su rg ica l m an -

agem en t o f ch ild re n w ith tem po ra l lob e ep ile ps y and m ass les ion s . N e u -

rosurger , ’ 1987 :21 :792 -797

6. D av idson S , F alco nen M A. O utcom e o f su rg e ry in 40 ch ild re n w ith tem po -

r al l ob e e p il ep sy . Lance t 19 75;1 :12 60 -1 2 63

7 . B abb T L, B row n W J. P ath o lo g ic fin d in gs in ep ile psy . in : En ge l JJ , e d . Sur -

g ic a l tre atm en to f th e ep ile ps ies . N ew Y ork : R av en , 1 98 7:5 11 -54 0

8 . B ro nen R A. Im ag ing find in g s in h ip po cam pa l sc le ros is : c o rre la tion w ith

pa tho logy . AJ N R 1991 :12 :933 -940

9 . Ja ckson G D, Be rko v ic SF , T re ss BM , K aln ins R M, Fab iny i G C, B la d in PF .

H ip po cam pa l sc le ros is ca n b e re lia b ly d e tec ted by m agne t ic re son an ce

imag ing . Neuro logy 1990 ;40 :1869-1 87 5

10 . B erko v ic S F , Ande rm an F , O liv ie r A , e t a l. H ip po cam pa l sc le ros is in tem -

p en a l lo be ep ilep sy d em on stra te d b y m agne tic resona nce im ag in g . An nNeuro l 1991 :29 :17 1 8 2

11 . Ja ck CR J, Sharb ro ug h FW , Tw om ey C K , e t a I . Tem pora l lob e se iz u res :

la te ra l iza tio n w ith M R vo lum e m ea su rem en ts o f the h ip po cam pa l fo rm a-

t ion. Rad io logy 1990 ;175 :423 -429

1 2. A sh ta ri M , B arr W B, S chau l N , B ogerts B . T hree -d im ens io na l low -a ng le

s ho t im ag ing and com pu te rize d vo lum e m easurem en t o f th e h ip po cam -

pu s in pa tie n ts w ith ch ro n ic ep ilep sy o f th e tem pora l lob e . AJ N R 1991 :

12 :941 -947

13 . B ro nen R A. E pile p sy : the ro le o f M R im ag in g . AJ R 1992 ;159 :1165-1174

14 . J ac kson G D, B erk ov ic S F, D un can JS , C onne ily A . O ptim iz in g the d ia gno -

s is o f h ipp ocam pa l sc le ros is us ing M R im ag ing . AJ N R 1993 :14 :753 -762

15 . C a sc ino G D , Ja ck C R , P aris i JE , e t a l. M agne tic re so nan ce im ag in g -

b as ed v olu me s tu die s in t em p or al l ob e e pi le ps y: p at ho lo gic al c or re la tio ns .

A nn N euro l 1991 :30 :31 - 36

Documents

Hippo Cam Pal Sclerosis in Children With Intractable Temporal Lobe Epilepsy- Detection With MR Imaging