Local and remote epileptogenicity in focal cortical dysplasias and neurodevelopmental tumours

doi:10.1093/brain/awp242

Brain 2009: 132; 3072–3086

| 3072

BRAIN A JOURNAL OF NEUROLOGY

Local and remote epileptogenicity in focal cortical dysplasias and neurodevelopmental tumours Sandrine Aubert,1,2 Fabrice Wendling,3,4 Jean Regis,5 Aileen McGonigal,1,2 Dominique Figarella-Branger,6 Jean-Claude Peragut,5 Nadine Girard,7 Patrick Chauvel1,2,8 and Fabrice Bartolomei1,2,8 1 2 3 4 5 6 7 8

CHU Timone, Service de Neurophysiologie Clinique, Marseille, F-13000, France Universite´ de la Me´diterrane´e, LNN, Marseille, F-13000, France INSERM, U642, Rennes, F-35000, France Universite´ de Rennes 1, LTSI, Rennes, F-35000, France CHU Timone, Service de Neurochirurgie Fonctionnelle, Marseille, F-13000, France CHU Timone, Service d’Anatomopathologie, Marseille, F-13000, France CHU Timone, Service de Neuro-Radiologie, Marseille, F-13000, France INSERM, U751, Marseille, F-13000, France

Correspondence to: Pr Fabrice Bartolomei, MD, PhD, Service de Neurophysiologie Clinique, CHU Timone-264 Rue st Pierre, 13005-Marseille, France E-mail: [email protected]

During the pre-surgical evaluation of drug-resistant epilepsy, the assessment of the extent of the epileptogenic zone and its organization is a crucial objective. Indeed, the epileptogenic zone may be organized as a simple focal lesional site or as a more complex network (often referred to as the ‘epileptogenic network’) extending beyond the lesion. This distinction is particularly relevant in developmental lesions such as focal cortical dysplasias or dysembryoplastic neuroepithelial tumours and may determine both the surgical strategy and the prognosis. In this study, we have quantified the epileptogenic characteristic of brain structures explored by depth electrodes in 36 patients investigated by stereoelectroencephalography and suffering from focal drug-resistant epilepsy associated with focal cortical dysplasias or dysembryoplastic neuroepithelial tumours. This quantification was performed using the ‘Epileptogenicity Index’ method that accounts for both the propensity of a brain area to generate rapid discharges and the time for this area to get involved in the seizure. Epileptogenicity Index values range from 0 (no epileptogenicity) to 1 (maximal epileptogenicity). We determined Epileptogenicity Index from signals recorded in distinct brain structures including the lesional site. We studied the type of epileptogenic zone organization (focal versus network) and looked for a correlation with clinical data and post-surgical outcome. Mean Epileptogenicity Index in lesional regions was 0.87 (0.25), and 0.29 (0.30) in ‘non-lesional’ structures. The number of highly epileptogenic structures (defined by Epileptogenicity Index value 50.4) was 3.14 (1.87) in the whole population. We found that 31% of patients had only one epileptogenic structure (NEI50.4 = 1), therefore disclosing a strictly focal epileptogenic zone organization while 25 patients had more than one epileptogenic region, disclosing a network (61%) or bilateral (8%) epileptogenic zone organization. We observed a trend for a difference in seizure outcome according to the type of epileptogenic zone organization. Indeed, 57% of patients with network organization and 87% with focal organization were seizure-free while none of those with bilateral organization became seizure-free. The determination of Epileptogenicity Index computed from electrophysiological signals recorded according to the stereoelectroencephalography technique is a novel tool. Results suggest that it can help in the delineation of the epileptogenic zone associated with brain lesions and that it could be used in the definition of the subsequent surgical resection.

Received April 26, 2009. Revised July 6, 2009. Accepted August 16, 2009. Advance Access publication September 21, 2009 ß The Author (2009). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: [email protected]

Quantification of brain epileptogenicity

Brain 2009: 132; 3072–3086

| 3073

Keywords: partial epilepsy; rapid discharge; focal cortical dysplasia; intracerebral recordings; prognosis Abbreviations: DNET = dysembryoplastic neuroepithelial tumour; EI = Epileptogenicity Index; FCD = focal cortical dysplasia; MTLE = mesio-temporal lobe epilepsy; NDT = neurodevelopmental tumours; SEEG = stereoelectroencephalography

Introduction Despite the large number of available antiepileptic drugs, the proportion of patients with focal epilepsy with uncontrolled seizures remains constant at around 20–30% (French, 2007). In focal drug-resistant epilepsy, surgery can lead to the suppression of seizures in some patients (Engel, 1998; Prayson, 2000; Wiebe et al., 2001). Pre-surgical evaluation of patients candidate to surgery consists of several steps (Bartolomei et al., 2002) that aim at determining the epileptogenic zone, historically defined as the brain region(s) that generate seizures (Talairach and Bancaud, 1966). The objective of epilepsy surgery is indeed to suppress seizures by removing the epileptogenic zone. Invasive EEG recordings may be required to accurately determine the epileptogenic zone. In most cases, this type of electrophysiological investigation reveals that the seizure onset is characterized by the occurrence of high frequency, low amplitude rapid discharges in the beta/gamma range (Bancaud et al., 1965; Allen et al., 1992; Worrell et al., 2002, 2004; Wendling et al., 2003; Bartolomei et al., 2004, 2008). From these recordings, the definition of the epileptogenic zone is thus classically based on two criteria: the appearance of rapid discharges and the delay of involvement, in the ictal process, of explored brain structures. It has been shown that rapid discharges may be visible in several, anatomically separate structures during the period of seizure onset, raising the possibility of a network organization of the epileptogenic zone (Bancaud et al., 1970; Bartolomei et al., 2004, 2008). The definition of the epileptogenic zone is however problematic in these cases since no precise criteria exist to define the characteristics of the initial discharge in terms of its frequency and the delay of its appearance in a given structure. These aspects are indeed traditionally assessed simply by visual inspection of the stereoelectroencephalography (SEEG) recordings. Recently, we proposed a new method for quantifying the epileptogenicity of brain structures in patients explored with depth electrodes and SEEG. A first study was conducted in patients with medial temporal lobe epilepsy (MTLE) associated with hippocampal sclerosis or normal MRI (Bartolomei et al., 2008). It was based on the determination of the ‘Epileptogenicity Index’ (EI) that combines both spectral and temporal parameters related to the propensity of a brain area to generate rapid discharges and the time for this area to become involved in the seizure process (Bartolomei et al., 2008). This index ranging from 0 (no epileptogenicity) to 1 (maximal epileptogenicity) allowed us to quantify the epileptogenicity of each studied brain structure. This previous study showed that the epileptogenic zone in MTLE disclosed features of an epileptogenic network involving several distinct structures. The EI values also allowed us to estimate the extent of this network. Finally, it was concluded that the duration of epilepsy was correlated with the number of epileptogenic structures in MTLE. These results

advocated for a gradual extension of brain epileptogenicity over time in these epilepsies. In the present study, although we used a similar methodological approach based on the EI, the clinical context is different. We aimed at analysing the organization of the epileptogenic zone in partial epilepsies associated with neurodevelopmental tumours (NDT) or focal cortical dysplasia (FCD). These cortical malformations are frequently found in cases of pharmacoresistant epilepsy referred for surgical consideration (Wolf et al., 1993; Honavar et al., 1999; Pasquier et al., 2002). FCD accounts for 20–25% cases of patients with focal epilepsy (Kuzniecky et al., 1993; Tassi et al., 2002). Currently available high-resolution MRI techniques have improved detection of FCD compared to past decades, but small lesions are still quite difficult to detect (Duncan, 1997; Lee et al., 2001; McGonigal et al., 2007). NDT represent the second most frequent group of lesions found in focal pharmacoresistant epilepsy. These tumours correspond to dysembryoplastic neuro-epithelial tumours (DNET) and gangliogliomas. These are congenital lesions that probably develop during embryogenesis, and share some clinical, radiological and neuropathological features. For NDT, most series have found a prevalence between 15% and 30% of all operated patients with focal pharmacoresistant epilepsy (Daumas-Duport, 1993, 1999; Prayson et al., 1993). Data from intracranial EEG including SEEG and electrocorticography (ECoG) in human patients have confirmed that FCDs are intrinsically epileptogenic (Palmini et al., 1995; Gambardella et al., 1996; Chassoux et al., 2000; Ferrier et al., 2001; Tassi et al., 2001). However, in many cases, it is established that the epileptogenic zone extends beyond the structural lesion seen on MRI (Chassoux et al., 2000; Colombo et al., 2003). Therefore, such cases provide an opportunity to study the relationship between focal brain lesions and the regions of seizure generation. This study was performed to assess potential links between the epileptogenic zone organization (as characterized by EI values) and certain clinical, neuroradiological or histopathological parameters, and to determine, among these potential factors, those that may influence post-operative prognosis.

Materials and methods Patient selection and SEEG recordings Among 138 consecutive patients who underwent SEEG in our centre between 2000 and 2007, 36 patients with focal pharmacoresistant epilepsy and neuroradiological findings suggestive of FCD or neurodevelopmental tumour (dysembryoplastic neuroepithelial tumour, ganglioglioma) were selected. We also included patients with nonlesional MRI in whom the surgical specimen pathological examination led to the diagnosis of FCD (Table 1).

F F F M F M F F F F M M F F M F

M M M F M

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 33 34 35 36

1.5 12 6 2.5 3.5

5 5 3 1.5 16 16 17 8 4 7 6 2 5 3.5 7 14

4 1.5 3 4.5 0.5 14 12 5 2.5 10 2 5 5 12 12

22 11 10 38 12

21 15 14 7 11 17 11 9 25 16 25 27 20 12 31 2

18 5 12 27 28 36 22 20 32 26 45 12 3 27 23

Y Y Y Y Y

Y Y N Y Y N N Y Y N N Y Y Y Y Y

Y Y Y Y Y Y Y Y Y Y N N Y Y N

No (surgery contra-indicated) No (not yet operated on) GK GK GK R parietal corticectomy R parietal corticectomy R lateral temporal corticectomy R temporal lobectomy R parietal corticectomy R frontal corticectomy L prefontal corticectomy R temporal lobectomy R premotor corticectomy L prefrontal and orbitofrontal corticectomy HS/R hippocampus R temporal lobectomy FCD/L SMA R frontal corticectomy – L premotor corticectomy FCD/R superior frontal sulcus R prefrontal corticectomy FCD/L anterior cingulate gyrus lL prefrontal corticectomy – R frontal corticectomy – R premotor corticectomy FCD/R occipito-temporal region R occipito-temporal corticectomy FCD/L anterior frontal region L prefrontal corticectomy – R lateral occipito-temporal corticectomy – R temporo-frontal corticectomy FCD/R occipital region R temporo-posterior corticectomy FCD/L premotor cortex L frontal corticectomy FCD/L mesial parietal region L premotor and cingular corticectomy FCD/L temporo-parieto-occipital region L occipito-temporal corticectomy DNET/R middle temporal gyrus No (surgery proposed to the patient but refused) FCD/L temporomesial region L temporal lobectomy DNET/R superior temporal gyrus R temporal lobectomy Ganglioglioma/L temporomesial region L temporal lobectomy FCD/L mesial occipital region L occipito-temporal corticectomy FCD/L parietal operculum Lesionectomy

FCD/R precentral region FCD/R temporo-polar region FCD/R central region FCD/R parieto-occipital region FCD/L precentral region FCD/R intraparietal sulcus FCD/R parietal region FCD/R superior temporal gyrus FCD/R hippocamus FCD/R parieto-occipital region – – FCD/R temporopolar region and insula FCD/R prefrontal region –

Surgery

N Y Y Y N

Y Y Y Y Y Y Y Y Y Y Y N N N NA –

– – – – – Y Y Y Y Y Y N N NA Y

NTW F NTW B F

F F NTW F NTW NTW NTW NTW NTW F NTW B NTW NTW F F

NTW NTW NTW B F NTW F F NTW NTW NTW NTW NTW NTW NTW

NDT NDT NDT NDT NDT

(ganglioglioma*) (DNET*) (DNET*) (DNET*) (ANET *)

(*) (*) (*) (*) (*) (*) (*) (*) (*) (*) (*) (*) (*) (*) (*)

(gliosis*) (gliosis*) Ia (*) Ia (*) Ib (*) Ib (*) Ib (*) Ib (*) Ib (*) IIa (*) FCD IIa FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb FCD IIb NDT

FCD FCD FCD FCD FCD FCD FCD FCD FCD FCD FCD FCD FCD FCD FCD

Complete Epileptogenic Final diagnosis removal zone organization (*confirmed by histopathology)

5.5 3.5 6 1 2

2 5.5 2.5 4 7 5.5 1.5 4 5 6.5 7 5 3 7 NA –

– – 5.5 6 – 1.5 2 6.5 4 1 6.5 4 2 NA 1

II Ia Ia IV IV

Ib Ia Ia Ia Ia Ia Ia II II II II III III IV NA –

– IV IV IV Ia Ia Ia Ia Ia Ia III IV NA Ia

Follow-up Engel (years) class

| Brain 2009: 132; 3072–3086

F = female; M = male; N = no; Y = yes; DNET = dysembryoplastic neuroepithelial tumour; FCD = focal cortical dysplasia; HS = hippocampal sclerosis; L = left; R = right; GK = gamma-knife; B = bilateral epileptogenic zone organization; F = focal epileptogenic zone organization; NTW = network epileptogenic zone organization; NA = not applicable; SO = surgical outcome (Engel’s class); *final pathological diagnosis confirmed by histopathology.

F M M F M F F F F F M F M M F

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Epilepsy Lesion Lesion Patient Sex Age at type/ epilepsy duration on localization MRI (years) onset

Table 1 Clinical, neuroradiological and surgical outcome in the studied population

3074 S. Aubert et al.

Quantification of brain epileptogenicity Neuro-radiological findings suggestive of FCD were defined as follows: local thickening of a part of the cortex, blurring of the grey matter to white matter interface, and increased signal (in Fluid Attenuated Inversion Recovery and T2-weighted sequences) in the underlying white matter (Barkovich et al., 1988; Sisodiya, 2000; Lee et al., 2001; Urbach et al., 2002; Colombo et al., 2003; WiddessWalsh et al., 2005), sometimes in a wedge-shaped tail extending radially to the ventricle (‘transmantle dysplasia’) (see example in Fig. 2b). A lesion being located in extra-temporal areas was a supplementary argument in favour of this diagnosis (Palmini et al., 1991, 1994; Hirabayashi et al., 1993; Chassoux et al., 2000). Diagnosis of NDT could be evoked if particular features were present: a well-demarcated, cortical (rarely both cortical and subcortical), cystic lesion, sometimes heterogeneous, enhanced or not by contrast injection, showing calcifications or adjacent bone erosion (DaumasDuport, 1993; Raymond et al., 1995; Ostertun et al., 1996;

Brain 2009: 132; 3072–3086

| 3075

Chassoux et al., 2000; Stanescu Cosson et al., 2001); most of the time, NDT are located in the temporal lobe (Kirkpatrick et al., 1993; Raymond et al., 1995; Degen et al., 2002; Shin et al., 2002; Kral et al., 2003) (see example in Figs 1b and 3b). All patients had a comprehensive evaluation including detailed history and neurological examination, neuropsychological testing, routine MRI, surface electroencephalography (EEG) and stereoelectroencephalography (SEEG, depth electrodes). SEEG exploration was carried out during long-term video-EEG monitoring, as part of patients’ normal clinical care. Recordings were performed using intracerebral multiple contact electrodes (10–15 contacts, length: 2 mm, diameter: 0.8 mm, 1.5 mm apart) placed intracerebrally according to Talairach’s stereotactic method (Talairach et al., 1992). In this study, all patients required invasive recordings after the non-invasive phase. In these cases (including patients with temporal lobe lesions), SEEG was indicated when the epileptogenic zone was

Figure 1 Patient no 36. (A) Sketch of the SEEG exploring the left hemisphere. Each electrode contains 10–15 contacts. Electrode OF0 investigated the opercular frontal region (external contacts) and the anterior insula (internal contacts). Electrode L0 was placed in the parietal operculum exploring a probable NDT, visible on the MRI (B). The route of electrode L0 is shown on the MRI passing through the lesion. P0 : parietal electrode reached the posterior part of the cingulate gyrus. T0 recorded the superior temporal gyrus (lateral contacts) and the insular cortex (temporal part, internal contacts). A0 recorded the middle temporal gyrus (lateral contacts) and the amygdala (internal contacts). (B) MRI (Flair sequence) showing the lesion (NDT) in the left parietal operculum. (C) SEEG recordings from three regions: L0 9–10: bipolar derivation from the lesional site (L0 ); T0 7–8: bipolar derivation from the superior temporal gyrus (STG), P0 7–8: bipolar derivation from the parietal region (P0 ). Below are represented the increase in energy ratio ER[n] in the three SEEG signals from L, STG and P. The Page–Hinkley algorithm provides a detection time (orange marks) for each brain structure if involved in the generation of a rapid discharge. The first detection time is arbitrarily defined as the reference time N0 (L in this case). (D) EI values obtained in seven regions in this patient. Maximal values are observed in L and low values in other regions, corresponding to a very focal onset.

3076

| Brain 2009: 132; 3072–3086

suspected to be larger than the lesion and/or when extratemporal regions were suspected to be involved in temporal lobe seizures (temporoperisylvian or temporal plus seizures) and/or when bilaterality was suspected. The placement of electrodes was established, in each patient, based upon available non-invasive information providing hypotheses about the localization of the epileptogenic zone (see examples in Figs 1A, 2A and 3A). Therefore, the number of electrodes and their location were defined for each individual case, including the need to place contralateral electrodes. Pre-planning of the implantation was performed on 3D T1 MRI images using dedicated software (‘Brainvisa’, http://brainvisa.info) for surface rendering calculation, cortical anatomy analysis and sulci labelling (Mangin et al., 2004). The final determination of each electrode trajectory

S. Aubert et al. was elaborated on a workstation allowing for stereotactic registration of a preoperative stereotactic MR and a preoperative stereotactic telemetric angiography. The implantation accuracy was controlled per-operatively by telemetric X-ray imaging. A post-operative computerized tomography (CT) scan was then used to verify the absence of bleeding and the position of electrodes. In each patient, videoEEG recording was prolonged as long as necessary to allow recording of several ‘habitual’ seizures. We also performed intracerebral electrical stimulations, thus helping to determine the epileptogenic zone determination and to localize the electrodes with accuracy. Intracerebral electrodes were then removed and an MRI was performed, permitting visualization of the trajectory of each electrode (3D T1-weighted images and T2-weighted coronal images,

C SEEG

TBI-2 L 10-11 CU 1-2

A

ER[n]

B

EI values

D

Figure 2 Patient no 23. (A) Sketch of the SEEG exploring the right hemisphere. Electrode CU explored the intraparietal sulcus (external contacts) and the occipital cuneiform gyrus (internal contacts); LI recorded the posterior part of the lateral occipital sulcus on its external contacts and the lingual gyrus on its internal contacts. L was placed in the lesion and reached lateral occipital area (external contacts) and the anterior calcarine fissure (internal contacts). The route of electrode L is shown on the MRI passing through the lesion. Electrode GC was situated in the supramarginalis gyrus and in the isthmus of the cingulate gyrus (on its external and internal contacts, respectively). TO investigated the inferior temporal gyrus (external contacts) and the fusiform gyrus, then the anterior calcarine fissure on its internal contacts. Electrodes B and C explored the middle temporal gyrus (external contacts) and the anterior and posterior hippocampus, respectively. TB was placed in the temporo-basal area (external contacts) and in the entorhinal cortex (internal contacts). (B) MRI (flair sequence) showing an FCD-like lesion in the right lateral temporo-occipital region. (C) SEEG recordings from three regions: the right entorhinal cortex (TB1-2), the lesion (L10–11), and the right internal occipital region (CU1–2). Below are represented the increase in energy ratio ER[n] in the three SEEG signals. In that case, the first detection time found by the Page–Hinkley algorithm was the one recorded in the lesion (L10–11). (D) EI values obtained in seven regions in this patient. Maximal values are observed in L but EI values 40.4 were also recorded in other structures (entorhinal cortex, internal occipital region, supramarginale gyrus and fusiform gyrus), corresponding to a network epileptogenic zone organization.

Quantification of brain epileptogenicity

| 3077

Brain 2009: 132; 3072–3086

L 2-3 L ‘2-3 G 1-2

SEEG

C

A

ER[n]

B

EI values

D

Figure 3 Patient no 4. (A) Sketch of the SEEG exploring the right hemisphere (seven electrodes) and the left hemisphere (four electrodes). Electrode GL recorded the posterior part of the lateral occipital sulcus on its external contacts and the lingual gyrus on its internal contacts. L was placed in the lesion and reached V5 area (external contacts) and the internal occipital region (internal contacts). The route of electrode L0 is shown on the MRI reaching the lesion. L0 was placed symmetrically to L, in the left hemisphere. Electrode G was situated in the supramarginale gyrus and in the isthmus of the cingulate gyrus (on its external and internal contacts, respectively), and G0 in the homologous contralateral region. T investigated the posterior part of the right superior temporal gyrus (external contacts) and the insula on its internal contacts. Electrode C (resp. C0 ) explored the right (resp. left) middle temporal gyrus (external contacts) and the right (resp. left) posterior hippocampus. TB was placed in the temporo-basal area (external contacts) and in the entorhinal cortex (internal contacts). Electrode B recorded the right middle temporal gyrus (external contacts) and the right anterior hippocampus, and B’ was its symmetrical. (B) MRI (flair sequence) showing a dysplastic lesion in the right internal occipital region. (C) SEEG recordings from three regions: the right internal occipital region (lesion, L2–3), the left internal occipital region (L0 2–3), and the right posterior cingulate gyrus (G1–2). Below are represented the increase in energy ratio ER[n] in the three SEEG signals. In that case, the first detection time found by the Page–Hinkley algorithm was recorded in two structures, L and L0 . (D) EI values obtained in seven regions in this patient. Maximal values are observed in L and L0 (bilateral epileptogenic zone organization).

Siemens 1.5 T). Finally, CT scan/MRI data fusion was performed to accurately check the anatomical location of each contact along the electrode trajectory. Signals were recorded on a 128 channel DeltamedTM system. They were sampled at 256 or 512 Hz and recorded on a hard disk (16 bits/ sample) using no digital filter. Two hardware filters were present in the acquisition procedure. The first is a high-pass filter (cut-off frequency equal to 0.16 Hz at 3 dB) used to remove very slow variations that sometimes contaminate the baseline. The second is an anti-aliasing low-pass filter (cut-off frequency equal to 97 Hz at 256 Hz or 170 Hz at 512 Hz).

Signal analysis: definition and computation of the EI The EI was first proposed in a recent study (Bartolomei et al., 2008), where a detailed description is provided. In this section, we provide a short summary of the main features of this index, along with the main methodological aspects. The EI is a normalized quantity (ranging from 0 to 1) computed from depth-EEG signals. This quantity accounts for both (i) the propensity of a given brain structure to generate a rapid discharge; and (ii) the delay of involvement of this structure with

3078

| Brain 2009: 132; 3072–3086

S. Aubert et al.

respect to seizure onset. Indeed, it is generally accepted that these two features (high frequency oscillations and delay of involvement) are relevant indicators of the epileptogenic nature of the recorded underlying neuronal systems. From a methodological viewpoint, the EI can be estimated using a two-stage procedure. First, the signal energy ratio (ER) between high [ (12.4–24 Hz) and

(24–90 Hz)] and low [(3.4–7.4 Hz) and  (7.4–12.4 Hz)] frequency bands of the EEG is computed, over a sliding window, from the signal spectral density (w) (squared modulus of its Fourier transform):

ER½n ¼

E þ E E þ E

where

Z Esub-band ¼

ðwÞdw

sub-band

where ‘sub-band’ denotes the , , and frequency sub-bands classically used to categorize rhythms reflected by EEG signals and where n denotes the current position of the window, in time (Figs 1C, 2C and 3C). Second, change-points are detected in the ER[n] quantity which is sensitive to frequency changes in the signal. This detection is automatic. It makes use of an optimal algorithm (‘cumulative sum algorithm’ or ‘CUSUM’) (Hinkley, 1970, p. 1954) to determine the time instant when ER[n] increases significantly, i.e. when – activity (that is predominant in background SEEG signals) changes into  activity (that is predominant in SEEG signals during rapid discharges) (Figs 1C, 2C and 3C). These two steps allow for the computation of the EIi from the SEEG signal si recorded from brain structure Si: Ndi þH

EIi ¼

X

1 ER½n, Ndi  N0 þ  n¼N

40

di

where N0 is the time instant corresponding to seizure onset (defined hereafter), Ndi is the detection time in signal si recorded from structure Si and H is the time interval over which ER[n] is integrated. From this equation, it can be observed that the sooner structure Si gets involved in the seizure, the higher the index EIi. Parameter  accounts for the particular where Si is the first structure that generates the fast activity (Ndi ¼ N0 , seizure onset) and avoids division by zero. It was arbitrarily set to 1. Another particular case is when structure Si does not generate fast activity. In this case, Ndi is set to be equal to the time corresponding to seizure termination. This operation leads to very low value of EIi, as expected. Parameter H was set to be equal to 5 s. This value corresponds to the average duration of rapid discharges. Finally, in order to obtain a normalized value ranging from 0 (no epileptogenicity) to 1 (maximal epileptogenicity) for considered structures Si, EIi values were divided by the maximal value obtained in each patient. In the sequel, normalized EIi values are simply denoted by ‘EI values’.

Statistical analysis Statistical analysis was performed (i) to assess potential links between the epileptogenic zone organization (as characterized by EI values) and some clinical, neuroradiological or histopathological parameters; and (ii) to determine, among these potential factors, the one(s) that could influence post-operative prognosis. Due to the amount of data to be processed, only one seizure (considered as representative of the patient’s usual seizures) was

analysed for EI determination. The epileptogenic zone organization was considered as being either focal, network or bilateral. Focal organization corresponded to a situation where seizures were generated by only one structure characterized by an EI value 40.4. Network organization was considered present in situations where seizures originated in at least two distant structures disclosing EI value 40.4. Finally, bilateral epileptogenic zone organization corresponded to seizures whose onset rapidly involved homotopic regions (disclosing EI 40.4). The localization of the lesions or/and the epileptogenic zone was defined as frontal, temporal or ‘posterior’ (grouping together parietal and occipital cases). In order to evaluate the relationship between epileptogenicity and distance between studied sites, we estimated the Euclidian distance between the recorded regions and compared with EI values. Statistical analysis was performed in order to compare the data between different groups using a rank-analysis of variance (Kruskal– Walis) for quantitative data. Bivariate analysis used the non-parametric Mann–Whitney test. Chi square (or Fisher’s exact test when appropriate) was used for qualitative data. An analysis with logistical regression was realized in order to determine factors potentially influencing the epileptogenic zone organization. A P-value smaller than or equal to 0.05 was considered to be statistically significant. Statistical tests were realized with the software statistical package for the social sciences (SPSS) for Windows, version 13.1 and Statview 5.0. for Windows.

Results Clinical data Clinical data are summarized in Table 1. Thirty-six patients were included in the study (15 males and 21 females). Mean age at epilepsy onset was 6.6 years (range 6 months–17 years; SD 4.7) and mean epilepsy duration was 19.2 years (range 2–45 years; SD 10). Among the 36 patients, 8 had a normal MRI despite repeated explorations. For the patients whose MRI was abnormal, several features could be found, suggestive of FCD (24 cases), or of DNET/ganglioglioma (four cases). Ten MRI-visible lesions were located in the frontal lobe (three in the precentral cortex), nine in the temporal lobe and nine in posterior cortices (parietal, occipital, or parieto-occipital region). Following SEEG, 33/36 patients went on to have surgical or radio-surgical treatment. Surgery was contraindicated in one patient because of involvement of the language area (left temporo-basal FCD); one patient refused surgery and another is awaiting surgery. Among the 33 patients, radiosurgery was proposed in three cases, targeting the visible MRI lesion, also determined as the epileptogenic zone: two cases of probable dysplastic lesions of the primary motor cortex and one case of NDT located in the visual cortex (calcarine region). Poor outcome was observed in these three cases but low doses were used to avoid functional deficit (three are in class IV). Thirty patients underwent surgical resection, the limits of which were accurately defined and tailored for each patient on the basis of SEEG. Most of them underwent cortectomy (23 cases); six patients had lobectomy (mainly temporal anterior lobectomy), one had lesionectomy with peri-operative ECoG. Mean follow-up was 4.1 years (min = 6 months to max = 7 years; SD 2.24). Two cases with a follow-up of less than 1 year were not included in the

Quantification of brain epileptogenicity analysis. Thus surgical results involve 28 patients. Among 28, 16 patients were in class I (15 IA, 1 IB) (57.2%); 5 (17.8%) in class II and 7 (25%) in class III or IV. Age at epilepsy onset was lower in the non-seizure free (NSF) group [NSF 4.5 versus seizure free (SF) 8.5; P = 0.03, Mann–Whitney] but duration of epilepsy before surgery was not significantly different (P = 0.44, NSF 20.2 versus SF 19.2). In 75% of all cases (21/28 patients with surgical treatment), resection of the lesion was considered as complete, based on peri-operative neurosurgical record and on post-operative MRI. In these cases, 76% (16/21) were in Engel class I whereas, if resection was incomplete, 84% (6/7) were in class III or IV. Thus, the complete resection of the lesion was significantly associated with a better seizure outcome (P = 0.0004).

Histopathological features Histopathological diagnosis was available in 30 patients (for three patients who underwent gamma-knife, histopathological study was not available). For the eight patients with normal MRI, analysis of the resection piece displayed features of FCD and 6/8 were Taylor-type FCD (FCD-II). As shown in Table 1, the most common histopathological diagnosis after surgical resection was FCD in 23/30 patients (77%), and more frequently FCD II, representing 47% of our population (14 patients). FCD I was found in 23% of all cases (seven patients); five cases could be included as NDT, three DNET (10%), one ganglioglioma (3.3%) and one case with histopathologic diagnosis of angiocentric glioma, a newly described lesion (Louis et al., 2007) that corresponds to benign developmental lesions or low-grade tumours, with long-standing epilepsy and low proliferation indices on histopathology, without malignant evolution in most cases (Preusser et al., 2007, Sugita et al., 2008). Other diagnoses included two cases with non-specific changes (gliosis) despite very suggestive MRI appearances of FCD. No positive correlation was found between histopathological findings and post-operative outcome (P = 0.19, Fisher’s exact test). In the group of patients for whom both MRI and histopathological diagnosis was available (n = 15), we found that FCD was over-represented in the frontal lobe (6/15, 40%). Only five patients (33%) displayed an FCD in the temporal lobe and the remaining four (27%) were located in posterior regions. Other lesional types (DNET, ganglioglioma) were preferentially located in the temporal lobe. Finally, after MRI or/and histopathological analysis, we retained the diagnosis of FCD in 30 patients and of NDT in six patients; these patients were included in the signal analysis study. When analysing post-operative outcome in this group, only 26 FCD and 5 NDT could be evaluated (three non-operated patients and two with a follow-up after 51 year); in the FCD group, 14/26 were in class I according to Engel’s score (55%), 4/26 were in class II (15%) and 8/26 were in class III or IV (30%). If we only consider patients with conventional curative surgery (excluding three palliative radiosurgical procedures, all in class IV) the proportion of FCD patients in class I was 61% (14/23 patients). When strictly considering FCD diagnosis assessed by histopathological findings (23 patients) and removing the two patients with 51 year

Brain 2009: 132; 3072–3086

| 3079

follow-up, we found 57% patients (12/21) in class I, 19% (4/21) in class II, and 24% (5/21) in class III or IV.

SEEG signal analysis SEEG recordings (average = 7.5 electrodes per patient) permitted to localize the epileptogenic zone in all selected patients; the frontal lobe was the most frequently involved at seizure onset (17 patients, 48%). The implantation of contralateral electrodes was required in the majority of patients (20/26) including 10/17 patients with the epileptogenic zone located in the frontal lobe, 4/9 patients with the epileptogenic zone in the temporal lobe, and 6/10 patients with the epileptogenic zone located in posterior regions. For 19 patients (52.7% of all cases), seizure started in the right hemisphere and in the left hemisphere for 14 patients (38.8%). Rapid involvement (51 s) of controlateral homotopic regions was observed in 3 patients.

Organization of the epileptogenic zone In order to determine the epileptogencity of analysed brain areas, we selected several cortical areas structures (hereafter named ‘structures’, seven per patient) that were crucial for understanding the anatomo-functional organization of involved systems. In agreement with the characteristics of our population, these structures differed from one patient to another (online supplementary table). Among the recorded structures, some were considered as ‘lesional’ when contacts of the electrodes were localized in (or near) the lesion visible on MRI or disclosed neurophysiological features highly suggestive of FCD (such as focal slow waves or seizure like patterns) in the group with ‘negative-MRI’ (Chassoux et al, 2000). The EI was then calculated for each selected structure. Mean EI value, averaged over all the structures, was 0.379 (0.35). Mean EI in lesional regions was 0.87 (0.25) and 0.29 (0.30) in ‘non-lesional’ structures (P50.001, Mann– Whitney). This result confirmed the intrinsic epileptogenic nature of the studied lesional sites. In addition, and as shown in Fig. 4A, values for FCD tended to be higher than values for NDT (P = 0.07, Mann–Whitney). Values for FCD are grouped around 1 (mean 0.85; median 1; SD 0.31), while more dispersion is observed for NDT (mean 0.60; median 0.56; SD 0.36). In addition, results suggest that intrinsic epileptogenicity is probably higher for FCD than neurodevelopmental tumour. These findings prompted us to define an accurate cut-off of EI values above which we could reasonably postulate that a structure is highly ‘epileptogenic’. Figure 4B shows a box plot representation of EI values in the lesional and extra-lesional sites. Since lesional sites are also epileptogenic in this subset of patients, we established this limit as EI40.4. This value corresponded to the 90th percentile of EI values in the lesion. Therefore, most of the lesional sites disclose values greater than 0.4. This value is close to the cut-off value chosen in our previous study in patients with mesio-temporal lobe epilepsy (Bartolomei et al., 2008). However, as illustrated in Fig. 4B, a large number of structures in extra-lesional sites also have values greater than 0.4.

3080

| Brain 2009: 132; 3072–3086

S. Aubert et al.

Number of patients

A

12 10 8 6 4 2 0 0

1

2

3

4

5

6

7

8

9

10

11

12

B

12

Number of patients

NEI ≥ 0.4

10 8 NSF SF

6 4 2 0 N

Figure 4 (A) Box plot of the EI values in the two main types of observed lesions (FCD: focal cortical dysplasia; NDT: neurodevelopmental tumours). (B) Box plots of EI values from lesional sites and extra lesional sites. Horizontal lines indicate the 10th, 25th, 50th, 75th and 90th percentiles, respectively. Values below the 10th and above the 90th percentiles are indicated by a cross.

This strongly suggests that epileptogenicity extends beyond the limit of the visible lesion in a large number of patients. We then studied, for each patient, the number of highly epileptogenic structures (involved in the epileptogenic zone) as defined by an EI value 50.4 (denoted by NEI50.4). The mean NEI50.4 was 3.14 (1.87) in the whole population. We analysed the type of epileptogenic zone organization for each patient according to our definition of focal, network and bilateral epileptogenic zone organization as described above (see ‘Materials and methods’ section). Figure 5A discloses histograms of the distribution of NEI50.4 in the whole population. We found that 11 patients had only one epileptogenic structure (NEI50.4 = 1), therefore disclosing a strictly focal epileptogenic zone organization (31%). Figure 1 shows one typical example of focal organization. In 25 patients, more than one area belonged to the epileptogenic zone. Among them, 22 disclosed a network epileptogenic zone organization (61%) (see example in Fig. 2) and three a bilateral epileptogenic zone representation (8%) (see example in Fig. 3). Therefore, the majority of patients with FCD or NDT were characterized by a network organization of the epileptogenic zone. We observed a significant inverse relationship between EI values and distances between lesional site and other studied regions (P50.0001). However, the mean distances were not different considering sites disclosing high EI (40.4) versus sites disclosing EI50.4 (EI40.4: mean 40.59  17 mm; EI50.4: 40.32  24 mm, P = 0.54).

F

B

Epileptogenic zone organization

Figure 5 (A) Histogram showing the number of structures disclosing EI values 40.4 in the whole patient population. (B) Number of patients becoming seizure-free (SF) or nonseizure free (NSF) after surgery, according to the epileptogenic zone organization type. F = focal; N = network; B = bilateral.

This result suggests that distance is not the main factor influencing epileptogenicity, this probably being more linked to anatomical/functional connectivity. Some differences appeared when focusing on the lobe involved in seizure onset and epileptogenic zone organization (P = 0.0019, Chi square). We have separated the anatomical site of seizure onset into three groups: frontal, temporal and posterior (including parietal, occipital and parietooccipital seizures). There were some differences in terms of epileptogenic zone organization among frontal, temporal and ‘posterior’ patients (P = 0.014). All patients who exhibited a bilateral epileptogenic zone organization were ‘posterior’ (three cases). When analysing ‘posterior’ versus ‘non-posterior’, according to the bilaterality (or not) of their epileptogenic zone organization, the difference increased (P = 0.0055). Thus, ‘posterior’ showed a bilateral epileptogenic zone organization more often. For patients whose epileptogenic zone was located in the frontal lobe, their organization had a tendency to be more regional than focal (P = 0.045). There was no statistical difference in terms of epileptogenic zone organization for patients with temporal lobe seizure onset, when compared to non-temporal patients (P = 0.40).

Relationship between clinical data, histopathological parameters and epileptogenic zone organization We looked at whether the epileptogenic zone organization was correlated with clinical parameters. We first looked for a

Quantification of brain epileptogenicity correlation between epilepsy duration or age at epilepsy onset and the number of structures displaying a high EI (NEI50.4) using Spearman’s correlation test. No significant relationship was found for epilepsy duration (P = 0.21) or for age at onset (P = 0.93). In the same way, we did not find significant differences in these two parameters between the three epileptogenic zone organization types (P = 0.28, age at onset; F = 2.4, P = 0.09 epilepsy duration, Kruskal Wallis). We were unable to find any statistical correlation between histopathological features and the organization of the epileptogenic zone (FCD versus other diagnoses, P = 0.39, Chi square; FCD I versus FCD II, P = 0.72, Chi square). Since, in our first study using EI quantification in mesio-temporal lobe epilepsy (Bartolomei et al, 2008), we found a relationship between epilepsy duration and the number of epileptogenic structures, we looked at whether differences could be found by separately analysing FCD affecting mesial temporal structures and FCD in other locations. Figure 6 shows the relationship between epilepsy duration and NEI50.4 in the temporal lobe, frontal lobe and posterior locations (defined as epileptogenic area including the lesional site located in the occipital or/and parietal areas). No relationship was found in frontal and posterior location (frontal lobe P = 0.96; posterior regions P = 0.97, Spearman). In contrast, for mesial temporal localization, a significant relationship was found between the epilepsy duration and the number of structures displaying a high EI (P = 0.02). A test of logistical regression with multivariate analysis was performed to see if clinical features (sex, age at epilepsy onset, epilepsy duration, lobar location of epileptogenic zone, presence or not of a visible lesion on MRI) influenced the type of epileptogenic zone organization. We did not find any significant relationship for

Brain 2009: 132; 3072–3086

| 3081

the analysed parameters [age at epilepsy onset: P = 0.94; epilepsy duration: P = 0.29; lobar location of epileptogenic zone (frontal: P = 0.50; temporal: P = 0.89; posterior: P = 0.72); presence or not of a visible lesion on MRI: P = 0.76]. In addition, when comparing FCD and NDT, the number of structures displaying a high EI (NEI50.4) was not significantly different even though we observed a tendency for NDT to be more focal (mean NEI50.4 = 2.1  1.8) than FCD (mean NEI50.4 = 3.8  1.7) (P = 0.08, Mann–Whitney).

Relationship between epileptogenic zone organization and surgical outcome We then looked at whether the epileptogenic zone organization was a potential factor influencing surgical outcome. The number of epileptogenic structures (NEI50.4) was not significantly different with regard to surgical outcome (SF = 3.2, NSF = 3.4; P = 0.80). We observed a trend for a difference in seizure outcome according to the type of organization (P = 0.06, Chi square, Fig. 5B). Indeed, 57% of patients with network organization and 87% with focal organization were seizure-free, while none of those with bilateral organization became seizure-free.

Discussion This article is based on the study of 36 patients suffering from focal pharmacoresistant epilepsy related to FCD or NDTs, investigated with SEEG during their pre-surgical evaluation. The main objective was to quantify the epileptogenicity of the lesional

Figure 6 Relationship between epilepsy duration and NEI50.4 in the temporal lobe (A), frontal lobe (B) and posterior locations (C).

3082

| Brain 2009: 132; 3072–3086

sites and distant investigated structures and to correlate these with clinical data.

General features of FCDs and NDT in the studied population Until now, FCD has been extensively studied, but intracranial data are often based on subdural approaches (grids, strips) or per-operative ECoG. Relatively few reports studying FCD with intracerebral electrodes have previously been published (Chassoux et al., 2000; Tassi et al., 2001; Francione et al., 2003). Intracerebral electrodes, as used in the SEEG method, are particularly suited for studying the epileptogenicity of FCD (Chassoux et al., 2000; Tassi et al., 2001). Indeed, they can catch parasulcal and mesial areas and also explore different anatomically separate sites along the same trajectory. Histopathological analysis, when available, revealed that most of the patients in the present study (23/30 patients, 77%) had FCD. In agreement with the literature (Palmini et al., 1991, 1994; Hirabayashi et al., 1993) we found that frontal lobe localization was the most frequent in patients with FCD. It is noteworthy that in our study, all patients with normal MRI (8/8) had FCD. This result is in agreement with other authors who have demonstrated that the diagnosis of FCD is relatively common amongst all patients operated with normal MRI (Palmini et al., 1995; Tassi et al., 2001, 2002; Colombo et al., 2003; Alarcon et al., 2006; McGonigal et al., 2007). Another interesting fact is that six cases in the present series had FCD-II (‘Taylor-type FCD’), a result that differs from previous studies where FCD was of type I when MRI was normal (Kresk et al., 2008a; Chassoux, 2008). In the whole FCD group, 55% were in class I according to Engel’s score, but if we focus on the post-operative outcome in the patients who underwent curative surgery, the results were slightly better (61% in class I). This post-operative outcome is close to the proportion of FCD patients in class I observed in other studies, ranging from 64% to 79% (Chassoux et al., 2000; Tassi et al., 2001; Urbach et al., 2002; Cohen-Gadol et al., 2004; Kral et al., 2007). Less favourable outcome after FCD surgery (about 40% class I) has also been reported in older studies (Sisodiya, 2000). In our population, no relationship was found between histopathological features and post-operative outcome, but the fact that the FCD-II group was over-represented in our population is an important bias. However, post-operative outcome in our FCD-I group was good as 4/6 patients (67%) were in class Ia after a mean follow-up of 4 years. Moreover, all FCD-I patients with complete resection of the lesion were in class Ia. The populations included in these different studies were, however, notably heterogeneous and this may explain the discrepancies between published results [different population sizes, patient selection (children/adults), follow-up duration, or histopathological classification]. Our study confirms that, in FCD, complete resection of the lesion is crucial for a good post-operative outcome, and increases the percentage of patients in class I from 57% to 76%. The importance of this factor has already been demonstrated in FCD

S. Aubert et al. (Chassoux et al., 2000; Edwards et al., 2000; Hong et al., 2000; Cohen-Gadol et al., 2004; Krsek et al., 2009) and other lesional types, like NDTs (Aronica et al., 2001, 2007; Lee et al., 2001; Nolan et al., 2004). Another important positive prognosis factor is an older age at epilepsy onset (after 8.5 years old), a finding previously described (Palmini et al., 1991; Hirabayashi et al., 1993; Polkey, 1996; Wyllie, 1998).

Quantification of the epileptogenic zone and intrinsic epileptogenicity We propose a quantification of brain structure epileptogenicity associated with FCD or NDT. The quantity is based on signal analysis of rapid discharges (high frequency oscillations) observed at seizure onset (Allen et al., 1992; Alarcon et al., 1995; Wendling et al., 2003; Worrell et al., 2004; Jirsch et al., 2006). To this aim, we have used a quantity that combines the signal energy in defined frequency bands (appearance of beta–gamma oscillations concomitantly with the attenuation of slower alpha–theta oscillations) and the delay with which rapid discharges occur. It is noteworthy that several experimental (Traub et al., 2001) and computational modelling (Wendling et al., 2005) studies have also demonstrated the existence of a relationship between the epileptogenicity of the neuronal tissue and its propensity to generate fast oscillations. EI values were found to be higher in the lesional sites for a great majority of patients. This result adds further confirmation that FCD and NDT are intrinsically epileptogenic as suggested by earlier reports (Ferrer et al., 1992; Mattia et al., 1995, Palmini et al., 1995; Morioka et al., 1999; Hong et al., 2000; Otsubo et al., 2005). In addition, this result indicates that the EI method is well-suited to quantify the epileptogenicity of a brain structure. Furthermore, our statistical analysis showed that intrinsic epileptogenicity in FCD tends to be higher than in NDT, a result in agreement with previous studies (Palmini et al., 1995; Rosenow et al., 1998). Even if the lesional sites are generally the ones disclosing highest epileptogenicity, our study highlights the fact that, in these patients, distant sites may also show early involvement by rapid discharges. As mentioned above, these properties allow these regions to be defined as highly epileptogenic and, therefore, part of the epileptogenic zone. According to this assumption, we defined the three different types of organization of the epileptogenic zone: ‘focal’, ‘network’ and ‘bilateral’. The ‘focal’ category was defined by a very focal onset within a single epileptogenic structure. This could be seen as corresponding to the classical notion of an epileptic ‘focus’ and accounted for 31% of our patients. In 8% of cases we observed a bilateral organization, these three cases corresponding to posterior epilepsies. The majority of patients (61%) had a more complex organization (‘epileptogenic network’), with at least two distant sites disclosing high epileptogenicity. This number is particularly high with regard to previous studies that generally consider that in FCD, the lesional and epileptogenic zones tend to co-localize, albeit without quantification of this aspect (Chassoux et al., 2000; Tassi et al., 2001, 2002; Kresk et al., 2008a, b).

Quantification of brain epileptogenicity In addition, it is now well-established that the epileptogenic zone can be more extended than the visible lesion on MRI (Ferrier et al., 2001; Tassi et al., 2002, Colombo et al., 2003, Francione et al., 2003), and/or histopathological findings (Tassi et al., 2001). In the paper by Chassoux et al. (2000), using the same SEEG approach, 20% of patients with FCD were considered to have a larger epileptogenic zone than the dysplastic cortex. In a more recent paper (Widdess-Walsh et al., 2007) the role of subdural grid evaluation for FCD was studied. The authors found that a diffuse, non-localized onset was found in 30% of cases and that a complex pattern (with two or three sites of onset) may be observed in 41% of patients. Nevertheless, despite the fact that it is has been suggested that the epileptogenic zone is often larger than the visible lesion, a focal organization is still considered to be the rule in FCD. Our study is the first to quantify precisely the epileptogenicity of these lesions and thus to demonstrate that the epileptogenicity may affect remote regions, distant from the lesional site. In addition, the extent of the epileptogenicity appears to be a prognostic factor for post-operative outcome since a focal organization was associated with a better Engel’s score. On the opposite, bilateral epileptogenicity was always associated with unfavourable outcome. Interestingly, we found a propensity for patients with frontal lesions to develop a network epileptogenic zone organization rather than a focal one. These findings are in agreement with human in vivo studies showing large functional connections within the frontal lobes and between the two frontal lobes (Lacruz et al., 2006). Besides, in our study, posterior cases often show bilateral organization. It is known that posterior epilepsies tend to present diffuse or even bilateral interictal spikes on surface EEG (Gavaret et al., 2008). Only a few studies (Krsek et al., 2008b) have reported that, in FCD patients, ictal EEG can show early bilateral discharge, but to our knowledge no case with bilateral posterior ictal onset in a patient with FCD has yet been reported. Rapid multilobar involvement has been more frequently described (Chassoux et al., 2000). The extension of the epileptogenic zone to involve contralateral homotopic structures by a secondary epileptogenic process appears to be the most probable mechanism. The high functional and anatomical connectivity between posterior regions (Clarke and Miklossy, 1990; Riedel et al., 2004; Peltier et al., 2007; Teipel et al., 2008) may support this phenomenon. We cannot, however, rule out the possibility that posterior cases with bilateral seizure onset could be due to bilateral FCD, but, in these particular cases, as the patients are never operated bilaterally, we do not have any histopathological data available from the contralateral lobe. Indeed, bilateral FCD can occur (Hong et al., 2000), especially in posterior regions (Chassoux et al., 2000; Fukui et al., 2001; Francione et al., 2003; Fogarasi et al., 2004). In our FCD population, we failed to identify clear clinical or histological factors influencing the type of organization of the epileptogenic zone. In contrast to previous observations in mesio-temporal lobe epilepsy (Bartolomei et al., 2008), we did not find a correlation between epilepsy duration and the extension of the epileptogenic zone. However, this result may be due to differences in the localization of the epileptogenic zone. Indeed,

Brain 2009: 132; 3072–3086

| 3083

a significant trend for a relation between epilepsy duration and extension of the epileptogenicity (NEI50.4) was observed in the cases affecting mesial temporal lobe. This result is in agreement with previous studies showing a positive correlation between epilepsy duration and development of pharmacoresistance, in lesions such as DNET and gangliogliomas, mainly affecting the temporal lobe (Weissman et al., 1996; Itoh et al., 1997; Lanzieri et al., 1997; Chassoux et al., 2000; Aronica et al., 2001; Hennessy et al., 2001; Valenti et al., 2002; Luyken et al., 2003; Nolan et al., 2004). The underlying mechanisms are probably related to the synaptic alterations observed during secondary epileptogenesis (Khalilov et al., 2003). The features observed in the ‘network’ organization of the epileptogenic zone are indeed reminiscent of those seen in secondary epileptogenesis processes, which occur not only at the site of a lesion but also in distant areas in animal models (Morrell, 1959; Wilder et al., 1969) and in humans (Morrell, 1985, 1989). It is interesting that secondary epileptogenesis is easier to induce in limbic temporal lobe structures than in other sites in the brain (Niedieck et al., 1990, Hirsch et al., 1991a,b; Parent et al., 1997; Bertram, 2007) and can lead to a mirror focus (Wilder et al., 1969; Morrell, 1985). These processes seem to mainly affect younger subjects, in both animal (Fanelli and McNamara, 1986) and human studies (Morrell, 1985), thus suggesting that the earlier the onset of epilepsy, the higher the risk of secondary epileptogenesis.

Concluding remarks This study has concluded that the EI can be a useful tool for quantifying and defining the epileptogenicity in FCD and neurodevelopmental tumour. We found that many cerebral structures are involved at seizure onset (on average, more than three per patient). A majority of patients (61%) had a network epileptogenic zone organization, showing that most of these ‘focal’ lesions display an epileptogenic zone that involves other structures, often distant ones; bilateral organization can also be seen, especially in posterior FCD. The precise determination of the epileptogenic zone is an important prognostic factor and with the help of the EI, we could confirm that epileptogenic zone organization influences surgical outcome. Furthermore, the number of epileptogenic structures increased with epilepsy duration in the mesio-temporal epileptic cases, which raises the issue of secondary epileptogenesis in this population of pharmaco-resistant epileptic patients.

Supplementary material Supplementary material is available at Brain online.

References Alarcon G, Binnie CD, Elwes RDC, Polkey CE. Power spectrum and intracranial EEG patterns at seizure onset in partial epilepsy. Electroencephalogr Clin Neurophysiol 1995; 94 (5): 326–37.

3084

| Brain 2009: 132; 3072–3086

Alarcon G, Valentı´n A, Watt C, Selway R, Lacruz ME, Elwes R, et al. Is it worth pursuing surgery for epilepsy in patients with normal neuroimaging? J Neurol Neurosurg Psychiatry 2006; 77: 474–80. Allen PJ, Fish DR, Smith SJM. Very high frequency rhythmic activity during SEEG suppression in frontal lobe epilepsy. Clin Neurophysiol 1992; 82: 155–9. Aronica E, Leenstra S, van Veelen CW, van Rijen PC, Hulsebos TJ, Tersmette AC, et al. Glioneuronal tumors and medically intractable epilepsy: a clinical study with long-term follow-up of seizure outcome after surgery. Epilepsy Res 2001; 43: 179–91. Aronica E, Redeker S, Boer K, Spliet WG, van Rijen PC, Gorter JA, et al. Inhibitory networks in epilepsy-associated gangliogliomas and in the perilesional epileptic cortex. Epilepsy Res 2007; 74: 33–44. Bancaud J, Angelergues R, Bernouilli C, Bonis A, Bordas-Ferrer M, Bresson M, et al. Functional stereotaxic exploration (SEEG) of epilepsy. Electroencephalogr Clin Neurophysiol 1970; 28: 85–6. Bancaud J, Talairach J, Bonis A, Schaub C, Szilka G, Morel P. La ste´re´oence´phalographie dans l’e´pilepsie: informations neurophysiopathologiques apporte´es par l’investigation fonctionnelle ste´re´otaxique. Paris, Masson 1965. Barkovich AJ, Chuang SH, Norman D. MR of neuronal migration anomalies. AJR Am J Roentgenol 1988; 150: 179–87. Bartolomei F, Barbeau E, Gavaret M, Guye M, McGonigal A, Regis J, et al. Cortical stimulation study of the role of rhinal cortex in deja vu and reminiscence of memories. Neurology 2004; 63: 858–64. Bartolomei F, Chauvel P, Wendling F. Epileptogenicity of brain structures in human temporal lobe epilepsy: a quantified study from intracerebral EEG. Brain 2008; 131: 1818–30. Bartolomei F, Guye M, Gavaret M, Regis J, Wendling F, Raybaud C, et al. [The presurgical evaluation of epilepsies]. Rev Neurol (Paris) 2002; 158: 4S55–64. Bertram E. The relevance of kindling for human epilepsy. Epilepsia 2007; 48 (Suppl 2): 65–74. Chassoux F, Devaux B, Landre E, Turak B, Nataf F, Varlet P, et al. Stereoelectroencephalography in focal cortical dysplasia: a 3D approach to delineating the dysplastic cortex. Brain 2000; 123 (Pt 8): 1733–51. Chassoux F. Malformations of cortical development: which strategy is best? Neurochirurgie 2008; 54: 272–81. Clarke S, Miklossy J. Occipital cortex in man: organization of callosal connections, related myelo- and cyto-architecture, and putative boundaries of functional visual areas. J Comp Neurol 1990; 298: 188–214. Cohen-Gadol AA, Ozduman K, Bronen RA, Kim JH, Spencer DD. Longterm outcome after epilepsy surgery for focal cortical dysplasia. J Neurosurg 2004; 101: 55–65. Colombo N, Tassi L, Galli C, Citterio A, Lo Russo G, Scialfa G, et al. Focal cortical dysplasias: MR imaging, histopathologic, and clinical correlations in surgically treated patients with epilepsy. AJNR Am J Neuroradiol 2003; 24: 724–33. Daumas-Duport C. Dysembryoplastic neuroepithelial tumours. Brain Pathol 1993; 3: 283–95. Daumas-Duport C, Varlet P, Bacha S, Beuvon F, Cervera-Pierot P, Chodkiewicz JP. Dysembryoplastic neuroepithelial tumors: nonspecific histological forms – a study of 40 patients. J Neurooncol 1999; 41: 267–80. Degen R, Ebner A, Lahl R, Leonhardt S, Pannek HW, Tuxhorn I. Various findings in surgically treated epilepsy patients with dysembryoplastic neuroepithelial tumors in comparison with those of patients with other low-grade brain tumors and other neuronal migration disorders. Epilepsia 2002; 43: 1379–84. Duncan JS. Imaging and epilepsy. Brain 1997; 120 (Pt 2): 339–77. Edwards JC, Wyllie E, Ruggeri PM, Bingaman W, Luders H, Kotagal P, et al. Seizure outcome after surgery for epilepsy due to malformation of cortical development. Neurology 2000; 55: 1110–4. Engel J Jr. Classifications of the International League Against Epilepsy: time for reappraisal. Epilepsia 1998; 39: 1014–7.

S. Aubert et al. Fanelli RJ, McNamara JO. Effects of age on kindling and kindled seizureinduced increase of benzodiazepine receptor binding. Brain Res 1986; 362: 17–22. Ferrer I, Pineda M, Tallada M, Oliver B, Russi A, Oller L, et al. Abnormal local-circuit in epilepsie partialis continua associated with focal cortical dysplasia. Acta Neuropathol 1992; 83: 647–52. Ferrier CH, Alarcon G, Engelsman J, Binnie CD, Koutroumanidis M, Polkey CE, et al. Relevance of residual histologic and electrocorticographic abnormalities for surgical outcome in frontal lobe epilepsy. Epilepsia 2001; 42: 363–71. Fogarasi A, Neuwirth M, Hegyi M, Siegler Z, Barsi P, Gyorsok Z, et al. Bilateral epilepsy surgery in a 4-year old child. Neuropediatrics 2004; 35: 360–3. Francione S, Vigliano P, Tassi L, Cardinale F, Mai R, Lo Russo G, et al. Surgery for drug-resistant epilepsy in children with focal cortical dysplasia: anatomical-clinical correlations and neurophysiological data in 10 patients. J Neurol Neurosurg Psychiatry 2003; 74: 1493–501. French JA. Refractory epilepsy: clinical overview. Epilepsia 2007; 48 (Suppl 1): 3–7. Fukui K, Morioka T, Nishio S, Kawamura T, Wu CM, Uesaka T, et al. Multiple focal cortical dysplasias presenting with intractable epilepsy: case report. No To shinkei 2001; 53: 961–7. Gambardella A, Palmini A, Andermann F, Dubeau F, Da Costa JC, Quesney LF, et al. Usefulness of focal rhythmic discharges on scalp EEG of patients with focal cortical dysplasia and intractable epilepsy. Electroencephalogr Clin Neurophysiol 1996; 98: 243–9. Gavaret M, Trebuchon A, Bartolomei F, Marquis P, McGonigal A, Wendling F, et al. Source localization of scalp-EEG interictal spikes in posterior cortex epilepsies investigated by HR-EEG and SEEG. Epilepsia 2008; 50: 276–89. Hennessy MJ, Elwes RD, Honavar M, Rabe-Hesketh S, Binnie CD, Polkey CE. Predictors of outcome and pathological considerations in the surgical treatment of intractable epilepsy associated with temporal lobe lesions. J Neurol Neurosurg Psychiatry 2001; 70: 450–8. Hinkley D. Inference about the change point from cumulative sum-tests. Biometrika 1970; 58: 509–23. Hirabayashi S, Binnie CD, Janota I, Polkey CE. Surgical treatment of epilepsy due to cortical dysplasia: clinical and EEG findings. J Neurol Neurosurg Psychiatry 1993; 56: 765–70. Hirsch LJ, Spencer SS, Spencer DD, Williamson PD, Mattson RH. Temporal lobectomy in patients with bitemporal epilepsy defined by depth electroencephalography. Ann Neurol 1991a; 30: 347–56. Hirsch LJ, Spencer SS, Williamson PD, Spencer DD, Mattson RH. Comparison of bitemporal and unitemporal epilepsy defined by depth electroencephalography. Ann Neurol 1991b; 30: 340–6. Honavar M, Janota I, Polkey CE. Histological heterogeneity of dysembryoplastic neuroepithelial tumour: identification and differential diagnosis in a series of 74 cases. Histopathology 1999; 34: 342–56. Hong SC, Kang KS, Seo DW, Hong SB, Lee M, Nam DH, et al. Surgical treatment of intractable epilepsy accompanying cortical dysplasia. J Neurosurg 2000; 93: 766–73. Itoh M, Morita T, Houdou S, Kato S, Ohama E, Mizushima M, et al. Two cases of dysembryoplastic neuroepithelial tumors with intractable complex partial seizures. J Child Neurol 1997; 12: 67–70. Jirsch JD, Urrestarazu E, LeVan P, Olivier A, Dubeau F, Gotman J. High frequency oscillations in human focal seizures. Brain 2006; 129: 1593–608. Khalilov I, Holmes GL, Ben-Ari Y. in vitro formation of a secondary epileptogenic mirror focus by interhippocampal propagation of seizures. Nature Neuroscience 2003; 6: 1079–85. Kirkpatrick PJ, Honavar M, Janota I, Polkey CE. Control of temporal lobe epilepsy following en bloc resection of low-grade tumors. J Neurosurg 1993; 78: 19–25. Kral T, Clusmann H, Blumcke I, Fimmers R, Ostertun B, Kurthen M, et al. Outcome of epilepsy surgery in focal cortical dysplasia. J Neurol Neurosurg Psychiatry 2003; 74: 183–8.

Quantification of brain epileptogenicity Kral T, von Lehe M, Podlogar M, Clusmann H, Sussmann P, Kurthen M, et al. J Neurol Neurosurg 2007; 78: 853–6. Krsek P, Maton B, Korman B, Pacheco-Jacome E, Jayakar P, Dunoyer C, et al. Different features of histopathological subtypes of pediatric focal cortical dysplasia. Ann Neurol 2008a; 63: 758–69. Krsek P, Pieper T, Karlmeier A, Hildebrandt M, Kolodziejczyk D, Winkler P, et al. Different presurgical characteristics and seizure outcomes in children with focal cortical dysplasia type I or II. Epilepsia 2008b; 30: 1–13. Krsek P, Maton B, Jayakar P, Dean P, Korman B, Rey G, et al. Incomplete resection of focal cortical dysplasia is the main predictor of poor outcome. Neurology 2009; 72: 217–23. Kuzniecky R, Murro A, King D, Morawetz R, Smith J, Powers R, et al. Magnetic resonance imaging in childhood intractable partial epilepsies: pathologic correlations. Neurology 1993; 43: 681–7. Lacruz ME, Garcı´a Seoane JJ, Valentin A, Selway R, Alarcon G. Frontal and temporal functional connections of the living human brain. European Journal of Neuroscience 2006; 26: 1357–70. Lanzieri CF, Bangert BA, Tarr RW, Shah RS, Lewin JS, Gilkeson RC. Neuroradiology case of the day. Dysembryoplastic neuroepithelial tumor. AJR Am J Roentgenol 1997; 169: 293, 297–8. Lee SK, Choe G, Hong KS, Nam HW, Kim JY, Chung CK, et al. Neuroimaging findings of cortical dyslamination with cytomegaly. Epilepsia 2001; 42: 850–6. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK (eds). WHO classification of tumours of the central nervous system. Lyon: IARC, 2007. Luyken C, Blumcke I, Fimmers R, Urbach H, Elger CE, Wiestler OD, et al. The spectrum of long-term epilepsy-associated tumors: long-term seizure and tumor outcome and neurosurgical aspects. Epilepsia 2003; 44: 822–30. Mangin JF, Riviere D, Cachia A, Duchesnay E, Cointepas Y, Papadopoulos-Orfanos D, et al. A framework to study the cortical folding patterns. Neuroimage 2004; 23 (Suppl 1): S129–38. Mattia D, Olivier A, Avoli M. Seizure-like discharges recorded in human dysplastic neocortex maintained in vitro. Neurology 1995; 45: 1391–5. McGonigal A, Bartolomei F, Regis J, Guye M, Gavaret M, Trebuchon-Da Fonseca A, et al. Stereoelectroencephalography in presurgical assessment of MRI-negative epilepsy. Brain 2007; 130: 3169–83. Morrell F. Experimental epilepsy in animals. Arch Neurol 1959; 1: 141–6. Morrell F. Secondary epileptogenesis in man. Arch Neurol 1985; 42: 318–35. Morrell F. Varieties of human epileptogenesis. J Clin Neurophysiol 1989; 42: 318–35. Morioka T, Nishio S, Ishibashi H, Muraishi M, Hisada K, Shigeto H, et al. Intrinsic epileptogenicity of focal cortical dysplasia as revealed by magnetoencephalography and electrocorticography. Epilepsy Res 1999; 33: 177–87. Niedieck T, Franke HG, Degen R, Ettlinger G. The development of independant foci in epileptic patients. Arch Neurol 1990; 47: 406–11. Nolan MA, Sakuta R, Chuang N, Otsubo H, Rutka JT, Snead OC III, et al. Dysembryoplastic neuroepithelial tumors in childhood: long-term outcome and prognostic features. Neurology 2004; 62: 2270–6. Ostertun B, Wolf HK, Campos MG, Matus C, Solymosi L, Elger CE, et al. Dysembryoplastic neuroepithelial tumors: MR and CT evaluation. AJNR Am J Neuroradiol 1996; 17: 419–30. Otsubo H, Iida M, Okuda C, Ochi A, Pang E, Weiss SK, et al. Neurophysiologic findings of neuronal migration disorders: intrinsic epileptogenicity of focal cortical dysplasia on electroencephalography, electrocorticography, and magnetoencephalography. J Child Neurol 2005; 20: 357–63. Page E. Continuous inspection schemes. Biometrika 1954; 41: 100–15. Parent JM, Yu TW, Leibowitz RT, Geschwind DH, Sloviter RS, Lowenstein DH. Dentate granule cell neurogenesis is increased by seizures and contributes to aberrant network reorganization in the adult rat hippocampus. J Neurosci 1997; 17: 3727–38. Palmini A, Andermann F, Olivier A, Tampieri D, Robitaille Y, Andermann E, et al. Focal neuronal migration disorders and intractable partial epilepsy: a study of 30 patients. Ann Neurol 1991; 30: 741–9.

Brain 2009: 132; 3072–3086

| 3085

Palmini A, Gambardella A, Andermann F, Dubeau F, da Costa JC, Olivier A, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995; 37: 476–87. Palmini A, Gambardella A, Andermann F, Dubeau F, da Costa JC, Olivier A, et al. Operative strategies for patients with cortical dysplastic lesions and intractable epilepsy. Epilepsia 1994; 35 (Suppl 6): S57–71. Pasquier B, Peoc HM, Fabre-Bocquentin B, Bensaadi L, Pasquier D, Hoffmann D, et al. Surgical pathology of drug-resistant partial epilepsy. A 10-year-experience with a series of 327 consecutive resections. Epileptic Disord 2002; 4: 99–119. Peltier S, Stilla R, Mariola E, LaConte S, Hu X, Sathian K. Activity and effective connectivity of parietal and occipital cortical regions during haptic shape perception. Neuropsychologia 2007; 45: 476–83. Polkey CE. Cortical dysplasia: resective surgery in children. In: Guerrini R, Andermann F, Canapicchi R, Roger J, Zifkin BG, Pfanner P, editors. Dysplasias of cerebral cortex and epilepsy. Philadelphia: LippincottRaven; 1996. p. 435–9. Prayson RA. Clinicopathological findings in patients who have undergone epilepsy surgery in the first year of life. Pathol Int 2000; 50: 620–5. Prayson RA, Estes ML, Morris HH. Coexistence of neoplasia and cortical dysplasia in patients presenting with seizures. Epilepsia 1993; 34: 609–15. Preusser M, Hoischen A, Novak K, Czech T, Prayer D, Hainfellner JA, et al. Angiocentric glioma: report of clinico-pathologic and genetic findings in 8 cases. Am J Surg Pathol 2007; 31: 1709–18. Raymond AA, Fish DR, Sisodiya SM, Alsanjari N, Stevens JM, Shorvon SD. Abnormalities of gyration, heterotopias, tuberous sclerosis, focal cortical dysplasia, microdysgenesis, dysembryoplastic neuroepithelial tumour and dysgenesis of the archicortex in epilepsy. Clinical, EEG and neuroimaging features in 100 adult patients. Brain 1995; 118 (Pt 3): 629–60. Riedel E, Stephan T, Marx E, Deutschlander A, Bruning R, Brandt T. Areas MT/V5 and their transcallosal connectivity in cortical dysplasia by fMRI. Neuroreport 2004; 15: 1877–81. Rosenow F, Luders H, Dinner DS, Prayson RA, Mascha E, Wolgamuth BR, et al. Histopathological correlates of epileptogenicity as expressed by electrocorticographic spiking and seizure frequency. Epilepsia 1998; 39: 850–6. Shin JH, Lee HK, Khang SK, Kim DW, Jeong AK, Ahn KJ, et al. Neuronal tumors of the central nervous system: radiologic findings and pathologic correlation. Radiographics 2002; 22: 1177–89. Sisodiya SM. Surgery for malformations of cortical development causing epilepsy. Brain 2000; 123 (Pt 6): 1075–91. Stanescu Cosson R, Varlet P, Beuvon F, Daumas Duport C, Devaux B, Chassoux F, et al. Dysembryoplastic neuroepithelial tumors: CT, MR findings and imaging follow-up: a study of 53 cases. J Neuroradiol 2001; 28: 230–40. Sugita Y, Ono T, Ohshima K, Niino D, Ito M, Toda K, et al. Brain surface spindle cell glioma in a patient with medically intractable partial epilepsy: a variant of monomorphous angiocentric glioma? Neuropathology 2008; 28: 516–20. Talairach J, Bancaud J. Lesion, ‘‘irritative’’ zone and epileptogenic focus. Confin Neurol 1966; 27: 91–4. Talairach J, Tournoux P, Musolino A, Missir O. Stereotaxic exploration in frontal epilepsy. Adv Neurol 1992; 57: 651–88. Tassi L, Pasquier B, Minotti L, Garbelli R, Kahane P, Benabid AL, et al. Cortical dysplasia: electroclinical, imaging, and neuropathologic study of 13 patients. Epilepsia 2001; 42: 1112–23. Tassi L, Colombo N, Garbelli R, Francione S, Lo Russo G, Mai R, et al. Focal cortical dysplasia: neuropathological subtypes, EEG, neuroimaging and surgical outcome. Brain 2002; 125: 1719–32. Teipel SJ, Pogarell O, Meindl T, Dietrich O, Sydykova D, Hunklinger U, et al. Regional networks underlying interhemispheric connectivity: an EEG and DTI study in healthy ageing and amnestic mild cognitive impairment. Human Brain Mapping 2008; 30: 2098–119.

3086

| Brain 2009: 132; 3072–3086

Traub RD, Whittington MA, Buhl EH, LeBeau FEN, Bibbig A, Boyd S, et al. A possible role for gap junctions in generation of very fast EEG oscillations preceding the onset of, and perhaps initiatig, seizures. Epilepsia 2001; 42: 153–70. Urbach H, Scheffler B, Heinrichsmeier T, von Oertzen J, Kral T, Wellmer J, et al. Focal cortical dysplasia of Taylor’s balloon cell type: a clinicopathological entity with characteristic neuroimaging and histopathological features, and favorable postsurgical outcome. Epilepsia 2002; 43: 33–40. Valenti MP, Froelich S, Armspach JP, Chenard MP, Dietemann JL, Kerhli P, et al. Contribution of SISCOM imaging in the presurgical evaluation of temporal lobe epilepsy related to dysembryoplastic neuroepithelial tumors. Epilepsia 2002; 43: 270–6. Wiebe S, Blume WT, Girvin JP, Eliasziw M. Effectiveness and efficiency of surgery for Temporal Lobe Epilepsy Study Group. N Engl J Med 2001; 345: 311–8. Weissman Z, Michowitz S, Shuper A, Kornreich L, Amir J. Dysembryoplastic neuroepithelial tumor: a curable cause of seizures. Pediatr Hematol Oncol 1996; 13: 463–8. Wendling F, Bartolomei F, Bellanger JJ, Bourien J, Chauvel P. Epileptic fast intracerebral EEG activity: evidence for spatial decorrelation at seizure onset. Brain 2003; 126: 1449–59.

S. Aubert et al. Wendling F. Neurocomputational models in the study of epileptic phenomena. J Clin Neurophysiol 2005; 22: 285–7. Widdess-Walsh P, Kellinghaus C, Jeha L, Kotagal P, Prayson R, Bingaman W, et al. Electro-clinical and imaging characteristics of focal cortical dysplasia: correlation with pathological subtypes. Epilepsy Res 2005; 67: 25–33. Widdess-Walsh P, Jeha L, Nair D, Kotagal P, Bingaman W, Najm I. Subdural electrode analysis in focal cortical dysplasia. Neurology 2007; 69: 660–7. Wilder BJ, King RL, Schmidt RP. Cortical and subcortical secondary epileptogenesis. Neurology 1969; 19: 643–52. Wolf HK, Wiestler OD. Surgical pathology of chronic epileptic seizure disorders. Brain Pathol 1993; 3: 371–80. Worrell GA, So EL, Kazemi J, O’brien TJ, Mosewich RK, Cascino CD, et al. Focal ictal beta discharge on scalp EEG predicts excellent outcome in frontal lobe epilepsy surgery. Epilepsia 2002; 43: 277–82. Worrell GA, Parish L, Cranstoun SD, Jonas R, Baltuch G, Litt B. Highfrequency oscillations and seizure generation in neocortical epilepsy. Brain 2004; 127: 1496–506. Wyllie E, Comair YG, Kotagal P, Bulacio J, Binagaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998; 44: 740–8.