j ourna l h om epa ge: www.elsev ier .com/ locate /ep i lepsyres
pilepsy surgery in Neurofibromatosis Type 1
armen Barbaa, Thomas Jacquesb,c, Philippe Kahaned, Tilman Polstere,ean Isnardf, Frans S.S. Leijteng, Cigdem Ozkarah, Laura Tassi i,lavio Giordanoj, Maura Castagnak, Alison Johnc, Buge Özl, Caroline Salonm,athalie Streichenbergern, Judith Helen Crosso, Renzo Guerrini a,p,∗
Pediatric Neurology Unit, Children’s Hospital Meyer-University of Florence, Florence, ItalyUCL Institute of Child Health, London, UKDepartment of Histopathology, Great Ormond Street Hospital NHS Trust, London, UKNeurology Department and INSERM U-836-UJF-CEA, Grenoble University Hospital, Grenoble, FranceKlinik Mara I, Bethel Epilepsy Center, Bielefeld, GermanyFunctional Neurology and Epileptology Unit, Hôpital Neurologique, Hospices Civils de Lyon, Bron, FranceUniversity Medical Center Utrecht, Rudolf Magnus Institute, Utrecht, The NetherlandsDepartment of Neurology, Cerrahpasa Medical Faculty, Istanbul, TurkeyEpilepsy Surgery Center Claudio Munari, Ospedale Niguarda Cà Granda, Milan, ItalyPediatric Neurosurgery Unit, Children’s Hospital Meyer-University of Florence, Florence, ItalyPathological Anatomy III Division, Department of Surgery, University of Pisa, Pisa, ItalyNeuropathology Division of Pathology Department, Cerrahpasa Medical School, Istanbul, TurkeyPathology Department, Grenoble University Hospital, Grenoble, FrancePathology Department, Hôpital Neurologique, Hospices Civils de Lyon, Bron-Faculty of Medicine Claude Bernard Lyon1, FranceThe Prince of Wales’s Chair of Childhood Epilepsy, UCL-Institute of Child Health, Great Ormond Street Hospital for Children &ational Centre for Young People with Epilepsy, London, UKIRCCS Stella Maris, Pisa, Italy
eceived 23 September 2012; received in revised form 4 December 2012; accepted 27 February 2013vailable online 16 April 2013
KEYWORDSNeurofibromatosisType 1;Epilepsy surgery;
Summary Epilepsy is relatively uncommon in patients with Neurofibromatosis Type 1 (NF1)and seizures are usually well controlled with antiepileptic treatment. However, pharmacore-sistance has been reported in patients with NF1 and MRI evidence of malformations of corticaldevelopment or glioneuronal tumours. Available information on epilepsy surgery in NF1 is
DNET; limited to a few patients with gliomas and glioneuronal tumours who underwent lesionectomies.
Temporal lobe We conducted a survey amongst 25 European epilepsy surgery centres to collect patients withNF1 who had undergone surgery for drug-resistant seizures and identified 12 patients fromeight centres. MRI abnormalities were present in all patients but one. They were unilateraltemporal in eight, bilateral temporal in one and multilobar or hemispheric in two. Seizures
∗ Corresponding author at: Pediatric Neurology Unit and Laboratories, Children’s Hospital A. Meyer-University of Florence, Viale Pieraccini4, 50139 Firenze, Italy. Tel.: +39 0555662573; fax: +39 0555662329.
originated from the temporal lobe in ten patients, from the temporo-parieto-occipital region inone, and were bitemporal in one. One year after surgery eight patients were seizure free, onehad worthwhile improvement and the remaining three had experienced no benefit. Postoperativeoutcome, available at 2 years in ten patients and at 5 years in three, remained stable in all but onewhose seizures reappeared. Histology revealed dysembryoplastic neuroepithelial tumour (DNET)in five patients, hippocampal sclerosis in four, mixed pathology in one and polymicrogyria in one.No histological abnormality was observed in the remaining patient.Epilepsy surgery can be performed effectively in patients with NF1 provided a single and well-delimited epileptogenic zone is recognized. The high prevalence of DNETs in this series mightsuggest a non-fortuitous association with NF1.
s res
wbesaPC
pdocc
rah
M
P
WsNTfAFNcDlinl(fcs
To reduce the variability in data collection, we asked theparticipating centres to complete a questionnaire (AppendixB) on general patients’ characteristics, NF1 and epilepsy
Neurofibromatosis Type 1 (NF1) is a common autosomaldominant neurocutaneous condition, with an estimated inci-dence of 1 in 2500—3000 individuals (Ferner and Jackson,2011) and a minimum prevalence of 1 in 4—5000. Mutationsin NF1 on chromosome 17q11.2 result in loss of function ofneurofibromin, a guanosine triphosphatase—activating pro-tein and account for up to 95% of cases (Messiaen et al.,2000). Loss of neurofibromin results in increased prolifer-ation and tumorigenesis in neurocutaneous tissues (Yohay,2009; Williams et al., 2009). Increased expression of neu-rofibromin during embryonic development suggests that theNF1 gene may be important in the differentiation and matu-ration of cells in the central nervous system (CNS) (Gutmannet al., 2007). Abnormalities of NF1 gene may therefore berelated to developmental malformations, both in animalmodels and in humans (Gutmann et al., 2007; Bielschowskyand Rose, 1927; Rosman and Pearce, 1967). A neuropatho-logical study of the whole brain in ten NF1 patients withmental retardation and epilepsy found macroscopic disor-ders of cerebral architecture and subcortical heterotopia(Rosman and Pearce, 1967). Dysembryoplastic neuroepithe-lial tumours (DNETs), gangliogliomas (Parizel et al., 1991;Lellouch-Tubiana et al., 1998; Kannuki et al., 1996; Vinchonet al., 2000; Fedi et al., 2004; Rodriguez et al., 2008)hemimegalencephaly (Flores-Sarnat, 2002) and unspecifiedmalformations of cortical development (Balestri et al.,2003; Vivarelli et al., 2003) have been reported in somepatients.
Prevalence of epilepsy in patients with NF1 ranges from4 to 13% (Korf et al., 1993; Kulkantrakorn and Geller, 1998;Ferner and Jan, 2001). Seizures are usually well controlledwith antiepileptic medication (Korf et al., 1993) and havea better prognosis than they do in other neurocutaneoussyndromes (Kulkantrakorn and Geller, 1998; Motte et al.,1993). The first reports on epilepsy in NF1 failed to find anyconsistent relationship between structural lesions, if any,and the site of seizure onset in patients with complex par-tial seizures (Korf et al., 1993; Kulkantrakorn and Geller,1998). Créange et al. (1999) described five adult patientswith epilepsy associated with aqueductal stenosis, subduralhaematoma, cerebral hamartoma and meningioma, whileVivarelli et al. (2003) and Hsieh et al. (2011) reported on
a few NF1 patients with seizures caused by various typesof tumours. Most of these patients were well controlled bytreatment or seizure free after lesionectomy. Even patients
hac
erved.
ith hemimegalencephaly and NF1 have been reported toe well controlled (Flores-Sarnat, 2002). Conversely, MRIvidence of other malformations of cortical developmentuch as focal cortical dysplasia and polymicrogyria is usu-lly accompanied by drug-resistant epilepsy (Rosman andearce, 1967; Balestri et al., 2003; Vivarelli et al., 2003;hang et al., 2006).
The indications for, and outcome of, epilepsy surgery inatients with NF1, have not been specifically addressed buteserve in our opinion adequate attention. Type and aeti-logy of seizures associated to NF1 are protean and theirorrelation with MRI visible structural lesions are yet to belarified (Ferner and Jackson, 2011).
We report on 12 patients with NF1 and drug-esistant focal epilepsy who underwent epilepsy surgerynd describe their clinical presentation, outcome andistopathology.
ethods
atients
e conducted a survey amongst 25 European epilepsyurgery centres to collect observations of patients withF1 who had undergone surgery for drug-resistant seizures.he study was initiated in 2009 within the European Task-orce for Epilepsy Surgery in Children (UTASK) (Appendix) and was coordinated by the Children’s Hospital Meyer,lorence, Italy. Inclusion criteria were: (a) diagnosis ofF1 according to the presence of any 2 of the followinglinical features (National Institutes of Health Consensusevelopment Conference, 1988): (1) at least 6 café-au-
ait spots >1.5 cm after puberty and >0.5 mm in prepuberalndividuals; (2) intertriginous freckling; (3) at least 2 Lischodules; (4) at least 2 neurofibromas of any type or ateast 1 plexiform neurofibroma; (5) optic pathway gliomasOPGs); (6) distinctive bony lesions; and (7) a first-degreeamily relative with NF1; (b) previous surgery for pharma-oresistant epilepsy; (c) at least one year follow-up afterurgery.
istory and features, genetic testing, EEG, neuroimagingnd neuropsychological findings, surgical approach, out-ome and histopathology.
3
Cs
V
Tasislp
N
WicsFMaotma2
N
Wadsd
S
Acapa
H
Iwqwfie(
R
P
Wcytoites(d
mip(twt
pso
t(
Bi2pfitw
V
Atitn2o3a
N
Pwa
86
The ethical committee on human experimentation of thehildren’s Hospital ‘‘Meyer’’, Florence, Italy approved thetudy. Informed consent was obtained for all patients.
ideo-EEG recordings
he scalp interictal and ictal EEG and the seizure semiology,s well as invasive recordings, were included in the analy-is. Data concerning the extent of the epileptogenic zone.e. the neurophysiologically defined cortical area whereeizures are initiated (Luders et al., 1993) and the corre-ation with the epileptogenic lesion were obtained for allatients.
euroimaging
hen available, pre-operative and postoperative brain MRImages of all patients were reviewed in the coordinatingentre. Owing to differing acquisition parameters and MRIcanners, analysis was conducted of only T1, T2 and/orLAIR sequences. Both epileptogenic lesions and additionalRI abnormalities were described (Table 1). We defineds ‘‘dysplasia’’ all epileptogenic lesions characterized byne or more of the following features: abnormal corticalhickness, blurring of the grey/white matter junction, whiteatter signal abnormalities, abnormal gyral/sulcal pattern
nd focal and/or lobar hypoplasia/atrophy (Colombo et al.,009).
europsychology
hen available, pre and post-operative neuropsychologicalssessment was provided by the participating centres. Sinceifferent protocols were used, we included in the analy-is only cognitive level (IQ) and the description of specificeficits, if any.
urgery
ll centres provided a detailed description of the surgi-al procedure and assessed the completeness of resection,ccording to the presurgical evaluation protocol andost-operative MRI. The surgical outcome was evaluatedccording to Engel’s classification (Engel et al., 1993).
istopathology
n order to reduce interobserver variability, each specimenas first evaluated by the local neuropathologist and subse-uently reviewed by an independent neuropathologist who
as unaware of the previous report. Cases were classi-ed against the WHO classification of CNS tumours (Louist al., 2007) and the ILAE classification of cortical dysplasiaBlümcke et al., 2011).
g1e(
C. Barba et al.
esults
atients
e collected 12 patients with NF1 who had been surgi-ally treated for intractable seizures having at least oneear follow-up in eight of the 25 invited centres. Two cen-res failed to respond while no patient with NF1 had beenperated in the remaining invited centres. The 12 patientsncluded in the study were retrospectively identified throughhe local epilepsy surgery databases according to the studyntry criteria. They accounted for 0.5% of all the epilepsyurgery cases treated in the eight participating centresrange 0.08—1.6). The local epileptologist reviewed all theata and filled the questionnaire.
The study group consisted of eight females and fourales; mean age 30.4 years (±12.9). Six patients had a fam-
ly history of NF1. Genetic testing was performed in fouratients: mutation analysis by Sanger sequencing in threepatients 4, 6, 10) and FISH (fluorescence in situ hybridiza-ion) analysis of 17q11.2 in one (patient 7). Mutation analysisas positive in the three patients while FISH was uninforma-
ive.Pregnancy and delivery had been fully normal in nine
atients; two patients were born preterm and one haduffered asphyxia due to placental abruption. Early devel-pment had been delayed in three patients.
Neurological examination was normal in five patients;he remaining seven exhibited various neurological signsTable 1).
The mean age at seizure onset was 10.2 years (±10.8).efore surgery, all patients were pharmacoresistant accord-
ng to the ILAE definition of drug-resistance (Kwan et al.,010) and most were on polytherapy (Table 2). Elevenatients exhibited complex partial seizures, associated inve with secondary generalization. One patient had epilep-ic spasms. Seizure frequency was daily in five patients,eekly in five and monthly in two.
ideo-EEG recordings
ll patients underwent scalp video-EEG recordings to cap-ure seizures. Two patients also underwent extraoperativenvasive recordings (one SEEG, one subdural grids) andwo intraoperative electrocorticography. Seizures origi-ated from the temporal lobe in ten patients (patients 1,, 4, 5, 7—12); from the temporo-parieto-occipital region inne (patient 6) and were clearly bitemporal in one (patient). Seizure semiology and interictal and ictal EEG featuresre described in Table 1.
euroimaging
reoperative MRI was available for review in nine patients,hile for three patients (1, 11,12) a detailed report wasvailable. MRI abnormalities correlating with the epilepto-
enic zone were detected in all but one patient (patient). They were unilateral temporal in eight patients, bilat-ral temporal in one and multilobar or hemispheric in twoTable 1 and Fig. 1). Additional MRI abnormalities were
Epilepsy Surgery
In N
eurofibromatosis
Type 1
387
Table 1 Summary of information on the 12 patients included in the study.
Pt ID Sex Age at firstseizure(years)
Age atsurgery(years)
NE Cognitivelevel
Specificcognitivedeficits
Interictal EEG Ictal EEG Intracranialrecordings
Ictal semiology
1 Female 2 22 N Borderline Memory L T sharp andslow waves
L T Not performed Epigastric sensation,unresponsiveness, gesturalautomatisms, sialorrhea,post-ictal aphasia
2 Female 1.6 30 N N Language L T sharp andslow waves
L T SEEG: L T mesialictal onset zone
a) Motionless staring,oroalimentary automatisms,R arm dystonia and mouthdeviation to the R,post-ictal aphasia; b) R legand arm paresthaesia,possible SG; c) déjà-vu anddéjà-vecu
3 Male 19 35 N NA NA L T spikes andslow waves
BiT Not performed a) Motionless staring,oroalimentary automatismswith sialorrhea, bilateraldystonia, eye and headdeviation to the L, gesturalautomatisms; b) motionlessstaring, bilateral dystoniclimb posturing, lookingaround
4 Male 2 17 L hemiparesis Mildimpairment
Memory R C spikes R F Grid: R T ictalonset zone
a) Paleness, stiffening(L > R), generalized jerking;b) unresponsiveness; c)feeling odd or dizzy withblurred visionindependently or prior toseizures
5 Female 2.5 16 L hemiplegia Mildimpairment
Memory R T spikes R T EcoG: R T spikes Fear, oroalimentary andgestural automatisms
6 Female 0.4 6 R hemiparesis Mildimpairment
Language L FT spikes L FTO Not performed a) Motionless staring, eyeflutter, unresponsiveness; b)staring, R arm stiffening, Larm flexion, moaning,crying; c) drop attack
7 Female 19 24 N N None R FT sharpwaves
R FT EcoG: R anteriorand mesial Tspikes
Epigastric aura, motionlessstaring, blinking, head andeye deviation to the L, thento the R, unresponsiveness
388
C. Barba
et al.
Table 1 (Continued )
Pt ID Sex Age at firstseizure(years)
Age atsurgery(years)
NE Cognitivelevel
Specificcognitivedeficits
Interictal EEG Ictal EEG Intracranialrecordings
Ictal semiology
8 Female 0.6 31 R eye exotropia Severeimpairment
NA Bilateral CTPsharp waves(R > L)
R FT Not performed Head and eye deviation tothe L, bilateral dystonicarm posturing
9 Female 3 8 Clumsiness Borderline Language R FT spikes andsharp waves
R medial andposterior T
Not performed Spasms
10 Male 3 11 Clumsiness Mildimpairment
Memory L FT spikes andslow waves
L T Not performed Epigastric sensation, fear,oroalimentary and gesturalautomatisms,unresponsiveness
11 Male 30 39 N N Verbal memory R T sharp andslow waves
R T Not performed Awakening, motionlessstaring, vocalization,grimacing, then pedallingleg movements,unresponsiveness,oroalimentary and gesturalautomatisms, postictalnaming deficit
12 Female 4 39 Clumsiness Mildimpairment
Verbal memory L TP sharp andslow waves
L T Not performed Dizziness, oral and gesturalautomatisms, groaning,motionless staring, R handjerking, SG
Pt ID Epileptogeniclesions onbrain MRI
Additional MRIabnormalities
Surgicalprocedure
Histology 1 year seizureoutcome
2 yearseizureoutcome
5 yearseizureoutcome
Completenessof resection
1 None None L T lobectomy HS with granule celldispersion, whitematter heterotopias
Ia Ia NA Yes
2 L HS Severalsubcorticalhyperintensi-ties
L T lobectomy HS and granule celldispersion
Ia Ia Ia Yes
3 R operculo-insular andL T mesial-cysticlesions
None L T mesiallesionectomy
LT DNET and Ropercolo-insularlow-gradeastrocytoma
III IV NA Yes
4 RhemisphericPMG
Bilateralhippocampalhyperintensity
R Tcorticectomy
PMG and folialsclerosis with granulecell dispersion
IV IV IV No
Epilepsy Surgery
In N
eurofibromatosis
Type 1
389
Table 1 (Continued )
Pt ID Epileptogeniclesions onbrain MRI
Additional MRIabnormalities
Surgicalprocedure
Histology 1 year seizureoutcome
2 yearseizureoutcome
5 yearseizureoutcome
Completenessof resection
5 R HS Hydrocephaluswithventriculardilatation,aqueductalstenosis, basalgangliahamartoma
Figure 1 (A) Patient 7: right temporo-mesial solid/multicystic tumour (white arrow) (B) Patient 7: postoperative MRI (righttemporal lobectomy) (C) Patient 3: left temporo-mesial and (D) right insular lesions (white arrows) (E) Patient 10: left tem-poral hyperintensity (white arrow) (F) Patient 10: postoperative MRI (left temporal lobectomy) (G) Patient 9: right temporaldysplasia extending towards the temporo-occipital junction (white arrow) (H) Patient 9: postoperative MRI (right temporal cor-
nsity
p
t(str
S
Ffs(bdpsrfsi7(
H
Tcsscs
ticectomy+lesionectomy) (I) Patient 8: right temporal hyperintelobectomy).
observed in eight patients (Table 1). In particular threepatients (7, 10, 12) disclosed contralateral hippocampalhyperintensity. A patient with bitemporal seizures (patient3) had two different lesions one left temporo-mesial and oneright operculo-insular (see Table 1).
Post-operative MRI was performed in 11 patients. In theremaining patient (patient 4), the completeness of theresection was assessed on the basis of intracranial recordingswith knowledge of structural abnormality left in situ.
Neuropsychology
Preoperative neuropsychological assessment was availablein 11 patients: five exhibited mild cognitive impairment,one had severe impairment and five average or borderlinescores. Six patients exhibited memory deficits and three hadlanguage impairment (Table 1).
Postoperative assessment was available in nine patients:all exhibiting stable cognitive level, three had improved inspecific cognitive domains, such as attention and daily lifeactivities (patients 4, 6, 10). Four patients (patients 1, 7,11, 12) had experienced slight worsening of their memoryperformances.
Surgical strategy and techniques
Mean age at surgery was 23.2 ± 11.7 years old. Surgerywas tailored according to the results of presurgical evalu-ation protocol. In seven patients with epileptogenic lesions,highly concordant correlations allowed a single-step surgery.Also in the patient with normal brain MRI, invasive recor-dings were not performed, since scalp video-EEG recordingsclearly indicated a left temporal seizure onset zone. Inthe remaining four patients, surgery was tailored to the
results of invasive recordings. Overall, surgical proce-dures included seven lobectomies, one lesionectomy, twocorticectomy + lesionectomy, one corticectomy, one focalresection + disconnection (Table 1).
No major complications occurred in 11 patients; oneatient developed post-operative hydrocephalus.
The resection/disconnection of the lesion-related epilep-ogenic zone was judged to be complete in eight patientspatients 1, 2, 6, 7, 9—12). In the patient with bitemporaleizures (patient 3), a complete left temporal lesionec-omy was performed, while functional constraints preventedesection of the right temporo-insular lesion.
eizure outcome
ollow-up ranged from 1 year to 10 years. At one-yearollow-up eight patients (patients 1, 2, 6, 7, 9—12) wereeizure free (Engel’s class Ia) and three were unchangedEngel’s class IV; patients 4,5,8). In the patient withitemporal seizures (patient 3) left temporal seizures hadisappeared. For ten patients (patients 1—6, 8—11) a 2 yearost-operative follow up was available, indicating a stableurgical outcome in nine. Patient 3 relapsed and a reope-ation was planned. Three patients (patients 2, 4, 9) wereollowed for more than 5 years after surgery, with stableurgical outcome (Table 1). The treatment was withdrawnn six patients (patients 1, 2, 7, 9, 10, 12); one (patient) had seizure recurrence immediately after AED reductionTable 2).
istopathology
he pathology of the temporal lobe was reviewed in all 12ases and hippocampal tissue was available for review ineven. The hippocampus was abnormal in all seven caseshowing segmental neuronal loss either in a pattern oflassic/total hippocampal sclerosis (4 cases) or end folialclerosis (3 cases). Granule cell dispersion was present in 5
ut of the 7 cases.
Typical Dysembryoplastic Neuroepithelial Tumours (DNETf the ‘simple subtype’) were present in three whilewo further patients had cellular cortical glial infiltrates
392
Table 2 Summary of information on the medicaltreatment.
ithout specific diagnostic features, raising the possibilityf a diffuse/non-specific subtype of DNET (Fig. 2).
Three individuals exhibited malformations of corticalevelopment including polymicrogyria (1 case), white mat-er heterotopia (1 case) and a complex mixed malformationith features of focal cortical dysplasia (type IIa, 1 case).
Mixtures of pathology were present in five cases. Sec-ndary and non-specific changes were common, e.g. subpialChalin’s) gliosis. In particular nine cases showed prominenthanges in the white matter vessels. These vessels had large
erivascular spaces that contained pigment (lipofuscin)-aden macrophages. The changes resembled those of thehite matter angiopathy described by Hildebrandt et al.
2008).
cH
g
igure 2 The images show examples of the range of neuropatholof a dysembryoplastic neuroepithelial tumour H&E; (B) Patient 2 shlassical hippocampal sclerosis. There is neuronal loss in CA1, CA4 anLFB/Nissl); (C) Patient 4 showing granule cell dispersion and foliaatient 4); (D) Patient 12 showing subpial (Chaslin’s) gliosis lining
ranching sulci in a region of polymicrogyria (LFB/Nissl)(see also panitting within the white matter (LFB/Nissl). (G) Patient 9 showinghowing a focus of perivascular pigment-laden cells in the white ma
C. Barba et al.
In patient 3 stereotactic biopsy of the right temporo-nsular lesion revealed a low-grade astrocytoma.
The interobserver agreement rate was high (10/12atients: 83%). Discordant diagnoses (patients 10, 12) areeported in Table 1.
iscussion
he 12 patients with NF1 and drug-resistant focal seizurese are reporting underwent surgery for epilepsy. Eight of
hem were seizure free at one-year follow-up; one hadorthwhile improvement and the remaining three did notenefit from the procedure. In patients for whom follow-upata at 2 and 5 years were available, postoperative out-ome had remained stable. Surgical failures were relatedo incomplete resection of the lesion-related epileptogenicone as assessed on post-operative MRI, likewise reported inll surgical series (Spencer and Huh, 2008).
Although multiple, potentially epileptogenic lesionsncrease the risk of failures, patients in whom a sin-le epileptogenic zone is identified can still benefit frompilepsy surgery, as demonstrated, for example, in tuber-us sclerosis (Wu et al., 2010). In one of our patientsxperiencing surgical failure, a DNET was associated with
contralateral astrocytoma, both lesions appearing highlypileptogenic. Previous reports had pointed out thatpilepsy surgery in patients with NF1 yields inadequateesults if a glioneuronal lesion is associated with additionalotentially epileptogenic lesions (Lellouch-Tubiana et al.,998; Fedi et al., 2004). Our results indicate that operatingatients with NF1 on a single epileptogenic lesion entails anutcome similar to that of the general population of surgical
andidates with a single epileptogenic lesion (Spencer anduh, 2008).
In this study, all patients but one harboured epilepto-enic lesions that corresponded to, or were included in,
gy seen in the patients: (A) Patient 7 showing typical featuresowing a low power image of the hippocampus with features ofd CA3 and preservation in CA2. There is granule cell dispersionl sclerosis (DG-Dentate Gyrus, LFB/Nissl)(see also panel E forthe margins of a sulcus (GFAP). (E) Patient 4 showing fused,el E for patient 4). (F) Patient 8 showing neuronal heterotopias
a region normally-formed cortex (Nissl/LFB). (H) Patient 12tter H&E.
bdtG
ic2hpmds
C
Tdiuanpmapnet
sltos
A
ToC
mi
a
A
Sfj
R
Epilepsy Surgery In Neurofibromatosis Type 1
the epileptogenic zone. Previous reports on epilepsy in NF1had stressed the lack of a consistent relationship betweenMRI visible structural lesion and epileptogenic zone (Korfet al., 1993; Kulkantrakorn and Geller, 1998). However,these studies included large series of patients who had notbeen specifically explored with the purpose of surgical treat-ment of epilepsy (Korf et al., 1993; Kulkantrakorn and Geller,1998).
Patients with NF1 exhibit a wide range of structurallesions, the most frequent of which, unidentified brightobjects (UBOs) and tumours, do not show a preferentiallobar localization (Hsieh et al., 2011; Braffman et al., 1988;Korf et al., 1999; Alkan et al., 2005). In all patients but onein our series, the epileptogenic lesion was located in thetemporal lobe. In previous reports on epilepsy surgery in NF1most operated patients had extratemporal lesions (Lellouch-Tubiana et al., 1998; Kannuki et al., 1996; Fedi et al.,2004; Vivarelli et al., 2003). Although a selection bias ispossible, as patients with temporal lobe epilepsy are habit-ually overrepresented in epilepsy surgery series (Spencerand Huh, 2008), the high prevalence of temporal lobeepilepsy in our study might be aetiology related as DNETs aremore frequently found in the temporal lobe (Englot et al.,2012).
Four out of the 12 patients had hippocampal sclerosis.However, mesial temporal lobe sclerosis is the most com-mon indication for surgery in adults (Spencer and Huh, 2008)and its high prevalence in this series is likely due to selectionbias related to the surgical objective of the study. Hippocam-pal hyperintensities are frequently observed in patients withNF1 (Hsieh et al., 2011; Alkan et al., 2005), but it is unclearwhether they correspond to hippocampal sclerosis or not.In our series, three patients who were seizure free aftertemporal lobectomy disclosed contralateral hippocampalhyperintensity.
The high prevalence of DNETs in this series might sug-gest that the association of this developmental tumour withNF1 is not fortuitous. Neurofibromin, the protein productof the NF1 gene, acts as a tumour suppressor and as aconsequence of NF1 gene inactivation individuals with NF1develop tumours with increased frequency (Williams et al.,2009; Rodriguez et al., 2008; Bajenaru et al., 2003). Thelargest available neuropathological series including 100 NF1patients (Rodriguez et al., 2008) featured 49% pylociticastrocytomas, 27% diffuse astrocytomas, 17% low-gradeastrocytomas and only 2% gangliogliomas. DNETs account for14—18% of histologic diagnoses in epilepsy surgery series(Pasquier et al., 2002; Luyken et al., 2003) but they havebeen reported only anecdotically in isolated patients withNF1 who had undergone surgery for epilepsy (Lellouch-Tubiana et al., 1998; Kannuki et al., 1996; Fedi et al.,2004). The 40% prevalence of DNETs in our series has possiblyemerged as a consequence of its being the largest epilepsysurgery series in NF1 so far. The prevalence of pharmacore-sistant epilepsy in NF1 is low and the number of patientsundergoing surgical treatment, and histopathological stud-ies, is consequently small (0.5% of all epilepsy surgery casesin our study).
Interobserver concordance among neuropathologists washigh (10 of 12 patients), when compared to previous reportson intra and interobserver reproducibility in the classifica-tion of dysplastic lesions (Chamberlain et al., 2009) and
A
393
rain tumours (Gilles et al., 2008), The only two discordantiagnoses might be related to different classification atti-udes among neuropathologists (Chamberlain et al., 2009;illes et al., 2008).
Sixty-six percent of our patients were cognitivelympaired. Approximately 50% of patients with NF1 haveognitive impairment (Williams et al., 2009; North et al.,002). Co-occurrence of NF1 and refractory epilepsy mightave negatively influenced the cognitive profile. In the nineatients who had post-operative neuropsychological assess-ent, cognitive level remained stable and attention andaily life activities improved in three. Memory performanceslightly worsened in four.
onclusions
his study has some limitations. The sample size was smallue to the low prevalence of pharmacoresistant epilepsyn NF1. The eight participating epilepsy surgery centrestilized different selection criteria for surgical candidacynd data were collected retrospectively using heteroge-eous methods of ascertainment. Although these limitationsrevented us to draw definite conclusions, our resultsight suggest that intractable epilepsy in NF1 can be
menable to successful and safe surgical treatment. Theresence of bilateral MRI abnormalities and pre-existing cog-itive impairment does not represent a contraindication topilepsy surgery provided a single and well-delimited epilep-ogenic zone is recognized.
The low number of analyzed specimens and the epilepsyurgery objective of the study may have biased hystopatho-ogical results. However, the high prevalence of DNETs inhis series might suggest that the association of this devel-pmental tumour with NF1 is not fortuitous. Further largertudies are needed to confirm these findings.
cknowledgements
his study was made possible by the network organizationf the UTASK, European Taskforce for Epilepsy Surgery inhildren.
The authors report no conflict of interest concerning theaterials or methods used in this study or the findings spec-
fied in this paper.This research received no specific grant from any funding
gency in the public, commercial or not-for-profit sectors.
ppendix A. Supplementary data
upplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/.eplepsyres.2013.02.021
eferences
lkan, A., Sigirci, A., Kutlu, R., Ozcan, H., Erdem, G., Aslan, M.,Ates, O., Yakinci, C., Egri, M., 2005. Neurofibromatosis type 1:diffusion weighted imaging findings of brain. Eur. J. Radiol. 56,229—234.
ajenaru, M.L., Hernandez, M.R., Perry, A., Zhu, Y., Parada, L.F.,Garbow, J.R., Gutmann, D.H., 2003. Optic nerve glioma in micerequires astrocyte Nf1 gene inactivation and Nf1 brain heterozy-gosity. Cancer Res. 63, 8573—8577.
alestri, P., Vivarelli, R., Grosso, S., Santori, L., Farnetani, M.A.,Galluzzi, P., Vatti, G.P., Calabrese, F., Morgese, G., 2003. Mal-formations of cortical development in Neurofibromatosis type 1.Neurology 61, 1799—1801.
ielschowsky, M., Rose, M., 1927. Zur Kenntnis der zentralenVeranderungen bei Recklinghausenschen Krankheit. J. Psychol.Neurol. 35, 42.
lümcke, I., Thom, M., Aronica, E., Armstrong, D.D., Vinters,H.V., Palmini, A., Jacques, T.S., Avanzini, G., Barkovich, A.J.,Battaglia, G., Becker, A., Cepeda, C., Cendes, F., Colombo, N.,Crino, P., Cross, J.H., Delalande, O., Dubeau, F., Duncan, J.,Guerrini, R., Kahane, P., Mathern, G., Najm, I., Ozkara, C.,Raybaud, C., Represa, A., Roper, S.N., Salamon, N., Schulze-Bonhage, A., Tassi, L., Vezzani, A., Spreafico, R., 2011. Theclinicopathologic spectrum of focal cortical dysplasias: a con-sensus classification proposed by an ad hoc Task Force of theILAE Diagnostic Methods Commission. Epilepsia 52, 158—174.
raffman, B.H., Bilaniuk, L.T., Zimmerman, R.A., 1988. The cen-tral nervous system manifestations of the phakomatoses on MR.Radiol. Clin. N Am. 26, 773—800.
hamberlain, W.A., Cohen, M.L., Gyure, K.A., Kleinschmidt-DeMasters, B.K., Perry, A., Powell, S.Z., Qian, J., Staugaitis,S.M., Prayson, R.A., 2009. Interobserver and intraobserverreproducibility in focal cortical dysplasia (malformations of cor-tical development). Epilepsia 50, 2593—2598.
hang, B.S., Apse, K.A., Caraballo, R., Cross, J.H., Mclellan, A.,Jacobson, R.D., Valente, K.D., Barkovich, A.J., Walsh, C.A.,2006. A familial syndrome of unilateral polymicrogyria affectingthe right hemisphere. Neurology 66, 133—135.
olombo, N., Salamon, N., Raybaud, C., Ozkara, C., Barkovich,A.J., 2009. Imaging of malformations of cortical development.Epileptic. Disord. 11, 194—205.
réange, A., Zeller, J., Rostaing-Rigattieri, S., Brugières, P., Degos,J.D., Revuz, J., Wolkenstein, P., 1999. Neurological compli-cations of neurofibromatosis type 1 in adulthood. Brain 122,473—481.
ngel Jr., J., Van Ness, P.C., Rasmussen, T.B., Ojemann, L.M., 1993.Outcome with respect to epileptic seizures. In: Engel Jr., J.
(Ed.), Surgical Treatment of the Epilepsies. , second ed. RavenPress, New York, pp. 609—621.
nglot, D.J., Berger, M.S., Barbaro, N.M., Chang, E.F., 2012. Fac-tors associated with seizure freedom in the surgical resection ofglioneuronal tumors. Epilepsia 53, 51—57.
edi, M., Mitchell, L.A., Kalnins, R.M., Gutmann, D.H., Perry, A.,Newton, M., Brodtmann, A., Berkovic, S.F., 2004. Glioneuronaltumours in neurofibromatosis type 1: MRI-pathological study. J.Clin. Neurosci. 11, 745—747.
erner, R.E., Jackson, M., 2011. Neurofibromatoses. In: Shorvon,S.D., Andermann, F., Guerrini, R. (Eds.), The Causes ofEpilepsy: Common and Uncommon Causes in Adults and Children.Cambridge University Press, Cambridge, pp. 183—188.
erner, R.E., Jan, W., 2001. Epilepsy in neurofibromatosis. J. Neu-rol. Sci. 819, S280.
lores-Sarnat, L., 2002. Hemimegalencephaly: part 1. Genetic,clinical, and imaging aspects. J. Child. Neurol. 17, 373—384.
illes, F.H., Tavaré, C.J., Becker, L.E., Burger, P.C., Yates, A.J.,Pollack, I.F., Finlay, J.L., 2008. Pathologist interobserver vari-ability of histologic features in childhood brain tumors: resultsfrom the CCG-945 study. Pediatr. Dev. Pathol. 11, 108—117.
utmann, D.H., Zhan, Y., Hirbe, A., 2007. Developmental regu-
lation of a neuron-specific neurofibromatosis 1 isoform. Ann.Neurol. 46, 777—782.
ildebrandt, M., Amann, K., Schröder, R., Pieper, T., Kolodziejczyk,D., Holthausen, H., Buchfelder, M., Stefan, H., Blümcke, I.,
R
C. Barba et al.
2008. White matter angiopathy is common in pediatric patientswith intractable focal epilepsies. Epilepsia 49, 804—815.
sieh, H.Y., Fung, H.C., Wang, C.J., Chin, S.C., Wu, T., 2011.Epileptic seizures in neurofibromatosis type 1 are related tointracranial tumors but not to neurofibromatosis bright objects.Seizure 20, 606—611.
annuki, S., Bando, K., Soga, T., Matsumoto, K., Matsumoto, K.,Hirose, T., 1996. A case report of dysembryoplastic neuroepithe-lial tumour associated with neurofibromatosis type 1. No ShinkeiGeka 24, 183—188.
orf, B.R., Carrazana, E., Holmes, G.L., 1993. Patterns of seizuresobserved in association with neurofibromatosis 1. Epilepsia 34,616—620.
orf, B.R., Schneider, G., Poussaint, T.Y., 1999. Structural anoma-lies revealed by neuroimaging studies in the brains of patientswith neurofibromatosis type 1 and large deletions. Genet. Med.1, 136—140.
ulkantrakorn, K., Geller, T.J., 1998. Seizures in neurofibromatosis1. Pediatr. Neurol. 19, 347—350.
wan, P., Arzimanoglou, A., Berg, A.T., Brodie, M.J., Allen Hauser,W., Mathern, G., Moshé, 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.
ellouch-Tubiana, A., Bourgeois, M., Vekemans, Robain, O.,1998. Dysembryoplastic neuroepithelial tumors in two chil-dren with neurofibromatosis type 1. Acta Neuropathol. 90,319—322.
ouis, D.N., Ohgaki, H., Wiestler, O.D., Cavenee, W.K., Burger, P.C.,Jouvet, A., Scheithauer, B.W., Kleihues, P., 2007. The 2007 WHOclassification of tumours of the central nervous system. ActaNeuropathol. 114, 97—109.
uders, H.O., Engel, J., Munari, C., 1993. General principles. In:Engel, J. (Ed.), Surgical Treatment of the Epilepsies. RavenPress, NewYork, pp. 137—153.
uyken, C., Blumcke, I., Fimmers, R., Urbach, H., Elger,C.E., Wiestler, O.D., Schramm, J., 2003. The spectrumof long-term epilepsy-associated tumors: long-term seizureand tumor outcome and neurosurgical aspects. Epilepsia 44,822—830.
essiaen, L.M., Callens, T., Mortier, G., Beysen, D., Vandenbroucke,I., Van Roy, N., Speleman, F., Paepe, A.D., 2000. Exhaustivemutation analysis of the NF1 gene allows identification of 95%of mutations and reveals a high frequency of unusual splicingdefects. Hum. Mutat. 15, 541—555.
otte, J., Billard, C., Fejerman, N., Sfaello, Z., Arroyo, H., Dulac,O., 1993. Neurofibromatosis type one and West syndrome: arelatively benign association. Epilepsia 34, 723—726.
ational Institutes of Health Consensus Development Conference,1988. Neurofibromatosis. Conference statement. Arch. Neurol.45, 575—578.
orth, K., Hyman, S., Barton, B., 2002. Cognitive deficits in neu-rofibromatosis 1. J. Child Neurol. 17, 605—612.
arizel, P.M., Martin, J.J., Van Vyve, M., van den Hauwe, L.,De Schepper, A.M., 1991. Cerebral ganglioglioma and neurofi-bromatosis type I. Case report and review of the literature.Neuroradiology 33, 357—359.
asquier, B., Peoc, H.M., Fabre-Bocquentin, B., Bensaadi, L.,Pasquier, D., Hoffmann, D., Kahane, P., Tassi, L., Le Bas, J.F.,Benabid, A.L., 2002. Surgical pathology of drug-resistant partialepilepsy. A 10-year-experience with a series of 327 consecutiveresections. Epileptic. Disord. 4, 99—119.
odriguez, F.J., Perry, A., Gutmann, D.H., Leonard, J., Bryant, S.,Giannini, C., 2008. Gliomas in neurofibromatosis type 1: a clini-
copathologic study of 100 patients. J. Neuropathol. Exp. Neurol.67, 240—249.
osman, N.P., Pearce, J., 1967. The brain in multiple neu-rofibromatosis (von Recklinghausen’s disease): a suggested
