A cross-sectional MRI study of brain regional atrophy and clinical characteristics of temporal lobe epilepsy with hippocampal sclerosis

Page 1

E

Aaw

SSM

a

b

c

d

e

RA

0d

pilepsy Research (2012) 99, 156—166

journa l homepage: www.e lsev ier .com/ locate /ep i lepsyres

cross-sectional MRI study of brain regional atrophynd clinical characteristics of temporal lobe epilepsyith hippocampal sclerosis

aud Alhusaini a,b, Colin P. Dohertye, Cathy Scanlonb, Lisa Ronanb,inead Maguired, Gabor Borgulyaa, Paul Brennand, Norman Delantya,c,ary Fitzsimonsb, Gianpiero L. Cavalleri a,∗

Department of Molecular and Cellular Therapeutics, The Royal College of Surgeons in Ireland, Dublin 2, IrelandDepartment of Neurophysics, Beaumont Hospital, Dublin 9, IrelandDepartment of Neurology, Beaumont Hospital, Dublin 9, IrelandDepartment of Radiology, Beaumont Hospital, Dublin 9, IrelandDepartment of Neurology, St. James’s Hospital, Dublin 8, Ireland

eceived 18 August 2011; received in revised form 9 November 2011; accepted 13 November 2011vailable online 23 December 2011

KEYWORDSMesial temporal lobeepilepsy;Quantitative MRI;Endophenotype

SummaryPurpose: Applying a cross-sectional design, we set out to further characterize the significanceof extrahippocampal brain atrophy in a large sample of ‘sporadic’ mesial temporal lobe epilepsywith hippocampal sclerosis (MTLE + HS). By evaluating the influence of epilepsy chronicity onstructural atrophy, this work represents an important step towards the characterization ofMRI-based volumetric measurements as genetic endophenotypes for this condition.Methods: Using an automated brain segmentation technique, MRI-based volume measurementsof several brain regions were compared between 75 patients with ‘sporadic’ MTLE + HS and50 healthy controls. Applying linear regression models, we examined the relationship betweenstructural atrophy and important clinical features of MTLE + HS, including disease duration,lifetime number of partial and generalized seizures, and history of initial precipitating insults(IPIs).

Results: Significant volume loss was detected in ipsilateral hippocampus, amygdala, thalamus,and cerebral white matter (WM). In addition, contralateral hippocampal and bilateral cere-bellar grey matter (GM) volume loss was observed in left MTLE + HS patients. Hippocampal, amygdalar, and cerebral WM volume loss correlated with duration of epilepsy. This correla-tion was stronger in patients with prior IPIs history. Further, cerebral WM, cerebellar GM, andcontralateral hippocampal volume loss correlated with lifetime number of generalized seizures.

∗ Corresponding author. Tel.: +353 01 4022146; fax: +353 01 4022453.E-mail address: [email protected] (G.L. Cavalleri).

920-1211/$ — see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.eplepsyres.2011.11.005

Page 2

Quantitative MRI study and clinical characteristics of epilepsy 157

Conclusion: Our findings confirm that multiple brain regions beyond the hippocampus are involvedin the pathogenesis of MTLE + HS. IPIs are an important factor influencing the rate of regionalatrophy but our results also support a role for processes related to epilepsy chronicity. Theconsequence of epilepsy chronicity on candidate brain regions has important implications on

ndophenotypes.s res

osefpKmacadcai

castcte

M

S

Wrcrim‘

S

PPBcsThit(other than hippocampal sclerosis were excluded from the study.In total, 75 patients comprising 35 males and 40 females (meanage ± SD = 36.6 year ± 11.5) were included. Side of seizure focus

their application as genetic e© 2011 Elsevier B.V. All right

Introduction

Hippocampal sclerosis (HS) is the most common lesion asso-ciated with temporal lobe epilepsy (Bruton, 1988; Wieser,2004; Berg et al., 2010). HS is characterized pathologicallyby cellular loss and synaptic reorganization in hippocam-pal sub-regions and typically diagnosed through magneticresonance imaging (MRI) by increased hippocampal T2 sig-nal, and hippocampal atrophy (Berkovic et al., 1991).While MRI-defined ipsilateral hippocampal atrophy is a well-documented feature of mesial temporal lobe epilepsy withhippocampal sclerosis (MTLE + HS) (Bernasconi et al., 2003;Garcia-Finana et al., 2006; Labate et al., 2008), struc-tural atrophy has also been identified in other subcorticalregions including the amygdala (Van Paesschen et al., 1996;Bernasconi et al., 2003, 2005), thalamus (Dreifuss et al.,2001; Natsume et al., 2002; Labate et al., 2008; McDonaldet al., 2008), and basal ganglia (Decarli et al., 1998; Dreifusset al., 2001; Natsume et al., 2002). Further, atrophy in thecerebellum (Sandok et al., 2000; Hermann et al., 2005;McDonald et al., 2008) and multiple cortical regions havealso been described (Doherty et al., 2003; Bernasconi et al.,2004; Seidenberg et al., 2005; Mueller et al., 2006). Whetherthe observed atrophy across the hippocampus and otherregions is progressive and already present at the time ofseizure onset, or develops as a result of damage secondaryto recurrent seizure activity is a matter of ongoing debate.Further, the meaning and significance of extrahippocampalatrophy remains poorly understood.

Patients with MTLE + HS often report a prior history ofearly brain insult, such as febrile seizures, head trauma,or intracranial infection(s) (French et al., 1993; Wieser,2004). Such early injuries have been suggested to initi-ate progressive hippocampal atrophy (Meyer et al., 1954;Van Landingham et al., 1998; Mathern et al., 2002; Scottet al., 2003; Liu et al., 2005). Further, evidence from animalmodels suggests that hippocampal neural loss can occur fol-lowing severe seizure activity (Cavazos et al., 1994; Pitkänenet al., 2002). In humans, hippocampal neural loss has beenshown to occur following acute episodes of status epilepticus(Weishmann et al., 1997). Cumulative hippocampal damagesecondary to recurrent seizure activity has been supportedby some (Tasch et al., 1999; Briellmann et al., 2001, 2002;Kälviäinen et al., 1998; Kälviäinen and Salmenperä, 2002;Fuerst et al., 2003; Salmenperä et al., 2005), but not all(Cendes et al., 1993; Moran et al., 2001; Andrade-Valencaet al., 2003; Holtkamp et al., 2004; Liu et al., 2002, 2005)longitudinal and cross-sectional MRI studies. The discor-dant results from such MRI-based studies probably reflect

the complexity of epilepsy in that studies often differedin both the clinical characteristics of the patient popula-tion and the quantitative methods used to analyze the MRIdata.

(r‘s

erved.

Characterizing brain regional atrophy not only improvesur understanding of epilepsy generally, but can aidpecifically in genetic studies of the condition. MRI-basedndophenotypes have been proposed as a viable approachor improving the power of genetic mapping studies of com-lex forms of epilepsy, such as MTLE + HS (Helbig et al., 2008;asperaviciute et al., 2010). An important step in deter-ining the suitability of volumes of candidate brain regions

s endophenotype in epilepsy is the characterization of theontribution of disease chronicity to the observed regionaltrophy. Volumes of brain regions that are susceptible toamage by recurrent seizure activity (or antiepileptic medi-ations) will show lower heritability in patients with epilepsynd thus are less suitable as endophenotypes when mappedn healthy populations.

In this study, we set out to (i) build on previousross-sectional and longitudinal MRI studies to further char-cterize extra-hippocampal atrophy specifically in a largeample of patients with ‘sporadic’ MTLE + HS and (ii) quan-ify the degree to which clinical features of MTLE + HSorrelate with regional atrophy in an effort to determinehe nature of the association between regional atrophy andpilepsy chronicity.

ethods

tudy design

e compared MRI-based volumetric measurements of several brainegions between 75 ‘sporadic’ MTLE + HS cases and 50 healthyontrols. Focusing on brain regions showing significant volumeeductions in patients compared to healthy controls, we stud-ed the relationship between brain regional atrophy (as ‘‘outcomeeasure’’) and various clinical characteristics of the condition (as

‘explanatory variables’’).

tudy participants

atientsatients with a clinical diagnosis of MTLE + HS were recruited fromeaumont Hospital and St. James’s Hospital, both tertiary epilepsyentres in Dublin, Ireland. All patients underwent a comprehen-ive investigation that confirmed clinical features of MTLE + HS.he side of seizure activity focus was determined by a compre-ensive evaluation including a combination of seizure semiology,ctal and inter-ictal EEG/video-telemetry recordings, and qualita-ive inspection of MRI films for evidence of hippocampal sclerosisHS) by a neuroradiologist. Patients with evidence of any lesion

and hippocampal sclerosis) was on the left in 34 patients, on theight in 33 patients and bilaterally in 8 patients. All cases weresporadic’ and reported no family history of epilepsy or febrileeizures. Initial precipitating insults (IPIs) were reported by 46

Page 3

1

c(pwKah

COmk

af

M

MGTseflr

M

AfhrwcFeua2oM

Mm

Tepva

ugd

C

CCuutceeo

sv(ESpfpatM

DFsi(yap

vt‘‘smmMrs

bptMb

taps

a

R

M

HChuMt(olwtS

58

ases (61%). The following were considered as IPIs: febrile seizuresFS), intracranial infection(s) and status epilepticus. We excludedatients with head trauma history. Sixty six of our patients (88%)ere classified as refractory to medical treatment as defined bywan et al. (2010). Thirty-four of the 75 patients (45.3%) underwentselective amygdalohippocampectomy. In each of these 34 cases,

istological evidence of hippocampal sclerosis was confirmed.

ontrolsur control population consisted of 50 individuals comprising 25ales and 25 females (mean age ± SD = 28.4 year ± 5.1) with no

nown neurological or psychiatric illness.The research ethics committees of both hospitals independently

pproved this study and written informed consent was obtainedrom all participants.

R image acquisition

RI scans of the brain were acquired on all participants using 1.5 TE Signa scanner at Beaumont hospital. A three-dimensional (3D)1-weighted spoiled gradient (SPGR) sequence was acquired from aagittal localizer in the coronal plane. The following imaging param-ters were used: TR = 10.1, TE = 4.2, TI = 450 ms, one excitation,ip angle = 20.0, field of view = 24 cm × 24 cm, matrix = 256 × 256,esulting in 124 × 1.5 mm-thick image slices.

R image processing

ll MR images were processed using FreeSurfer, aully automated image analysis package (version 4.50,ttps://surfer.nmr.mgh.harvard.edu). FreeSurfer provides corticaleconstruction and volumetric segmentation of the MR images. Itas used to segment and produce volume measurements of theortex, cerebral WM, cerebellum, and subcortical structures. ThereeSurfer process has been described in detail previously (Dalet al., 1999; Fischl et al., 1999a,b, 2001, 2002, 2004) and hasndergone extensive investigations to assess its accuracy, validitynd applicability (Rosas et al., 2002; Han et al., 2006; Lee et al.,006; Dickerson et al., 2008; Morey et al., 2010). Additional detailsf how Freesurfer was applied are available in the Supplementaryethods.

RI-based volumetric measurement (‘‘outcomeeasures’’)

he volumes of the following subcortical structures were consid-red in this study: the hippocampus, thalamus, amygdala, putamen,allidum, and caudate in each hemisphere. We also explored theolume of total cerebral and cerebellar grey and white matters (GMnd WM).

All volumes were adjusted for age and intracranial volumes (ICV)sing a linear regression model equation derived from the controlroup (Free et al., 1995). See Supplementary Methods for furtheretails.

linical characteristics (‘‘explanatory variables’’)

ollection of clinical informationlinical information was collected from patients and their relativessing structured interviews at the time of scanning or through followp telephone calls. Interviews were supplemented with informa-ion collected from medical records when necessary. The following

linical variables were collected: age at onset of seizures, dis-ase duration, history of initial precipitating insults (IPIs), andstimates of the lifetime number of partial seizures and partial-nset seizures that evolved to generalized convulsive events. Partial

VWp

S. Alhusaini et al.

eizures and partial-onset seizures that evolved to generalized con-ulsive events [i.e., secondary generalized tonic-clonic seizuresSGTCS)] were defined according to the International League Againstpilepsy (ILAE) guidelines (Commission, 1981; Berg et al., 2010).ee Supplementary Methods for further details. The number ofartial seizures and SGTCS were calculated through patient andamily member reports and seizure diaries where available. Com-lete details of lifetime number of both types of seizures were onlyvailable on 50 patients. A copy of the actual questionnaire usedo collect the above information is available in the Supplementaryaterial.

ata analysis and statisticsirst, in order to detect regional atrophy that is related to the side ofeizure activity we divided the patients into three groups accord-ng to the side of seizure focus (i.e., side of HS): left MTLE + HSn = 34), right MTLE + HS (n = 33) and bilateral MTLE + HS (n = 8). Anal-sis of covariance (ANCOVA) was applied to compare ICV and agedjusted regional volumes within each of these three MTLE + HSatient groups to the control group (n = 50).

Next, we focused specifically on structures showing significantolume reduction (i.e., regional atrophy) in MTLE + HS patientso explore the relationship between regional atrophy as an‘outcome measure’’ and the clinical features described above as‘explanatory variables’’ using univariate and multivariate regres-ion models. For this analysis, we re-assembled regional volumetriceasurements for each patient into ipsilateral and contralateraleasurements according to the side of seizure focus. Further,ann—Whitney U test was used to compare the patient groups in

espect to epilepsy duration, age at seizure onset and frequency ofeizures.

The association between IPIs and regional atrophy was exploredy comparing the volumes of atrophied regions in patients with aositive history of IPIs (n = 46; 61.3%) to those patients with nega-ive history (n = 29; 38.7%) using ANCOVA. Further, we applied theann—Whitney U test to compare age at onset of habitual seizuresetween patients with a positive and negative history of IPIs.

The relationship between SGTCS and regional atrophy was fur-her investigated using ANCOVA where volumes of regions showingtrophy in patients who experience SGTCS (31 SGTCS + ve of 50atients; 62%) were compared to those who only experience partialeizures (19 SGTCS − ve of 50 patients; 38%).

All statistical analyses were performed using SPSS statisticalnalysis software (version 18.0).

esults

RI-based volumetric analysis

ippocampal volumeomparing patients with healthy controls, we found aighly significant reduction in ipsilateral hippocampal vol-me in patients with left (p < 0.0001) and right (p < 0.0001)TLE + HS. In addition, bilateral hippocampal volume reduc-

ion was detected in patients with bilateral MTLE + HSp < 0.0001). These results were expected given the previ-us qualitative MRI-based classification of our patients intoeft, right, and bilateral MTLE + HS. The group of patientsith left MTLE + HS also showed a significant reduction in

he volume of the contralateral hippocampus (p < 0.001).ee Fig. 1.

olumes of other subcortical structureshen the volumes of subcortical structures in MTLE + HS

atients other than the hippocampus were compared to

Page 4

Quantitative MRI study and clinical characteristics of epilepsy

Figure 1 A comparison of hippocampal volumes betweenMTLE + HS patients and healthy controls. Left (n = 34), right(n = 33), and bilateral (n = 8) MTLE + HS patients were comparedto healthy controls (n = 50). Volumes (adjusted for age and ICV)are reported as z-scores derived from the mean of the con-trols data. Error bands represent 95% confidence intervals (CI).**p < 0.001; ***p < 0.0001.

tvdp(oaeeieptp

VmWl(Bir

Figure 2 A comparison of amygdalar, thalamic, cerebral WM and ccontrols. Left (n = 34), right (n = 33), and bilateral (n = 8) MTLE + HS(adjusted for age and ICV) are reported as z-scores derived from the mintervals (CI). *p < 0.05; **p < 0.01; ***p < 0.001.

159

hose of the healthy controls, we observed significantolume loss across several additional structures. Amyg-alar volumes were reduced ipsilaterally in the groups ofatients with left MTLE + HS (p = 0.008) and right MTLE + HSp = 0.002). A trend of amygdalar volume reduction wasbserved bilaterally in patients with bilateral MTLE + HS (leftmygdala: p = 0.09; right amygdala: p = 0.101) although theffect was not statistically significant (see Fig. 2a). Ipsilat-ral volume reduction was also detected in the thalamusn left (p < 0.0001) and right MTLE + HS (p = 0.003) and bilat-rally in patients with bilateral MTLE + HS (left thalamus:= 0.003; right thalamus: p = 0.008) (see Fig. 2b). No statis-

ically significant volume reduction was detected across theallidum, putamen or the caudate.

olumes of cerebral and cerebellar grey and whiteatterse observed a reduction in total cerebral WM volume ipsi-

aterally in left (p < 0.0001) and right MTLE + HS patients

p = 0.002) compared to the healthy controls (see Fig. 2c).ilateral cerebral WM volume reduction was also observed

n bilateral MTLE + HS patients (left cerebral WM: p = 0.003,ight cerebral WM: p = 0.016). No statistically significant

erebellar GM volumes between MTLE + HS patients and healthypatients were compared to healthy controls (n = 50). Volumesean of the controls data. Error bands represent 95% confidence

Page 5

1

dbCerp

TaMpFfHwM

IGppeinihea1((pa

ToUi2a‘awTop(vnetaWtwwa1rop

TsUofpesaehTmcsltGep

cictanemSldnd(posdeWltwlpdae

PoGSsbi

60

ifference in total cerebral GM volume was observedetween the patient groups and the healthy controls.onsidering the cerebellum we observed a modest bilat-ral reduction in GM volume but this reduction appearedestricted to patients with left MTLE + HS (ipsilateral:= 0.019; contralateral: p = 0.016), see Fig. 2d.

he relationship between clinical variables and regionaltrophyann—Whitney U tests revealed no differences between theatient groups in age at seizure onset or duration of epilepsy.urther, patients with left and right MTLE + HS did not dif-er in terms of the frequency of partial seizures or SGTCS.owever, a higher frequency of SGTCS was noted in patientsith bilateral MTLE + HS when compared to left and rightTLE + HS patients.

PIs and regional atrophyiven the previous attention that IPIs have received asotential initiators of chronic and progressive hippocam-al atrophy (Meyer et al., 1954; Mathern et al., 2002; Liut al., 2005), we explored the nature of volumetric changesn patients with and without a history of IPIs. We observedo significant volume difference of any of the regions wenvestigated between patients with a positive and negativeistory of IPIs, although we note a trend of greater ipsilat-ral hippocampal volume reduction in those patients withpositive history of IPI (p = 0.089) (see Supplementary Fig.

). We also noticed that patients who had febrile seizuresFS) had a considerably earlier age at onset of seizuresmedian = 8 year, 1st and 3rd quartile: 2.75 and 15.25) com-ared to those with a negative history (median = 14 year, 1stnd 3rd quartile: 6 and 27; p = 0.032).

he relationship between duration of epilepsy, age atnset of seizures and regional atrophysing univariate linear regression modelling, we character-

zed the correlation between duration of epilepsy (mean2 years, range 1—52 years) as an ‘‘explanatory variable’’nd each measure of regional atrophy in turn as an‘outcome measure’’. We found ipsilateral hippocampal,mygdalar, and cerebral WM volume to correlate negativelyith duration of epilepsy (see Fig. 3a—c). As shown in

able 1, duration of epilepsy was found to predict 29.6%f ipsilateral hippocampal volume variance (beta = −0.6;< 0.0001), 12.9% of ipsilateral amygdala volume variance

beta = −0.4; p = 0.003), and 8.2% of ipsilateral WM volumeariance (beta = −0.3; p = 0.013). Duration of epilepsy didot relate to variance in ipsilateral thalamic, contralat-ral hippocampal, or cerebellar GM volumes. It is notablehat the negative correlations between duration of epilepsynd the volume of ipsilateral hippocampus and cerebralM are present in both patients with, and without a his-

ory of IPIs (Fig. 4a and c) while the negative correlationith amygdalar volume appears restricted to those patientsith a positive history of IPI (Fig. 4b). A similar univari-te regression model with age at onset of seizures (mean

4 years, range 1—54 years) as the ‘‘explanatory variable’’evealed a positive correlation between age at seizurenset and ipsilateral cerebral WM volume only (beta = 0.27;= 0.017).

tdp(

S. Alhusaini et al.

he relationship between estimated lifetime number ofeizures and regional atrophysing univariate regression model with lifetime numberf partial seizures as the ‘‘explanatory variable’’, weound no significant correlation between the number ofartial seizures and volume measurements of any of thexplored structures. However, a similar univariate regres-ion model, but with estimated lifetime number of SGTCSs the explanatory variable explained 28.7% of ipsilat-ral (beta = −0.535; p = 0.004) and 41.1% of contralateralippocampal volume variance (beta = −0.641; p = 0.001).he same model explained 15.6% of ipsilateral thala-ic (beta = −0.395; p = 0.041) and 26.4% of ipsilateral

erebral WM volume variance (beta = −0.514; p = 0.006),ee Table 1. Although reduction in volume of cerebel-ar GM was only observed in patients with left MTLE + HS,he model predicted approximately 25% of cerebellarM volume variance bilaterally in all patients (ipsilat-ral: beta = −0.501; p = 0.008; contralateral: beta = −0.507,= 0.013).

We next applied a multiple linear regression modelling toontrol for multicollinearity and to determine the degree ofndependence attributable to each explanatory variable inorrelations with regional atrophy. The model included all ofhe ‘‘explanatory variables’’ (i.e. duration of epilepsy, aget onset of seizures, history of IPIs and estimated lifetimeumber of partial seizures and SGTCS) and was applied toach measure of regional atrophy in turn as an ‘‘outcomeeasure’’. This model, when applied to the patients with

GTCS activity (n = 31), showed that the variance in ipsi-ateral hippocampal volume was independently related touration of epilepsy (beta = −0.73; p < 0.001) and lifetimeumber of partial seizures (beta = −0.35; p = 0.007). A bor-erline association with age at onset was also observedbeta = 0.29; p = 0.051). Strikingly, contralateral hippocam-al volume variance was only related to lifetime numberf SGTCS (beta = −0.66; p = 0.009) (see Table 2). Theame multivariate model predicted that ipsilateral amyg-alar volume variance was only related to duration ofpilepsy (beta = −0.65; p = 0.005), while that of ipsilateralM volume was related to IPI (beta = 0.45; p = 0.033) and

ifetime number of SGTCS (beta = −0.45; p = 0.048). Further,he model predicted that cerebellar GM volume varianceas mainly related to lifetime number of SGTCS (ipsi-

ateral: beta = −0.55 p = 0.039; contralateral: beta = −0.71,= 0.017). None of the clinical variables related indepen-ently to ipsilateral thalamic volume variance (see Table 2),lthough lifetime number of SGTCS appeared to have a mod-st negative correlation (beta = −0.40; p = 0.133).

atients with SGTCS vs. patients with partial seizuresnlyiven the strong negative correlation we observed betweenGTCS and contralateral hippocampal volume, we hypothe-ised that contralateral hippocampal volume reductions wille limited largely to those MTLE + HS patients experienc-ng SGTCS. Indeed, when we compared this group (n = 31)

o those who only experience partial seizures (n = 19), weetected a significant volume reduction in contralateral hip-ocampus (p < 0.041) in those patients experiencing SGTCSsee Supplementary Fig. 2.0).

Page 6

Quantitative MRI study and clinical characteristics of epilepsy 161

Figure 3 The correlation of duration of epilepsy with ipsilateral (circles) and contralateral (triangles) hippocampal, amygdalar,and cerebral white matter (WM) volume. Volumes (adjusted for age and ICV) are reported as z-scores derived from the mean of thecontrols data.

Table 1 Univariate analysis exploring duration of epilepsy and lifetime number of SGTCS as predictors of regional volume loss.

Brain region Duration of epilepsy Lifetime number of SGTCS

r2 p-Value r2 p-Value

HippocampusIpsilaterala 0.296 <0.0001 0.287 0.004Contralateralb 0.043 0.101 0.411 0.001

AmygdalaIpsilaterala 0.129 0.002 0.111 0.089

ThalamusIpsilaterala 0.009 0.420 0.156 0.041

Cerebral WMIpsilaterala 0.082 0.013 0.264 0.006

Cerebellar GMIpsilateralb 0.04 0.086 0.251 0.008Contralateralb 0.036 0.124 0.257 0.013

Results of Univariate linear regression models with duration of epilepsy as a single explanatory variable of regional atrophy in thesecond column (n = 75) and lifetime number of SGTCS as a single explanatory variable in the third column (n = 31) are shown. Volumeswere measured in mm3 and adjusted for ICV and age. SGTCS: secondarily generalized tonic-clonic seizures; r2: multiple regression(determination) coefficient; WM: white matter, GM: grey matter.Bold values represent significant p-values.

a Regions where volume reduction was detected in all patient groups.b Regions where volume reduction was only detected in left MTLE + HS.

Figure 4 The correlation between duration of epilepsy in patients with (triangles) and without (circles) initial precipitating insults(IPIs) with ipsilateral hippocampal, amygdalar, and cerebral white matter (WM) volume. Volumes (adjusted for age and ICV) arereported as z-scores derived from the mean of the controls data.

Page 7

162S.

Alhusainietal.

Table 2 Multivariate analysis exploring duration of epilepsy, age at onset, IPIs, and estimates of lifetime number of partial seizures and SGTCS as predictors of regionalatrophy in patients who experience SGTCS (n = 31).

Brain region r2 p-Value Duration of epilepsy Age at onset IPIs Partial Seizures SGTCS

Beta p-Value Beta p-Value Beta p-Value Beta p-Value Beta p-Value

HippocampusIpsilaterala 0.312 0.002 −0.73 <0.001 0.29 0.051 0.02 0.875 −0.35 0.007 −0.21 0.167Contralateralb 0.519 0.022 −0.14 0.537 0.06 0.943 −0.27 0.207 −0.26 0.203 −0.66 0.009

AmygdalaIpsilaterala 0.051 0.649 −0.65 0.005 0.16 0.398 0.251 0.169 −0.03 0.873 −0.09 0.675

ThalamusIpsilaterala 0.196 0.326 −0.13 0.607 0.17 0.490 0.01 0.984 −0.08 0.707 −0.40 0.133

Cerebral WMIpsilaterala 0.342 0.044 −0.09 0.682 0.377 0.09 0.449 0.033 −0.01 0.968 −0.45 0.048

Cerebellar GMIpsilateralb 0.184 0.337 0.05 0.851 −0.21 0.410 0.08 0.716 −0.03 0.887 −0.55 0.039Contralateralb 0.394 0.140 0.23 0.495 0.13 0.676 0.08 0.756 −0.15 0.557 −0.73 0.017

Results of a multiple regression model with duration of epilepsy, age at onset, history of initial precipitating insults (IPIs) and lifetime number of partial seizures and SGTCS included asexplanatory variables of regional atrophy are shown (n = 31). Volumes were measured in mm3 and adjusted for ICV and age. IPI: initial precipitating insult, SGTCS: secondary generalizedtonic-clonic seizures; r2: multiple regression (determination) coefficient; WM: white matter, GM: grey matter.Bold values represent significant p-values.

a Regions where volume reduction was detected in all patient groups.b Regions where volume reduction was detected in left MTLE + HS.

Page 8

sy

rpWhahsoweycactSBredc

tpiswtiecsantcoamiMR

MM

Qpaaisctrpv

Quantitative MRI study and clinical characteristics of epilep

Discussion

The primary goal of this study was to further characterizethe significance of extrahippocampal atrophy in MTLE + HS.Our findings indicate that multiple brain regions beyond thehippocampus are volumetrically affected in patients withMTLE + HS. Patients displayed significant volume reductionin ipsilateral amygdala, thalamus and cerebral WM. In addi-tion, patients with left MTLE + HS displayed volume loss incontralateral hippocampus and cerebellar GM bilaterally.These findings support previous observations in the litera-ture and solidify a picture of extra-hippocampal atrophy inMTLE + HS (Van Paesschen et al., 1996; Decarli et al., 1998;Sandok et al., 2000; Dreifuss et al., 2001; Natsume et al.,2002; Bernasconi et al., 2003, 2005; Labate et al., 2008;McDonald et al., 2008).

A secondary goal of this study was to determine thedegree to which clinical features of epilepsy correlated withextrahippocampal atrophy. Ipsilateral amygdala and cere-bral WM showed a progressive atrophy that is likely to bemagnified in the presence of prior IPIs history. Additionally,lifetime number of generalized seizures correlated signif-icantly with contralateral hippocampal, cerebral WM, andcerebellar GM atrophy and modestly with thalamic atrophy.

Brain regional atrophy, initial precipitating insults(IPIs), and duration of epilepsy

Our findings support the view that patients with prior IPIshistory have a faster rate of hippocampal atrophy (Mathernet al., 2002; Liu et al., 2005). Additionally, IPIs appeared toinfluence amygdalar and cerebral WM atrophy with progres-sive amygdalar volume loss apparently restricted to thosepatients with a positive IPIs history (see Fig. 4b). Associa-tions between amygdalar damage and IPIs (such as complexfebrile seizures and intracranial infection) have previouslybeen observed (Salmenperä et al., 2001; Kälviäinen andSalmenperä, 2002). Further, Hermann et al. reported a morepronounced reduction in cerebral WM volume in patientswith childhood-onset TLE; suggesting that early insults tothe developing brain are significant factors (Hermann et al.,2002).

However, it is important to note that the progressiveatrophy in ipsilateral hippocampus and cerebral WM wasalso observed in the absence of IPIs history, just that therate of atrophy appeared slower (see Fig. 4a and c). Also,the existence of ipsilateral hippocampal, amygdalar, andcerebral WM volume loss in patients without a prior IPIshistory (see Supplementary Fig. 1.0) suggests that IPIs (ascurrently understood) alone or in combination with normalage-related atrophy are insufficient to explain progressivehippocampal and/or extrahippocampal atrophy.

Brain regional atrophy and seizure activity

Evidence from some longitudinal MRI studies suggests

that progressive degeneration in ipsilateral hippocampusresults from damage secondary to recurrent seizure activ-ity (Briellmann et al., 2002; Fuerst et al., 2003). Wenote, however, that other longitudinal MRI studies found no

pSir

163

elationship between seizure frequency and progressive hip-ocampal damage (Holtkamp et al., 2004; Liu et al., 2005).e feel these discrepancies could in part be explained byeterogeneity in the studied populations and differences innalytical approach. The degree of atrophy appears to beeavily influenced by side of seizure onset; neither studytructured analysis in a manner that captured the effectf the side of seizure onset and both contained patientsith different epilepsy types (Holtkamp et al., 2004; Liut al., 2005). In our study, when we restricted our anal-sis to patients with detailed seizure activity history, alear relationship between ipsilateral hippocampal volumesnd lifetime number of partial seizures was detected afterontrolling for the effect of the other clinical variables. Fur-hermore, our findings suggest that patients who experienceGTCS are prone to contralateral hippocampal volume loss.ased on these observations it is tempting to speculate thatecurrent partial seizures contribute to progressive ipsilat-ral hippocampal atrophy, while contralateral hippocampalamage may result from SGTCS. Further work is required toonfirm or reject this hypothesis.

Although SGTCS may play a role in causing ipsilateralhalamic atrophy, the underlying mechanism appears com-lex and may be related to an interaction of several factorsncluding deafferantation of mesial temporal regions andeizure-induced damage (Mueller et al., 2010). In agreementith previous observations, the relationship between life-

ime number of SGTCS and cerebral WM atrophy detectedn our study suggests seizure-induced damage (Seidenbergt al., 2005). Similarly, the bilateral cerebellar atrophyould be secondary to damage caused by spreading SGTCSeizures or other factors related to epilepsy chronicity (e.g.,ntiepileptic medications) (Hermann et al., 2005). Alter-atively, these relationships may indicate a predispositiono seizure generalization. It is important to note here thaterebellar GM volume was reduced in left MTLE patientsnly, suggesting that these patients might be susceptible tomore severe form of MTLE + HS. Bilateral structural abnor-alities, including contralateral hippocampal volume loss

n patients with left MTLE compared to those with rightTLE have been observed previously (Bonliha et al., 2007;iederer et al., 2008).

RI-based regional volumes as endophenotypes inTLE + HS

uantitative measurements of candidate brain regions areotential endophenyotypes for epilepsy. However, their suit-bility could be compromised if the structures themselvesre influenced by the chronicity of epilepsy (e.g., seizure-nduced damage). Seizure-induced injury would render thetructure less heritable, impacting on its potential appli-ation in genetic mapping studies in epilepsy patients. Inhis context, we and others have demonstrated a clearelationship between duration of epilepsy and hippocam-al, amygdalar, and cerebral WM atrophy. Additionally, theolumes of ipsilateral cerebral WM, contralateral hippocam-

us, and bilateral cerebellar GM seem to be associated withGTCS frequency. These relationships may need to be takennto account when MRI-based volumes of these particularegions are being considered as potential endophenotypes

Page 9

1

itrto

C

T

A

TpF0

A

Scd

R

A

B

B

B

B

B

B

B

B

B

C

C

C

D

D

D

D

D

F

F

F

F

F

F

F

64

n MTLE + HS patients. The impact of disease chronicity onhese structures also raises the question as to whethereduced volumes, prior to onset of epilepsy, act as risk fac-ors for epilepsy itself. Studies involving unaffected siblingsf index cases would help shed light on this question.

onflicts of interest

he authors have no conflicts of interest to disclose.

cknowledgements

he authors thank all patients and participants who tookart in this study. This work was funded by Scienceoundation Ireland Research Frontiers Programme award8/RFP/GEN1538.

ppendix A. Supplementary data

upplementary data associated with this arti-le can be found, in the online version, atoi:10.1016/j.eplepsyres.2011.11.005.

eferences

ndrade-Valenca, L.P., Valenca, M.M., Ribeiro, L.T., Matos, A.L.,Sales, L.V., Velasco, T.R., Santos, A.C., Leite, J.P., 2003. Clini-cal and neuroimaging features of good and poor seizure controlpatients with mesial temporal lobe epilepsy and hippocampalatrophy. Epilepsia 44, 807—814.

erg, A.T., Berkovic, S.F., Brodie, M.J., Buchhalter, J., Cross, J.H.,van Emde Boas, W., Engel, J., French, J., Glauser, T.A., Mathern,G.W., Moshe, S.L., Nordli, D., Plouin, P., Scheffer, I.E., 2010.Revised terminology and concepts for organization of seizuresand epilepsies: report of the ILAE Commission on Classificationand Terminology, 2005—2009. Epilepsia 51, 676—685.

erkovic, S.F., Andermann, F., Olivier, A., Ethier, R., Melanson, D.,Robitaille, Y., Kuzniecky, R., Peters, T., Feindel, W., 1991. Hip-pocampal sclerosis in temporal lobe epilepsy demonstrated bymagnetic resonance imaging. Ann. Neurol. 29, 175—182.

ernasconi, N., Bernasconi, A., Caramanos, S.B., Antel, B., Ander-mann, F., Arnold, D.L., 2003. Mesial temporal damage intemporal lobe epilepsy: a volumetric MRI study of the hip-pocampus, amygdala, and parahippocampal region. Brain 126,462—469.

ernasconi, N., Duchesne, S., Janke, A., Lerch, J., Collins, D.L.,Bernasconi, A., 2004. Whole-brain voxel-based statistical anal-ysis of gray matter and white matter in temporal lobe epilepsy.Neuroimage 23, 717—723.

ernasconi, N., Natsume, J., Bernasconi, A., 2005. Progression intemporal lobe epilepsy: differential atrophy in mesial temporalstructures. Neurology 65, 223—228.

onliha, L., Rorden, C., Halford, J.J., Eckert, M., Appenzeller, S.,Cendes, F., Li, L.M., 2007. Asymmetrical extra-hippocampal greymatter loss related to hippocampal atrophy in patients withmedial temporal lobe epilepsy. J. Neurol. Neurosurg. Psychiatry78, 286—294.

riellmann, R.S., Newton, M.R., Wellard, M., Jackson, G.D., 2001.

Hippocampal sclerosis following brief generalized seizures inadulthood. Neurology 57, 315—317.

riellmann, R.S., Berkovic, S.F., Syngeniotis, A., King, M.A., Jack-son, G.D., 2002. Seizure-associated hippocampal volume loss:

F

S. Alhusaini et al.

a longitudinal magnetic resonance study of temporal lobeepilepsy. Ann. Neurol. 51, 641—644.

ruton, C.J., 1988. The Neuropathology of Temporal Lobe Epilepsy.Oxford University Press, New York.

avazos, J.E., Das, I., Sutula, T.P., 1994. Neuronal loss inducedin limbic pathways by kindling: evidence for induction of hip-pocampal sclerosis by repeated brief seizures. J. Neurosci. 14,3103—3121.

endes, F., Andermann, F., Gloor, P., Lopes-Cendes, I., Ander-mann, E., Melanson, D., Jones-Gotman, M., Robitaille, Y., Evans,A., Peters, T., 1993. Atrophy of mesial structures in patientswith temporal lobe epilepsy: cause or consequence of repeatedseizures? Ann. Neurol. 34, 795—801.

ommission of Classification and Terminology of the InternationalLeague Against Epilepsy, 1981. Proposal for revised clinicaland electroencephalographic classification of epileptic seizures.Epilepsia 22, 489—501.

ale, A.M., Fischl, B., Sereno, M., 1999. Cortical surface-basedanalysis: segmentation and surface reconstruction. Neuroimage9, 179—194.

ecarli, C., Hatta, G., Fazilat, S., Gaillard, W.D., Theodore,W.H., 1998. Extratemporal atrophy in patients with complexpartial seizures of left temporal origin. Ann. Neurol. 43,41—45.

ickerson, B.C., Fenstermacher, E., Salat, D.H., Wolk, D.A.,Maguire, R.P., Desikan, R., Pacheco, J., Quinn, B.T., Van derKouwe, A., Greve, D.N., Blacker, D., Albert, M.S., Killiany, R.J.,Fischl, B., 2008. Detection of cortical thickness correlates ofcognitive performance: reliability across MRI scan sessions, scan-ners, and field strengths. Neuroimage 39, 10—18.

oherty, C.P., Fitzsimons, M., Meredith, G., Thornton, J., McMackin,D., Farrell, M., Philips, J., Staunton, H., 2003. Rapid stereolog-ical quantitation of temporal neocortex in TLE. Magn. Reson.Imaging 21, 511—518.

reifuss, S.M., Vingerhoets, F., Lazeyras, F.P., Gonzales Andino, S.P.,Delavelle, J.M., Seeck, M.M., 2001. Volumetric measurementsof subcortical nuclei in patients with temporal lobe epilepsy.Neurology 57, 1636—1641.

ischl, B., Sereno, M., Dale, A.M., 1999a. Cortical surface-basedanalysis: II: inflation. Flattening, and a surface-based coordinatesystem. Neuroimage 9, 195—207.

ischl, B., Sereno, M.I., Tootell, R.B., Dale, A.M., 1999b. High-resolution intersubject averaging and a coordinate system forthe cortical surface. Hum. Brain Mapp. 8, 272—284.

ischl, B., Liu, A., Dale, A.M., 2001. Automated manifold surgery:constructing geometrically accurate and toplogically correctmodels of the human cerebral cortex. IEEE Trans. Med. Imaging20, 70—80.

ischl, B., Salat, D., Busa, E., Albert, M., Dieterich, M., Haselgrove,C., van Der Kouwe, A., Killiany, R., Kennedy, D., Klaveness, S.,Montillo, A., Markis, N., Rosen, B., Dale, A.M., 2002. Whole brainsegmentation: automated labeling of neuroanatomical struc-tures in the human brain. Neuron 33, 341—355.

ischl, B., van der Kouwe, A., Destrieux, C., Halgren, E., Seogonne,F., Salat, D., Busa, E., Seidman, L., Goldstein, J., Kennedy, D.,Caviness, V., Makris, N., Rosen, B., Dale, A.M., 2004. Automati-cally parcellating the human cerebral cortex. Cereb. Cortex 14,11—22.

ree, S.L., Bergin, P.S., Fish, D.R., Cook, M.J., Shorvon, S.D.,Stevens, J.M., 1995. Methods for normalization of hippocampalvolumes measured with MR. AJNR 16, 637—643.

rench, J.A., Williamson, P.D., Thadani, V.M., Darcey, T.M., Matt-son, R.H., Spencer, S.S., Spencer, D.D., 1993. Characteristics ofmedial temporal lobe epilepsy: I. Results of history and physical

examination. Ann. Neurol. 34, 774—780.

uerst, D., Shah, J., Shah, A., Watson, C., 2003. Hippocampal scle-rosis is a progressive disorder: a longitudinal volumetric MRIstudy. Ann. Neurol. 53, 413—416.

Page 10

sy

M

M

M

M

M

M

N

P

R

R

S

S

S

S

S

T

V

Quantitative MRI study and clinical characteristics of epilep

Garcia-Finana, M., Denby, C.E., Keller, S.S., Wieshmann, U.C.,Roberts, N., 2006. Degree of hippocampal atrophy is relatedto side of seizure onset in temporal lobe epilepsy. AJNR 27,1046—1052.

Han, X., Jovicich, J., Salat, D., van Der Kouwe, A., Quinn, B., Czan-ner, S., Busa, E., Pacheco, J., Albert, M., Killiany, R., Maguire,P., Rosas, D., Makris, N., Dale, A., Dickerson, B., Fischl, B., 2006.Reliability of MRI-derived measurements of human cerebral cor-tical thickness: The effects of field strength, scanner upgradeand manufacturer. Neuroimage 32, 180—194.

Helbig, I., Scheffer, I.E., Mulley, J., Berkovic, S.F., 2008. Navigat-ing the channels and beyond: unravelling the genetics of theepilepsies. Lancet Neurol. 7, 231—245.

Hermann, B., Seidenberg, M., Bell, B., 2002. The neurodevelop-mental impact of childhood onset temporal lobe epilepsy onbrain structure and function and the risk of progressive cognitiveeffects. Prog. Brain Res. 135, 429—438.

Hermann, B., Bayless, K., Hansen, R., Parrish, J., Seidenberg, M.,2005. Cerebellar atrophy in temporal lobe epilepsy. EpilepsyBehav. 7, 279—287.

Holtkamp, M., Schuchmann, S., Gottschalk, S., Meierkord, M., 2004.Recurrent seizures do not cause hippocampal damage. J. Neurol.46, 948—950.

Kälviäinen, R., Salmenperä, T., Partanen, K., Vainio, P., Riekki-nen, P., Pitkänen, A., 1998. Recurrent seizures may causehippocampal damage in temporal lobe epilepsy. Neurology 50,1377—1382.

Kälviäinen, R., Salmenperä, T., 2002. Do recurrent seizurescause neuronal damage? A series of studies with MRI vol-umetry in adults with partial seizures. Prog. Brain Res. 135,279—295.

Kasperaviciute, D., Catarino, C.P., Heinzen, E.L., Depondt, C., Cav-alleri, G.L., Caboclo, L.O., Tate, S.K., Jamnadas-Khoda, J.,Chinthapalli, K., Clayton, L.M., Shianna, K.V., Radtke, R.A.,Mikati, M.A., Gallentine, W.B., Husain, A.M., Alhusaini, S.,Leppert, D., Middleton, L.T., Gibson, R.A., Johnson, M.R.,Matthews, P.M., Hosford, D., Heuser, K., Amos, L., Ortega, M.,Zumsteg, D., Weiser, H.G., Steinhoff, B.J., Krämer, G., Hansen,J., Dorn, T., Kantanen, A.M., Gjerstad, L., Peuralinna, T., Her-nandez, D.G., Eriksson, K.J., Kälviäinen, R.K., Doherty, C.P.,Wood, N.W., Pandolfo, M., Duncan, J.S., Sander, J.W., Delanty,N., Goldstein, D.B., Sisodiya, S.M., 2010. Common genetic vari-ation and susceptibility to partial epilepsies: a genome-wideassociation study. Brain 133, 2136—2147.

Kwan, P., Arzimanoglou, A., Berg, A.T., Brodie, M.J., Allen Hauser,W., Mathern, G., Moshe, S.L., Perucca, E., Wiebe, S., French, J.,2010. Definition of drug resistant epilepsy: consensus proposalby the ad hoc Task Force of the ILAE Commission on TherapeuticStrategies. Epilepsia 51, 1069—1077.

Labate, A., Cerasa, A., Gambardella, A., Aguglia, U., Quattrone,A., 2008. Hippocampal and thalamic atrophy in mild temporallobe epilepsy: a VBM study. Neurology 71, 1094—1101.

Lee, J.K., Lee, J.M., Kim, J.S., Kim, I.Y., Evan, A.C., Kim, S.I., 2006.A novel quantitative cross-validation of different cortical surfacereconstruction algorithms using MRI phantom. Neuroimage 31,572—584.

Liu, R.S., Lemieux, L., Sander, J.W., Sisodiya, S.M., Duncan, J.S.,2002. Seizure-associated hippocampal volume loss: a longitudi-nal magnetic resonance study of temporal lobe epilepsy. Ann.Neurol. 52, 861.

Liu, R.S., Lemieux, L., Bell, G.S., Sisodiya, S.M., Bartlett, P.,Shorvon, S., Sander, J., Duncan, J., 2005. Cerebral damage inepilepsy: a population-based longitudinal quantitative MRI study.Epilepsia 46, 1482—1494.

Mathern, G.W., Adelson, P.D., Cahan, L.D., Leite, J.P., 2002. Hip-pocampal neuron damage in human epilepsy: Meyer’s hypothesisrevisited. Prog. Brain Res. 135, 237—252.

165

cDonald, C.R., Hagler Jr., D.J., Ahmadi, M.E., Tecoma, E., Iragui,V., Dale, A.M., Halgren, E., 2008. Subcortical and cerebellaratrophy in mesial temporal lobe epilepsy revealed by automaticsegmentation. Epilepsy Res. 79, 130—138.

eyer, A., Falconer, M., Beck, E., 1954. Pathological findings intemporal lobe epilepsy. J. Neurol. Neurosurg. Psychiatry 17,276—285.

oran, N.F., Lemieux, L., Kitchen, N.D., Fish, D.R., Shorvon,S.D., 2001. Extrahippocampal temporal lobe atrophy in tem-poral lobe epilepsy and mesial temporal sclerosis. Brain 124,167—175.

orey, R.A., Selgrade, E.S., Wagner 2nd, H.R., Huettel, S.A., Wang,L., McCarthy, G., 2010. Scan-rescan reliability of subcorticalbrain volumes derived from automated segmentation. Hum.Brain Mapp. 31, 1751—1762.

ueller, S.G., Laxer, K.D., Cashdollar, N., Buckley, S., Paul, C.,Weiner, M., 2006. Voxel-based optimized morphometry (VMB)of gray and white matter in temporal lobe epilepsy (TLE)with and without mesial temporal sclerosis. Epilepsia 47,900—907.

ueller, S.G., Laxer, K.D., Barakos, J., Cheong, I., Finlay, D., Gar-cia, P., Cardenas-Nicolson, V., Weiner, M.W., 2010. Involvementof the thalamocortical network in TLE with and without mesial-temporal sclerosis. Epilepsia 51, 1436—1445.

atsume, J., Bernasconi, N., Andermann, F., Bernasconi, A.,2002. MRI volumetry of the thalamus in temporal, extratem-poral, and idiopathic generalized epilepsy. Neurology 60,1296—1300.

itkänen, A., Nissinen, J., Nairismag, J., Lukasiuk, K., Grohn, O.H.,Miettinen, R., Kauppinen, R., 2002. Progression of neuronal dam-age after status epilepticus and during spontaneous seizures ina rat model of temporal lobe epilepsy. Prog. Brain Res. 135,67—83.

iederer, F., Lanzenberger, R., Kaya, M., Prayer, D., Serles, W.,Baumgarther, C., 2008. Network atrophy in temporal lobeepilepsy: a voxel-based morphometry study. Neurology 71,419—425.

osas, H.D., Liu, A.K., Hersch, S., Glessner, M., Ferrante, R.J.,Salat, D.H., van der Kouwe, A., Jenkins, B.G., Dale, A.M., Fis-chl, B., 2002. Regional and progressive thinning of the corticalribbon in Huntington’s disease. Neurology 58, 695—701.

almenperä, T., Kälviäinen, R., Partanen, K., Pitkanen, A., 2001.Hippocampal and amygaloid damage in partial epilepsy: a crosssectional MRI study of 241 patients. Epilepsy Res. 46, 69—82.

almenperä, T., Kononen, M., Roberts, N., Vanninen, R., Pitkanen,A., Kälviäinen, R., 2005. Hippocampal damage in newly diag-nosed focal epilepsy: a prospective MRI study. Neurology 64,62—68.

andok, E.K., O’Brien, T.G., Jack, C.R., So, E.L., 2000. Significanceof cerebellar atrophy in intractable temporal lobe epilepsy: aquantitative MRI study. Epilepsia 41, 1320—2000.

cott, R.C., King, M.D., Gadian, D.G., Neville, B.G.R., Connelly, A.,2003. Hippocampal abnormalities after prolonged febrile con-vulsion: a longitudinal MRI study. Brain 126, 2551—2559.

eidenberg, M., Kelly, K.G., Parrish, J., Geary, E., Dow, C.,Rutecki, P., Hermann, B., 2005. Ipsilateral and contralateralMRI volumetric abnormalities in chronic unilateral tempo-ral lobe epilepsy and their clinical correlates. Epilepsia 46,420—430.

asch, E., Cendes, F., Li, L.M., Dubeau, F., Andermann, F., Arnold,D.L., 1999. Neuroimaging evidence of progressive neuronal lossand dysfunction in temporal lobe epilepsy. Ann. Neurol. 45,568—576.

an Landingham, K.E., Heinz, E.R., Cavazos, G.E., Lewis, D.V.,

1998. Magnetic resonance imaging evidence of hippocampalsclerosis after prolonged focal febrile convulsions. Ann. Neurol.,43413—43426.

Page 11

1

V

W

1238—1241.

66

an Paesschen, W., Connelly, A., Johnson, C., Duncan, J.S.,1996. The amygdala and intractable temporal lobe epilepsy: aquantitative magnetic resonance imaging study. Neurology 47,

1021—1031.

eishmann, U.C., Woermann, F.G., Lemieux, L., Free, S.L.,Bartlett, P.A., Smith, S.J., Duncan, J.S., Stevens, J.M., Shorvon,S.D., 1997. Development of hippocampal atrophy: a serial

W

S. Alhusaini et al.

magnetic resonance imaging study in a patient who devel-oped epilepsy after generalised status epilepticus. Epilepsia 38,

ieser, H.G., 2004. ILAE Commission on Neurosurgery of epilepsy,ILAE Commission report: mesial temporal lobe epilepsy with hip-pocampal sclerosis. Epilepsia 45, 695—714.