- Email: [email protected]
Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery
Clinical Neurophysiology xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery Michael P. Malter a,b, Christina Bahrenberg a, Pitt Niehusmann c, Christian E. Elger a, Rainer Surges a,⇑ a b c
Department of Epileptology, University Hospital of Bonn, Germany Department of Neurology, University Hospital Cologne, Germany Department of Neuropathology, University Hospital of Bonn, Germany
a r t i c l e
i n f o
Article history: Accepted 21 June 2015 Available online xxxx Keywords: EEG Epilepsy surgery Hippocampal sclerosis Amygdalo-hippocampectomy
h i g h l i g h t s We investigated determining factors and its predictive value of ictal scalp EEG in people with unilat-
eral mesial temporal lobe epilepsy due to hippocampal sclerosis. Neither rhythmic activity at seizure-onset nor other scalp EEG-features appeared to predict postsur-
gical seizure outcome. Ictal pattern at seizure-onset on scalp EEG gets faster with increasing epilepsy duration, possibly via
time-dependent alterations of epileptogenic networks.
a b s t r a c t Objective: To investigate determining factors of the ictal scalp EEG pattern at seizure onset and its predictive value for postsurgical outcome in people with unilateral MTLE due to hippocampal sclerosis (MTLE-HS). Methods: Review of consecutive people with chronic MTLE-HS undergoing presurgical video-EEG telemetry. Exclusion criteria were additional epileptogenic lesions or seizure generators or compromised EEG traces at seizure-onset. Mixed linear or logistic regression models were used. Results: Inclusion of 63 patients with 219 seizures with a favorable outcome (no seizures or auras only) in 43 patients at last follow-up. Rhythmic activity at seizure-onset (RA) had a frequency of 4.7 ± 1.5/s (range 1–8/s), mostly localized in the anterior temporal region. Postsurgical seizure outcome was not associated with any clinical or electrophysiological feature. RA in the delta-band was more often observed with shorter epilepsy duration (p = 0.002). Conclusions: RA on scalp EEG gets faster with increasing epilepsy duration, possibly via time-dependent alterations of epileptogenic networks. Neither the frequency of RA nor other EEG-features appeared to predict postsurgical seizure outcome. Significance: The results challenge the view that if patients with apparent MTLE display RA in the deltaband, seizure-onset in neocortical structures rather than in temporo-mesial tissue should be considered and further investigations should be prompted. Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Mesial temporal lobe epilepsy due to hippocampal sclerosis (MTLE-HS) is a well-recognized syndrome which tends to be ⇑ Corresponding author at: Department of Epileptology, Sigmund-Freud-Straße 25, 53127 Bonn, Germany. Tel.: +49 228 287 14778; fax: +49 228 287 14328. E-mail address: [email protected] (R. Surges).
intractable to anticonvulsant drugs in a high proportion of patients (Semah et al., 1998; Kim et al., 1999). In MTLE-HS, however, resective epilepsy surgery frequently leads to a favorable outcome (Wiebe et al., 2001; Kumlien et al., 2002; Yasuda et al., 2006). Seizure freedom rates after standard operative procedures (anterior temporal lobectomy or selective amygdalo-hippocampectomy) range from 49% to 90% (Kilpatrick et al., 1999; Jutila et al., 2002; Jeong et al., 2005; Chin et al., 2006; Dupont et al., 2006; Spencer
http://dx.doi.org/10.1016/j.clinph.2015.06.035 1388-2457/Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
2
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
and Huh, 2008; Bien et al., 2013; Malter et al., 2013). In the past years, more selective resections such as selective amygdalohippocampectomy (SAHE) have been favored because they were considered to have a better cognitive postoperative outcome than anterior temporal lobectomy (ATLE) while seizure outcome was comparable in both procedures (Morino et al., 2006; Paglioli et al., 2006; Helmstaedter et al., 2008; Wendling et al., 2013). However, there is still a relevant failure rate of surgery with persistent seizures in patients with MTLE-HS. Causes for surgical failure are still matter of debate (Thom et al., 2010b). Appropriate presurgical assessment to identify hippocampal sclerosis as epileptogenic zone is essential for favorable postsurgical outcome. Ictal scalp EEG is considered to be reliable in lateralizing and localizing epileptic foci (Risinger et al., 1989; Walczak et al., 1992; Ebersole and Pacia, 1996; Pacia and Ebersole, 1997; Serles et al., 2000). Rhythmic theta and alpha activity were considered as predictors in lateralizing TLE (Walczak et al., 1992). The scalp EEG pattern of seizures arising from the hippocampus commonly consists of rhythmic activity (RA) between 5 and 9/s, whereas an RA below this frequency range appears to indicate more often seizure generation outside the mesial temporal structures (Ebersole and Pacia, 1996), possibly prompting further investigation using intracranial EEG electrodes. In view of these data, we aimed at investigating the predictive value for postsurgical seizure outcome as well as the influencing factors of the ictal pattern at seizure-onset on scalp EEG recordings in people with unilateral chronic MTLE-HS. To do so, we have applied very conservative inclusion criteria considering only those patients with concordant MRI and electro-clinical findings who underwent SAHE with subsequently confirmed hippocampal sclerosis in most cases.
2. Methods 2.1. Patient selection We have retrospectively reviewed consecutive patients with chronic MTLE-HS who underwent presurgical assessment at the Department of Epileptology at the University of Bonn between January 2003 and December 2010. Inclusion criteria comprised (1) presence of medically intractable MTLE with qualitative MRI features of unilateral HS without other pathology on cranial MRI, (2) concordant electro-clinical data assessed during video-EEG monitoring, (3) performance of SAHE, (4) histological confirmation of hippocampal pathology and (5) postsurgical follow-up of at least one year. Patients were excluded from further analysis when (1) MRI or video-EEG data showed or suggested additional seizure generators and (2) interpretation of EEG traces at clinical or electrophysiological seizure onset was impossible due to artifacts.
2.2. EEG analysis All patients underwent non-invasive video-EEG-monitoring with scalp electrodes according to the 10/20 system with additional anterior temporal electrodes of T1 and T2 or bilateral sphenoidal electrodes. At least one typical seizure per patient was recorded. EEG data acquisition was performed with a Stellate Harmonie digital video-EEG system (Version 5.4, Schwarzer GmbH/Natus, Germany) using up to 128 channels, a 200 Hz sampling rate and a 16 bit analogue-to-digital converter. Data were band pass filtered between 0.016 and 70 Hz. All ictal EEG recordings were reanalyzed for the study by at least one experienced senior consultant in epileptology (MPM, RS). Recordings which were difficult to interpret were analyzed and discussed together (MPM, RS) and consensus was reached. Unclear or heavily
compromised recordings were excluded. Data of interictal focus were taken from the records. The following ictal EEG features were analyzed: RA at seizure-onset (grouped into rhythmic theta-/alpha (5–9/s) vs. delta-waves (<5/s); localization of ictal activity (ipsilateral anterior temporal/sphenoidal, ipsilateral temporal, ipsilateral posterior temporal, right/left hemispheric, without localization); contralateral seizure propagation and its latency; latency of clinical signs and ictal discharges, appearance of interictal discharges (slowing, non-epileptic discharges, ipsilateral/contralateral epileptic discharges to side of surgery). Most of the patients had more than one seizure. If individual patients had seizures with different frequencies at seizure-onset, the predominant pattern determined the assignment to the delta- or theta-group. As the differences in frequencies at seizure-onset were usually small (in the range of 1/s), our approach appears acceptable. Furthermore clinical and demographical co-factors were considered. 2.3. MRI Brain MRIs were obtained according to a dedicated epilepsy protocol (Urbach et al., 2004) in clinical routine diagnostics using a 3 Tesla scanner, Philips, The Netherlands (Dept. of Neuroradiology) and hippocampi were visually classified according to qualitative criteria (increase in signal intensity on T2-weighted images, loss of internal structure, volume loss) by experienced neuroradiologists as well as by experienced epileptologists. 2.4. Surgery All patients underwent SAHE via a transsylvian or subtemporal approach (von Rhein et al., 2012). Seizure outcome was classified according to the ILAE-classification (Wieser et al., 2001). Patients with ILAE-classification I and II were rated as good clinical outcome. Standardized neuropathological analysis was performed in all patients (Blumcke et al., 2007). For the study, all available biopsy samples were re-evaluated. 2.5. Neuropathology Biopsy samples were fixed with formaldehyde overnight, embedded into paraffin, cut into 4 lm sections and mounted on slides (HistoBond, Marienfelde, Germany). Neuropathological analysis comprised at least hematoxylin and eosin staining as well as immunohistochemistry for glial fibrillary acidic protein (DakoCytomation, Glostrup, Denmark) and NeuN (Chemicon, Temecula). Qualitative assessment of hippocampal cell loss and reactive gliosis was performed by experienced neuropathologists as part of diagnostic standard procedure including a semiquantitative classification of the extent of neuronal cell loss in the subfields CA1–4. In accordance with others (Thom et al., 2010a), the CA4 sector was defined as the region within the arms of the DG. This region included cells of the polymorphic layer in addition to hilar pyramidal neurones. This semiquantitative classification ranged from no/ mild (0) cell loss, over moderate (+, <50%), and severe (++, 50–75%) to extensive cell loss (+++, >75%). Furthermore, all specimens were classified according to Wyler as well as the ILAE (Wyler et al., 1992; Blumcke et al., 2013). 2.6. Statistical analysis Statistical significance was tested using univariate logistic regression or mixed linear regression models allowing for seizure clustering. Due to multiple comparisons, p-values were adjusted according to the Holm–Bonferroni stepwise correction procedure,
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
3
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
p-value was considered significant if <0.05. Data are given as mean ± SD (range).
Table 2 Association between postsurgical seizure outcome and (para-) clinical and EEG features.
3. Results 3.1. General characteristics of patients and seizures We reviewed 1089 people with refractory focal epilepsy who underwent presurgical assessment. A total of 98 patients with chronic MTLE-HS and consecutive SAHE were identified of whom a further 35 people had to be excluded due to EEG artefacts or non-congruent findings in ictal scalp EEG. Finally, sixty-three patients (32 men, 31 women) with 219 seizures fulfilled the conservative inclusion criteria (see Table 1). Age at surgery was 38.7 ± 13.2 years (range 16–70) and epilepsy duration at time of surgery was 22.6 ± 14.0 years (range 3–63). Postsurgical follow-up amounted to 38.2 ± 25.4 months (range 11–112, in two patients only 11 months last follow-up). A favorable outcome (ILAE I-II) was achieved in 39 patients (62%) after 11–12 months and in 43 patients (68%) at last follow-up. Clinical data were without predictive value for seizure outcome (Table 2 for details). A total of 219 complex partial seizures were recorded, of which 110 (50.2%) secondarily developed into a generalized tonic–clonic seizure. Detailed ictal EEG data are given in Table 3. The RA at seizure-onset had a mean frequency of 4.8 ± 1.5/s (range 1–8/s). Initial RA in the theta/alpha band was observed in 124 (57%) seizures (mean frequency 5.9/s, range: 5–8/s) and in the delta band in 95 (43%) seizures (mean frequency 3.3/s, range: 1–4/s), illustrative examples for seizure onset pattern are given Figs. 1 and 2. Multiple seizures were recorded in 52 of the 63 patients and showed high intraindividual consistency in 39 of 52 patients,
Table 1 Patients characteristics (n = 63). Patients characteristics (N = 63)
Sex (m/F) Age of onset Age at surgery Epilepsy duration at surgery Invasive telemetry prior to surgery Postoperative MRI findings Surgical approach Anticonvulsant drug regime (changed versus unchanged) Interictal focus Localization interictal focus Localization ictal focus Propagation to contralateral side Latency symptoms and EEG Degree of hippocampal pathology
Seizure outcome 12 months (ILAE I–II vs. PIII) p-value
Seizure outcome last follow up (ILAE I–II vs. PIII) p-value
0.537 0.668 0.662 0.959
0.932 0.693 0.078 0.177
0.835
0.315
0.946
0.606
0.618 0.009+
0.536 0.369
0.629 0.173
0.553 0.103
0.372 0.936
0.194 0.610
0.958
0.366
*+
0.020
0.863
Significant p-values after Bonferroni–Holm correction are bold. * Degree of cell loss in hippocampal subfield CA3. + n.s. after Bonferroni–Holm stepwise correction procedure.
Table 3 Ictal scalp EEG characteristics in seizures. EEG characteristics in seizures
Number of seizures (total N = 219)
Initial ictal pattern
Rhythmic alpha/theta Rhythmic delta
124 95
Localizing of ictal discharges
Anterior temporal/ sphenoidal Medial temporal Posterior temporal Hemispheral Non-localizing
187 14 0 14 4
Gender Median age at epilepsy onset
32m/31f 16.1 (±2.4)
MRI findings HS left HS right
30 (48%) 33 (52%)
Propagation to contralateral side
EEG interictal Ipsilateral focus Contralateral focus Bilateral focus No focus
Yes No Artefacts
93 109 17
44 (70%) 2 (3%) 14 (22%) 3 (5%)
Latency until propagation
Median
10 s (0–74)
Latency clinic to EEG onset
Median number of seizures recorded per patient (SD) Median age at surgery in years (SD)
3.5 (±2.4) 38.7 (±13.3)
117 19 60
Surgical approach Transsylvian Subtemporal Unknown
Clinic before EEG EEG before clinic EEG/clinic simultaneously Not discriminable
40 (64%) 16 (25%) 7 (11%)
MRT postsurgical Sufficient Complications Unsufficient Data missing
53 (84%) 3 (5%) 2 (3%) 5 (8%)
Favorable postsurgical seizure outcome (ILAE I-II) 12 months Last available follow up
39 (62%) 43 (68%)
AED at postsurgical 12 month-follow up Unchanged Changed Reduced Increased
19 (30%) 5 (8%) 20 (32%) 19 (30%)
m: male; f: female; HS: hippocampal sclerosis; AED: anticonvulsant drugs.
23
i.e. frequency at seizure-onset ranged by only ±1/s. At the level of individual patients, 27 patients (43%) invariably displayed RA within the theta/alpha band, and 21 patients (33%) RA exclusively within the delta band. Fifteen patients (24%) had seizures with RA in the delta- and theta/alpha-band at seizure-onset. In the latter patient group, the intra-individual difference between slowest and fastest frequencies ranged from 1 to 4/s (range of absolute frequencies in all 15 patients from 2 to 7/s). Given the relatively small variability in the frequency range of ictal EEG pattern, the patients were assigned into the frequency group of which they displayed the majority of seizures. If they had the same number of seizures in both frequency bands, they were assigned into the delta-band group. Apart from the consideration on the group level (theta/ alpha- versus delta-band), we have also analyzed the data on an
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
4
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
Fig. 1. Initial ictal activity in the alpha/theta band in surface EEG. Original EEG traces are displayed in the conventional bipolar montage. The time interval between vertical green lines is 1 second.
Fig. 2. Initial ictal activity in the delta band in surface EEG. Original EEG traces are displayed in the conventional bipolar montage. The time interval between vertical green lines is 1 second.
individual level taking every seizure into account (without grouping the data). 3.2. Predictors of postsurgical seizure outcome The ictal pattern at seizure-onset was not associated with the postsurgical seizure outcome at one year (p = 0.9) or at last follow-up (p = 0.5), regardless of the statistical model used (i.e. whether comparison was made at the group level delta- versus theta-band or every seizure was individually taken into account). In 85% of the seizures, ictal pattern at seizure-onset was localized in the ipsilateral anterior temporal region. The localization was not associated with postsurgical seizure outcome at one year follow up (p = 0.4) or last follow up (p = 0.2). Contralateral seizure propagation was observed in 93 seizures (42%) with mean latency of 10 s (range 0–74 s). The occurrence
of contralateral propagation did not predict postsurgical seizure outcome (one-year-follow up p = 0.94, last follow up p = 0.6). Furthermore, the latency between onset of clinical signs and first EEG changes had again no apparent influence on the postsurgical seizure outcome. Interictal sharp-slow wave focus occurred in most patients: in 44 of 63 (70%) ipsilateral to the HS, bilateral in 14 of 63 (22%) and solely contralateral to the HS in 2 of 63 (3%) patients. In 3 of 63 (5%), no interictal sharp-slow wave focus was detected. Postsurgical seizure outcome was independent from the features of the interictal sharp-slow wave focus. However, any modification (increase, decrease of daily doses or change of substances) of the anticonvulsant drug treatment was associated with a poorer postsurgical seizure outcome one year after epilepsy surgery (Table 2). Hippocampal specimen could be analyzed in 57 of 63 patients (1 specimen was not available; the remaining 5 were of poor
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
5
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
quality). In 56 of 57 patients, hippocampal sclerosis of variable severity was confirmed, whereas no clear cell loss was detected in one patient. There was a weak association between the degree of cell loss in the hippocampal subfield CA3 and seizure-outcome one year after surgery (higher degree of cell loss was associated with better outcome; Table 3), but not with the Wyler or ILAE classification (p = 0.40 and 0.44, respectively).
3.3. Determining factors of ictal pattern at seizure-onset in MTLE-HS patients Surprisingly, the frequencies of RA at seizure-onset appeared to be significantly associated with epilepsy duration, see Fig. 3. People with initial RA in the theta band had longer epilepsy duration than those with RA in the delta band (18 years vs. 26.8 years, p = 0.002). In addition, the degree of cell loss in the hippocampal subfield CA4 was weakly associated with the EEG-frequency at seizure-onset (a lower frequency at seizure-onset was correlated with a lower degree of cell loss), but not with the Wyler or ILAE classification (p = 0.43 and 0.53, respectively). In further analysis, none of the other features was correlated with the EEG pattern at seizureonset (Table 4).
Table 4 Association between frequency of rhythmic activity at seizure-onset and para-clinical or other EEG features. Frequency of initial ictal pattern
p-Value
95% CI
Sex Age of onset Age at surgery Epilepsy duration at telemetry Outcome 1 year Outcome last FU Localization interictal focus Propagation to contralateral side Propagation time to contralateral side Latency symptoms and EEG Latency time symptoms and EEG Degree of hippocampal pathology
0.941 0.083 0.109 0.002 0.877 0.489 0.204 0.879 0.881 0.525 0.173 0.031*+
0.755–0.701 0.057–0.0035 0.0049–0.049 0.0144–0.0633 0.811–0.692 0.505–1.057 0.667–0.143 0.248–0.290 0.011–0.012 0.202–0.103 0.0042–0.023 0.048–0.996
Mixed linear regression analysis, adjusted for variable seizure numbers per patient. Significant p-values after Bonferroni–Holm correction are bold. * Degree of cell loss in subfield CA4. + n.s. after Bonferroni–Holm stepwise correction procedure.
in temporo-mesial tissue should be considered and further investigations should be prompted. 4.1. EEG pattern at seizure-onset
4. Discussion This retrospective study was designed to investigate the characteristics of ictal patterns on scalp EEG, their determining factors as well as their predictive value for postsurgical seizure outcome in people with unilateral MTLE-HS. In summary, we have found that ictal patterns at seizure-onset exclusively in the delta-band (21 of 63 patients, 33%) were only slightly less common than those exclusively in the theta-/alpha-band (27 of 63 patients, 43%) and that a considerable proportion of the patients (15 of 63 patients, 24%) displayed some variability in the frequency of RA within the deltaand theta-/alpha-band. Furthermore, neither the frequency of RA at seizure-onset nor other EEG-features appeared to predict postsurgical seizure outcome. These results challenge the view that if patients with apparent MTLE display RA in the delta-band at seizure-onset, seizure-onset in neocortical structures rather than
As a hallmark study, Ebersole and Pacia (1996) investigated scalp EEG recordings in 93 people with TLE and categorized several typical ictal scalp EEG pattern as predictors for hippocampal or neocortical temporal seizure onset. People with MTLE had ictal onset pattern of rhythmic theta activity in scalp EEG in 86% of cases, whereas initial delta rhythm was associated with neocortical temporal seizure onset in 84% of seizures. The authors reproduced their findings later in simultaneous scalp/intracranial EEG recordings in a smaller study group (Pacia and Ebersole, 1997), but they did not provide etiological or postsurgical data for further confirmation of their hypothesis on seizure onset. In a further study, the same working group (Assaf and Ebersole, 1999) investigated whether visual or quantitative ictal scalp EEG recordings predict surgical outcome after ATLE of patients with MTLE suggestive of unilateral HS. In line with their previous studies, patients with theta rhythm as ictal onset pattern on visual scalp EEG had a significant better
Fig. 3. Relation between initial seizure frequency (x/s) and disease duration (years) at presurgical evaluation.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
6
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
postsurgical outcome than those with delta rhythms, but HS could not be confirmed in all histopathological specimens. The different results between these studies and our work may be due to the different methods to prove mesiotemporal seizure-generation (findings in intracranial EEG recordings subsequent to non-invasive EEG recordings; combination of clinical features strongly suggestive for MTLE or by MRI demonstrating a tumor or by direct referring to surgery), heterogeneous etiologies of TLE and different surgical procedure. In line with our results, in a recent study including 152 seizures of 42 people with MTLE-HS, rhythmic theta/alpha activity was observed in about half of the patients (53.9%) and in about two third of those patients who were seizure-free after surgery (Sirin et al., 2013). However, attenuation of background activity, cessation of interictal epileptiform discharges and repetitive spiking were seen more frequently in the non-seizure-free group (59.1% of seizures). In contrast, Dericioglu and colleagues (Dericioglu et al., 2008) analyzed 71 ictal scalp EEG recordings of 25 patients and reported that at seizure-onset, cessation of interictal epileptiform discharges was the most frequent pattern (35.2%) followed by ipsilateral regional temporal rhythmic delta/theta activity (24%). They did not clearly differentiate between theta- or delta-rhythms so that comparison of data with ours is hampered. The differences between the results of Dericioglu (Dericioglu et al., 2008) and others could also be explained by different criteria of determination of seizure-onset and may reflect the fact that visual analysis of EEG is, at least to some extent, subjective and influenced by the experience and views of the reporting physician. In addition, we have shown some intraindividual variability of RA in about 24% of the patients, which is likely to contribute to the different proportions of seizures within the theta-/alpha- or delta-bands. 4.2. Other EEG features Contralateral seizure propagation especially with short latencies was considered to be a negative predictor for postsurgical seizure outcome in several studies of people with MTLE (Lieb et al., 1986; Adam et al., 1994; Schulz et al., 2000; Lee et al., 2006). In our patients, we have not found such correlations, maybe due to the strict inclusion criteria, resulting into a homogeneous patient group. In addition, previous studies have suggested that localized scalp ictal EEG pattern after the onset of grossly observable ictal behavior may be inadequate for localization and that only EEG patterns appearing before impaired consciousness are valid (Engel et al., 1981; Quesney and Gloor, 1985). Here we could not find associations between clinical and EEG onset. However, the ‘‘true” seizure-onset in people with MTLE can only be detected with intracranial depths electrodes which is an inherent problem of all studies investigating scalp EEG and cannot be overcome. Furthermore, the determination of the ‘‘true” clinical seizure onset can also be very challenging and may be influenced by the individual views of the reporting physician. Interictal sharp-slow wave focus was ipsilateral to resection side in the majority of our patients, and exclusively contralateral to the resection side in only 3% of our patients. Again, our results did not confirm previous findings of a poorer postsurgical seizure outcome when interictal sharp-slow wave focus was bilateral or contralateral to the resection side (Steinhoff et al., 1995; Schulz et al., 2000). 4.3. Determining factors of the EEG pattern at seizure-onset Surprisingly, our data suggest that in unilateral MTLE, ictal rhythms at seizure-onset on scalp EEG gets faster with increasing epilepsy duration, possibly via time-dependent alterations of epileptogenic networks, but does not predict postsurgical seizure outcome. A previous study has described a correlation between the degree of HS and the EEG frequency at seizure-onset (Vossler
et al., 1998). In our patients, only the degree of cell loss in the hippocampal subfield CA4 was weakly associated with the EEG-frequency at seizure-onset (a greater cell loss was associated with faster frequencies of RA at seizure-onset), possibly via alterations of the hippocampal networks that determine seizure activity. The exact underlying mechanisms of this observation, however, remain unclear. No other feature appeared to influence the pattern at seizure-onset. 4.4. Study limitations There are a number of limitations to our study. We have strictly selected a subgroup of people with unilateral MTLE-HS in a tertiary referral epilepsy center and excluded patients with inconsistent MRI and EEG findings or artefacts at ictal onset. Thus, our results are difficult to apply to people with MTLE caused by other etiologies than HS, to patients with MTLE of unknown origin and to those who display inconsistent findings or other difficulties (e.g. bilateral EEG pattern, contralateral EEG onset after clinical seizure-onset). We aimed at estimating the predictive value of ictal scalp EEG features for postoperative seizure outcome in a homogeneous uniform group of people with chronic unilateral MTLE-HS which reflects a common entity in tertiary epilepsy centers, diagnosed and treated with state of the art procedures. All patients were diagnosed by 3 Tesla-MRI, which enables physicians to improve ‘‘noninvasive” diagnosis of MTLE-HS as compared to 1.5 Tesla-MRI (Winston et al., 2013) and, in fact, MRI diagnosis of MTLE-HS matched in all operated histopathological specimen. Furthermore, only people undergoing SAHE were included, as we intended to relate the value of preoperative scalp EEG more precisely on mesio-temporal structures (which is not possible in the case of anterior temporal lobectomies). This conservative approach provided a relatively ‘‘pure” and homogeneous population, strengthening the validity of our results for this particular patient group. Interpretation of ictal EEG and simultaneous video files is commonly influenced at least to some extent by subjective interpretation of the reporting physician. Our EEG files were re-evaluated for the study by two experienced senior consultants in epileptology at a tertiary epilepsy center with profound expertise in epilepsy surgery, but inter-rater reliability was not systematically assessed. In conclusion, we have shown that in unilateral MTLE-HS, the ictal pattern at seizure-onset on scalp EEG gets faster with increasing epilepsy duration, possibly via time-dependent alterations of epileptogenic networks, but does not predict postsurgical seizure outcome. Conflict of interest M.P.M. received payments for congress participation, travel expenses, lecture and manuscript preparation from UCB and EISAI. C.B. and P.N. report no conflict of interest. C.E.E. received honoraria for consultancy, expert testimony and lectures from UCB Pharma, Desitin and Pfizer. He is employee of the Life and Brain Institute Bonn. R.S. has received speaker fees from Cyberonics, EISAI, Novartis and UCB. Study sponsorship or funding None. References Adam C, Saint-Hilaire JM, Richer F. Temporal and spatial characteristics of intracerebral seizure propagation: predictive value in surgery for temporal lobe epilepsy. Epilepsia 1994;35:1065–72.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx Assaf BA, Ebersole JS. Visual and quantitative ictal EEG predictors of outcome after temporal lobectomy. Epilepsia 1999;40:52–61. Bien CG, Raabe AL, Schramm J, Becker A, Urbach H, Elger CE. Trends in presurgical evaluation and surgical treatment of epilepsy at one centre from 1988–2009. J Neurol Neurosurg Psychiatry 2013;84:54–61. Blumcke I, Pauli E, Clusmann H, Schramm J, Becker A, Elger C, et al. A new clinicopathological classification system for mesial temporal sclerosis. Acta Neuropathol 2007;113:235–44. Blumcke I, Thom M, Aronica E, Armstrong DD, Bartolomei F, Bernasconi A, et al. International consensus classification of hippocampal sclerosis in temporal lobe epilepsy: a Task Force report from the ILAE Commission on Diagnostic Methods. Epilepsia 2013;54:1315–29. Chin PS, Berg AT, Spencer SS, Lee ML, Shinnar S, Sperling MR, et al. Patient-perceived impact of resective epilepsy surgery. Neurology 2006;66:1882–7. Dericioglu N, Oguz KK, Ergun EL, Tezer FI, Saygi S. Ictal/interictal EEG patterns and functional neuroimaging findings in subcortical band heterotopia: report of three cases and review of the literature. Clin EEG Neurosci 2008;39:43–9. Dupont S, Tanguy ML, Clemenceau S, Adam C, Hazemann P, Baulac M. Long-term prognosis and psychosocial outcomes after surgery for MTLE. Epilepsia 2006;47:2115–24. Ebersole JS, Pacia SV. Localization of temporal lobe foci by ictal EEG patterns. Epilepsia 1996;37:386–99. Engel Jr J, Rausch R, Lieb JP, Kuhl DE, Crandall PH. Correlation of criteria used for localizing epileptic foci in patients considered for surgical therapy of epilepsy. Ann Neurol 1981;9:215–24. Helmstaedter C, Richter S, Roske S, Oltmanns F, Schramm J, Lehmann TN. Differential effects of temporal pole resection with amygdalohippocampectomy versus selective amygdalohippocampectomy on material-specific memory in patients with mesial temporal lobe epilepsy. Epilepsia 2008;49:88–97. Jeong SW, Lee SK, Hong KS, Kim KK, Chung CK, Kim H. Prognostic factors for the surgery for mesial temporal lobe epilepsy: longitudinal analysis. Epilepsia 2005;46:1273–9. Jutila L, Immonen A, Mervaala E, Partanen J, Partanen K, Puranen M, et al. Long term outcome of temporal lobe epilepsy surgery: analyses of 140 consecutive patients. J Neurol Neurosurg Psychiatry 2002;73:486–94. Kilpatrick C, Cook M, Matkovic Z, O’Brien T, Kaye A, Murphy M. Seizure frequency and duration of epilepsy are not risk factors for postoperative seizure outcome in patients with hippocampal sclerosis. Epilepsia 1999;40: 899–903. Kim WJ, Park SC, Lee SJ, Lee JH, Kim JY, Lee BI, et al. The prognosis for control of seizures with medications in patients with MRI evidence for mesial temporal sclerosis. Epilepsia 1999;40:290–3. Kumlien E, Doss RC, Gates JR. Treatment outcome in patients with mesial temporal sclerosis. Seizure 2002;11:413–7. Lee SA, Yim SB, Lim YM, Kang JK, Lee JK. Factors predicting seizure outcome of anterior temporal lobectomy for patients with mesial temporal sclerosis. Seizure 2006;15:397–404. Lieb JP, Engel Jr J, Babb TL. Interhemispheric propagation time of human hippocampal seizures. I. Relationship to surgical outcome. Epilepsia 1986;27:286–93. Malter MP, Tschampa HJ, Helmstaedter C, Urbach H, von Lehe M, Becker A, et al. Outcome after epilepsy surgery in patients with MRI features of bilateral ammon’s horn sclerosis. Epilepsy Res 2013;105:150–7. Morino M, Uda T, Naito K, Yoshimura M, Ishibashi K, Goto T, et al. Comparison of neuropsychological outcomes after selective amygdalohippocampectomy versus anterior temporal lobectomy. Epilepsy Behav 2006;9:95–100. Pacia SV, Ebersole JS. Intracranial EEG substrates of scalp ictal patterns from temporal lobe foci. Epilepsia 1997;38:642–54. Paglioli E, Palmini A, Portuguez M, Paglioli E, Azambuja N, da Costa JC, et al. Seizure and memory outcome following temporal lobe surgery: selective compared
7
with nonselective approaches for hippocampal sclerosis. J Neurosurg 2006;104:70–8. Quesney LF, Gloor P. Localization of epileptic foci. Electroencephalogr Clin Neurophysiol Suppl 1985;37:165–200. Risinger MW, Engel Jr J, Van Ness PC, Henry TR, Crandall PH. Ictal localization of temporal lobe seizures with scalp/sphenoidal recordings. Neurology 1989;39:1288–93. Schulz R, Luders HO, Hoppe M, Tuxhorn I, May T, Ebner A. Interictal EEG and ictal scalp EEG propagation are highly predictive of surgical outcome in mesial temporal lobe epilepsy. Epilepsia 2000;41:564–70. Semah F, Picot MC, Adam C, Broglin D, Arzimanoglou A, Bazin B, et al. Is the underlying cause of epilepsy a major prognostic factor for recurrence? Neurology 1998;51:1256–62. Serles W, Caramanos Z, Lindinger G, Pataraia E, Baumgartner C. Combining ictal surface-electroencephalography and seizure semiology improves patient lateralization in temporal lobe epilepsy. Epilepsia 2000;41:1567–73. Sirin NG, Gurses C, Bebek N, Dirican A, Baykan B, Gokyigit A. A quadruple examination of ictal EEG patterns in mesial temporal lobe epilepsy with hippocampal sclerosis: onset, propagation, later significant pattern, and termination. J Clin Neurophysiol 2013;30:329–38. Spencer S, Huh L. Outcomes of epilepsy surgery in adults and children. Lancet Neurol 2008;7:525–37. Steinhoff BJ, So NK, Lim S, Luders HO. Ictal scalp EEG in temporal lobe epilepsy with unitemporal versus bitemporal interictal epileptiform discharges. Neurology 1995;45:889–96. Thom M, Liagkouras I, Elliot KJ, Martinian L, Harkness W, McEvoy A, et al. Reliability of patterns of hippocampal sclerosis as predictors of postsurgical outcome. Epilepsia 2010a;51:1801–8. Thom M, Mathern GW, Cross JH, Bertram EH. Mesial temporal lobe epilepsy: how do we improve surgical outcome? Ann Neurol 2010b;68:424–34. Urbach H, Hattingen J, von Oertzen J, Luyken C, Clusmann H, Kral T, et al. MR imaging in the presurgical workup of patients with drug-resistant epilepsy. Am J Neuroradiol 2004;25:919–26. von Rhein B, Nelles M, Urbach H, Von Lehe M, Schramm J, Helmstaedter C. Neuropsychological outcome after selective amygdalohippocampectomy: subtemporal versus transsylvian approach. J Neurol Neurosurg Psychiatry 2012;83:887–93. Vossler DG, Kraemer DL, Knowlton RC, Kjos BO, Rostad SW, Wyler AR, et al. Temporal ictal electroencephalographic frequency correlates with hippocampal atrophy and sclerosis. Ann Neurol 1998;43:756–62. Walczak TS, Radtke RA, Lewis DV. Accuracy and interobserver reliability of scalp ictal EEG. Neurology 1992;42:2279–85. Wendling AS, Hirsch E, Wisniewski I, Davanture C, Ofer I, Zentner J, et al. Selective amygdalohippocampectomy versus standard temporal lobectomy in patients with mesial temporal lobe epilepsy and unilateral hippocampal sclerosis. Epilepsy Res 2013;104:94–104. Wiebe S, Blume WT, Girvin JP, Eliasziw MEffectiveness, Efficiency of Surgery for Temporal Lobe Epilepsy Study G. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med 2001;345:311–8. Wieser HG, Blume WT, Fish D, Goldensohn E, Hufnagel A, King D, et al. ILAE Commission Report. Proposal for a new classification of outcome with respect to epileptic seizures following epilepsy surgery. Epilepsia 2001;42:282–6. Winston GP, Micallef C, Kendell BE, Bartlett PA, Williams EJ, Burdett JL, et al. The value of repeat neuroimaging for epilepsy at a tertiary referral centre: 16 years of experience. Epilepsy Res 2013;105:349–55. Wyler AR, Dohan FC, Schweitzer FB, Barry AD. A grading system for mesial temporal pathology (hippocampal sclerosis) from anterior temporal lobectomy. J Epilepsy 1992;5:220–5. Yasuda CL, Tedeschi H, Oliveira EL, Ribas GC, Costa AL, Cardoso TA, et al. Comparison of short-term outcome between surgical and clinical treatment in temporal lobe epilepsy: a prospective study. Seizure 2006;15:35–40.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery Michael P. Malter a,b, Christina Bahrenberg a, Pitt Niehusmann c, Christian E. Elger a, Rainer Surges a,⇑ a b c
Department of Epileptology, University Hospital of Bonn, Germany Department of Neurology, University Hospital Cologne, Germany Department of Neuropathology, University Hospital of Bonn, Germany
a r t i c l e
i n f o
Article history: Accepted 21 June 2015 Available online xxxx Keywords: EEG Epilepsy surgery Hippocampal sclerosis Amygdalo-hippocampectomy
h i g h l i g h t s We investigated determining factors and its predictive value of ictal scalp EEG in people with unilat-
eral mesial temporal lobe epilepsy due to hippocampal sclerosis. Neither rhythmic activity at seizure-onset nor other scalp EEG-features appeared to predict postsur-
gical seizure outcome. Ictal pattern at seizure-onset on scalp EEG gets faster with increasing epilepsy duration, possibly via
time-dependent alterations of epileptogenic networks.
a b s t r a c t Objective: To investigate determining factors of the ictal scalp EEG pattern at seizure onset and its predictive value for postsurgical outcome in people with unilateral MTLE due to hippocampal sclerosis (MTLE-HS). Methods: Review of consecutive people with chronic MTLE-HS undergoing presurgical video-EEG telemetry. Exclusion criteria were additional epileptogenic lesions or seizure generators or compromised EEG traces at seizure-onset. Mixed linear or logistic regression models were used. Results: Inclusion of 63 patients with 219 seizures with a favorable outcome (no seizures or auras only) in 43 patients at last follow-up. Rhythmic activity at seizure-onset (RA) had a frequency of 4.7 ± 1.5/s (range 1–8/s), mostly localized in the anterior temporal region. Postsurgical seizure outcome was not associated with any clinical or electrophysiological feature. RA in the delta-band was more often observed with shorter epilepsy duration (p = 0.002). Conclusions: RA on scalp EEG gets faster with increasing epilepsy duration, possibly via time-dependent alterations of epileptogenic networks. Neither the frequency of RA nor other EEG-features appeared to predict postsurgical seizure outcome. Significance: The results challenge the view that if patients with apparent MTLE display RA in the deltaband, seizure-onset in neocortical structures rather than in temporo-mesial tissue should be considered and further investigations should be prompted. Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Mesial temporal lobe epilepsy due to hippocampal sclerosis (MTLE-HS) is a well-recognized syndrome which tends to be ⇑ Corresponding author at: Department of Epileptology, Sigmund-Freud-Straße 25, 53127 Bonn, Germany. Tel.: +49 228 287 14778; fax: +49 228 287 14328. E-mail address: [email protected] (R. Surges).
intractable to anticonvulsant drugs in a high proportion of patients (Semah et al., 1998; Kim et al., 1999). In MTLE-HS, however, resective epilepsy surgery frequently leads to a favorable outcome (Wiebe et al., 2001; Kumlien et al., 2002; Yasuda et al., 2006). Seizure freedom rates after standard operative procedures (anterior temporal lobectomy or selective amygdalo-hippocampectomy) range from 49% to 90% (Kilpatrick et al., 1999; Jutila et al., 2002; Jeong et al., 2005; Chin et al., 2006; Dupont et al., 2006; Spencer
http://dx.doi.org/10.1016/j.clinph.2015.06.035 1388-2457/Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
2
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
and Huh, 2008; Bien et al., 2013; Malter et al., 2013). In the past years, more selective resections such as selective amygdalohippocampectomy (SAHE) have been favored because they were considered to have a better cognitive postoperative outcome than anterior temporal lobectomy (ATLE) while seizure outcome was comparable in both procedures (Morino et al., 2006; Paglioli et al., 2006; Helmstaedter et al., 2008; Wendling et al., 2013). However, there is still a relevant failure rate of surgery with persistent seizures in patients with MTLE-HS. Causes for surgical failure are still matter of debate (Thom et al., 2010b). Appropriate presurgical assessment to identify hippocampal sclerosis as epileptogenic zone is essential for favorable postsurgical outcome. Ictal scalp EEG is considered to be reliable in lateralizing and localizing epileptic foci (Risinger et al., 1989; Walczak et al., 1992; Ebersole and Pacia, 1996; Pacia and Ebersole, 1997; Serles et al., 2000). Rhythmic theta and alpha activity were considered as predictors in lateralizing TLE (Walczak et al., 1992). The scalp EEG pattern of seizures arising from the hippocampus commonly consists of rhythmic activity (RA) between 5 and 9/s, whereas an RA below this frequency range appears to indicate more often seizure generation outside the mesial temporal structures (Ebersole and Pacia, 1996), possibly prompting further investigation using intracranial EEG electrodes. In view of these data, we aimed at investigating the predictive value for postsurgical seizure outcome as well as the influencing factors of the ictal pattern at seizure-onset on scalp EEG recordings in people with unilateral chronic MTLE-HS. To do so, we have applied very conservative inclusion criteria considering only those patients with concordant MRI and electro-clinical findings who underwent SAHE with subsequently confirmed hippocampal sclerosis in most cases.
2. Methods 2.1. Patient selection We have retrospectively reviewed consecutive patients with chronic MTLE-HS who underwent presurgical assessment at the Department of Epileptology at the University of Bonn between January 2003 and December 2010. Inclusion criteria comprised (1) presence of medically intractable MTLE with qualitative MRI features of unilateral HS without other pathology on cranial MRI, (2) concordant electro-clinical data assessed during video-EEG monitoring, (3) performance of SAHE, (4) histological confirmation of hippocampal pathology and (5) postsurgical follow-up of at least one year. Patients were excluded from further analysis when (1) MRI or video-EEG data showed or suggested additional seizure generators and (2) interpretation of EEG traces at clinical or electrophysiological seizure onset was impossible due to artifacts.
2.2. EEG analysis All patients underwent non-invasive video-EEG-monitoring with scalp electrodes according to the 10/20 system with additional anterior temporal electrodes of T1 and T2 or bilateral sphenoidal electrodes. At least one typical seizure per patient was recorded. EEG data acquisition was performed with a Stellate Harmonie digital video-EEG system (Version 5.4, Schwarzer GmbH/Natus, Germany) using up to 128 channels, a 200 Hz sampling rate and a 16 bit analogue-to-digital converter. Data were band pass filtered between 0.016 and 70 Hz. All ictal EEG recordings were reanalyzed for the study by at least one experienced senior consultant in epileptology (MPM, RS). Recordings which were difficult to interpret were analyzed and discussed together (MPM, RS) and consensus was reached. Unclear or heavily
compromised recordings were excluded. Data of interictal focus were taken from the records. The following ictal EEG features were analyzed: RA at seizure-onset (grouped into rhythmic theta-/alpha (5–9/s) vs. delta-waves (<5/s); localization of ictal activity (ipsilateral anterior temporal/sphenoidal, ipsilateral temporal, ipsilateral posterior temporal, right/left hemispheric, without localization); contralateral seizure propagation and its latency; latency of clinical signs and ictal discharges, appearance of interictal discharges (slowing, non-epileptic discharges, ipsilateral/contralateral epileptic discharges to side of surgery). Most of the patients had more than one seizure. If individual patients had seizures with different frequencies at seizure-onset, the predominant pattern determined the assignment to the delta- or theta-group. As the differences in frequencies at seizure-onset were usually small (in the range of 1/s), our approach appears acceptable. Furthermore clinical and demographical co-factors were considered. 2.3. MRI Brain MRIs were obtained according to a dedicated epilepsy protocol (Urbach et al., 2004) in clinical routine diagnostics using a 3 Tesla scanner, Philips, The Netherlands (Dept. of Neuroradiology) and hippocampi were visually classified according to qualitative criteria (increase in signal intensity on T2-weighted images, loss of internal structure, volume loss) by experienced neuroradiologists as well as by experienced epileptologists. 2.4. Surgery All patients underwent SAHE via a transsylvian or subtemporal approach (von Rhein et al., 2012). Seizure outcome was classified according to the ILAE-classification (Wieser et al., 2001). Patients with ILAE-classification I and II were rated as good clinical outcome. Standardized neuropathological analysis was performed in all patients (Blumcke et al., 2007). For the study, all available biopsy samples were re-evaluated. 2.5. Neuropathology Biopsy samples were fixed with formaldehyde overnight, embedded into paraffin, cut into 4 lm sections and mounted on slides (HistoBond, Marienfelde, Germany). Neuropathological analysis comprised at least hematoxylin and eosin staining as well as immunohistochemistry for glial fibrillary acidic protein (DakoCytomation, Glostrup, Denmark) and NeuN (Chemicon, Temecula). Qualitative assessment of hippocampal cell loss and reactive gliosis was performed by experienced neuropathologists as part of diagnostic standard procedure including a semiquantitative classification of the extent of neuronal cell loss in the subfields CA1–4. In accordance with others (Thom et al., 2010a), the CA4 sector was defined as the region within the arms of the DG. This region included cells of the polymorphic layer in addition to hilar pyramidal neurones. This semiquantitative classification ranged from no/ mild (0) cell loss, over moderate (+, <50%), and severe (++, 50–75%) to extensive cell loss (+++, >75%). Furthermore, all specimens were classified according to Wyler as well as the ILAE (Wyler et al., 1992; Blumcke et al., 2013). 2.6. Statistical analysis Statistical significance was tested using univariate logistic regression or mixed linear regression models allowing for seizure clustering. Due to multiple comparisons, p-values were adjusted according to the Holm–Bonferroni stepwise correction procedure,
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
3
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
p-value was considered significant if <0.05. Data are given as mean ± SD (range).
Table 2 Association between postsurgical seizure outcome and (para-) clinical and EEG features.
3. Results 3.1. General characteristics of patients and seizures We reviewed 1089 people with refractory focal epilepsy who underwent presurgical assessment. A total of 98 patients with chronic MTLE-HS and consecutive SAHE were identified of whom a further 35 people had to be excluded due to EEG artefacts or non-congruent findings in ictal scalp EEG. Finally, sixty-three patients (32 men, 31 women) with 219 seizures fulfilled the conservative inclusion criteria (see Table 1). Age at surgery was 38.7 ± 13.2 years (range 16–70) and epilepsy duration at time of surgery was 22.6 ± 14.0 years (range 3–63). Postsurgical follow-up amounted to 38.2 ± 25.4 months (range 11–112, in two patients only 11 months last follow-up). A favorable outcome (ILAE I-II) was achieved in 39 patients (62%) after 11–12 months and in 43 patients (68%) at last follow-up. Clinical data were without predictive value for seizure outcome (Table 2 for details). A total of 219 complex partial seizures were recorded, of which 110 (50.2%) secondarily developed into a generalized tonic–clonic seizure. Detailed ictal EEG data are given in Table 3. The RA at seizure-onset had a mean frequency of 4.8 ± 1.5/s (range 1–8/s). Initial RA in the theta/alpha band was observed in 124 (57%) seizures (mean frequency 5.9/s, range: 5–8/s) and in the delta band in 95 (43%) seizures (mean frequency 3.3/s, range: 1–4/s), illustrative examples for seizure onset pattern are given Figs. 1 and 2. Multiple seizures were recorded in 52 of the 63 patients and showed high intraindividual consistency in 39 of 52 patients,
Table 1 Patients characteristics (n = 63). Patients characteristics (N = 63)
Sex (m/F) Age of onset Age at surgery Epilepsy duration at surgery Invasive telemetry prior to surgery Postoperative MRI findings Surgical approach Anticonvulsant drug regime (changed versus unchanged) Interictal focus Localization interictal focus Localization ictal focus Propagation to contralateral side Latency symptoms and EEG Degree of hippocampal pathology
Seizure outcome 12 months (ILAE I–II vs. PIII) p-value
Seizure outcome last follow up (ILAE I–II vs. PIII) p-value
0.537 0.668 0.662 0.959
0.932 0.693 0.078 0.177
0.835
0.315
0.946
0.606
0.618 0.009+
0.536 0.369
0.629 0.173
0.553 0.103
0.372 0.936
0.194 0.610
0.958
0.366
*+
0.020
0.863
Significant p-values after Bonferroni–Holm correction are bold. * Degree of cell loss in hippocampal subfield CA3. + n.s. after Bonferroni–Holm stepwise correction procedure.
Table 3 Ictal scalp EEG characteristics in seizures. EEG characteristics in seizures
Number of seizures (total N = 219)
Initial ictal pattern
Rhythmic alpha/theta Rhythmic delta
124 95
Localizing of ictal discharges
Anterior temporal/ sphenoidal Medial temporal Posterior temporal Hemispheral Non-localizing
187 14 0 14 4
Gender Median age at epilepsy onset
32m/31f 16.1 (±2.4)
MRI findings HS left HS right
30 (48%) 33 (52%)
Propagation to contralateral side
EEG interictal Ipsilateral focus Contralateral focus Bilateral focus No focus
Yes No Artefacts
93 109 17
44 (70%) 2 (3%) 14 (22%) 3 (5%)
Latency until propagation
Median
10 s (0–74)
Latency clinic to EEG onset
Median number of seizures recorded per patient (SD) Median age at surgery in years (SD)
3.5 (±2.4) 38.7 (±13.3)
117 19 60
Surgical approach Transsylvian Subtemporal Unknown
Clinic before EEG EEG before clinic EEG/clinic simultaneously Not discriminable
40 (64%) 16 (25%) 7 (11%)
MRT postsurgical Sufficient Complications Unsufficient Data missing
53 (84%) 3 (5%) 2 (3%) 5 (8%)
Favorable postsurgical seizure outcome (ILAE I-II) 12 months Last available follow up
39 (62%) 43 (68%)
AED at postsurgical 12 month-follow up Unchanged Changed Reduced Increased
19 (30%) 5 (8%) 20 (32%) 19 (30%)
m: male; f: female; HS: hippocampal sclerosis; AED: anticonvulsant drugs.
23
i.e. frequency at seizure-onset ranged by only ±1/s. At the level of individual patients, 27 patients (43%) invariably displayed RA within the theta/alpha band, and 21 patients (33%) RA exclusively within the delta band. Fifteen patients (24%) had seizures with RA in the delta- and theta/alpha-band at seizure-onset. In the latter patient group, the intra-individual difference between slowest and fastest frequencies ranged from 1 to 4/s (range of absolute frequencies in all 15 patients from 2 to 7/s). Given the relatively small variability in the frequency range of ictal EEG pattern, the patients were assigned into the frequency group of which they displayed the majority of seizures. If they had the same number of seizures in both frequency bands, they were assigned into the delta-band group. Apart from the consideration on the group level (theta/ alpha- versus delta-band), we have also analyzed the data on an
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
4
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
Fig. 1. Initial ictal activity in the alpha/theta band in surface EEG. Original EEG traces are displayed in the conventional bipolar montage. The time interval between vertical green lines is 1 second.
Fig. 2. Initial ictal activity in the delta band in surface EEG. Original EEG traces are displayed in the conventional bipolar montage. The time interval between vertical green lines is 1 second.
individual level taking every seizure into account (without grouping the data). 3.2. Predictors of postsurgical seizure outcome The ictal pattern at seizure-onset was not associated with the postsurgical seizure outcome at one year (p = 0.9) or at last follow-up (p = 0.5), regardless of the statistical model used (i.e. whether comparison was made at the group level delta- versus theta-band or every seizure was individually taken into account). In 85% of the seizures, ictal pattern at seizure-onset was localized in the ipsilateral anterior temporal region. The localization was not associated with postsurgical seizure outcome at one year follow up (p = 0.4) or last follow up (p = 0.2). Contralateral seizure propagation was observed in 93 seizures (42%) with mean latency of 10 s (range 0–74 s). The occurrence
of contralateral propagation did not predict postsurgical seizure outcome (one-year-follow up p = 0.94, last follow up p = 0.6). Furthermore, the latency between onset of clinical signs and first EEG changes had again no apparent influence on the postsurgical seizure outcome. Interictal sharp-slow wave focus occurred in most patients: in 44 of 63 (70%) ipsilateral to the HS, bilateral in 14 of 63 (22%) and solely contralateral to the HS in 2 of 63 (3%) patients. In 3 of 63 (5%), no interictal sharp-slow wave focus was detected. Postsurgical seizure outcome was independent from the features of the interictal sharp-slow wave focus. However, any modification (increase, decrease of daily doses or change of substances) of the anticonvulsant drug treatment was associated with a poorer postsurgical seizure outcome one year after epilepsy surgery (Table 2). Hippocampal specimen could be analyzed in 57 of 63 patients (1 specimen was not available; the remaining 5 were of poor
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
5
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
quality). In 56 of 57 patients, hippocampal sclerosis of variable severity was confirmed, whereas no clear cell loss was detected in one patient. There was a weak association between the degree of cell loss in the hippocampal subfield CA3 and seizure-outcome one year after surgery (higher degree of cell loss was associated with better outcome; Table 3), but not with the Wyler or ILAE classification (p = 0.40 and 0.44, respectively).
3.3. Determining factors of ictal pattern at seizure-onset in MTLE-HS patients Surprisingly, the frequencies of RA at seizure-onset appeared to be significantly associated with epilepsy duration, see Fig. 3. People with initial RA in the theta band had longer epilepsy duration than those with RA in the delta band (18 years vs. 26.8 years, p = 0.002). In addition, the degree of cell loss in the hippocampal subfield CA4 was weakly associated with the EEG-frequency at seizure-onset (a lower frequency at seizure-onset was correlated with a lower degree of cell loss), but not with the Wyler or ILAE classification (p = 0.43 and 0.53, respectively). In further analysis, none of the other features was correlated with the EEG pattern at seizureonset (Table 4).
Table 4 Association between frequency of rhythmic activity at seizure-onset and para-clinical or other EEG features. Frequency of initial ictal pattern
p-Value
95% CI
Sex Age of onset Age at surgery Epilepsy duration at telemetry Outcome 1 year Outcome last FU Localization interictal focus Propagation to contralateral side Propagation time to contralateral side Latency symptoms and EEG Latency time symptoms and EEG Degree of hippocampal pathology
0.941 0.083 0.109 0.002 0.877 0.489 0.204 0.879 0.881 0.525 0.173 0.031*+
0.755–0.701 0.057–0.0035 0.0049–0.049 0.0144–0.0633 0.811–0.692 0.505–1.057 0.667–0.143 0.248–0.290 0.011–0.012 0.202–0.103 0.0042–0.023 0.048–0.996
Mixed linear regression analysis, adjusted for variable seizure numbers per patient. Significant p-values after Bonferroni–Holm correction are bold. * Degree of cell loss in subfield CA4. + n.s. after Bonferroni–Holm stepwise correction procedure.
in temporo-mesial tissue should be considered and further investigations should be prompted. 4.1. EEG pattern at seizure-onset
4. Discussion This retrospective study was designed to investigate the characteristics of ictal patterns on scalp EEG, their determining factors as well as their predictive value for postsurgical seizure outcome in people with unilateral MTLE-HS. In summary, we have found that ictal patterns at seizure-onset exclusively in the delta-band (21 of 63 patients, 33%) were only slightly less common than those exclusively in the theta-/alpha-band (27 of 63 patients, 43%) and that a considerable proportion of the patients (15 of 63 patients, 24%) displayed some variability in the frequency of RA within the deltaand theta-/alpha-band. Furthermore, neither the frequency of RA at seizure-onset nor other EEG-features appeared to predict postsurgical seizure outcome. These results challenge the view that if patients with apparent MTLE display RA in the delta-band at seizure-onset, seizure-onset in neocortical structures rather than
As a hallmark study, Ebersole and Pacia (1996) investigated scalp EEG recordings in 93 people with TLE and categorized several typical ictal scalp EEG pattern as predictors for hippocampal or neocortical temporal seizure onset. People with MTLE had ictal onset pattern of rhythmic theta activity in scalp EEG in 86% of cases, whereas initial delta rhythm was associated with neocortical temporal seizure onset in 84% of seizures. The authors reproduced their findings later in simultaneous scalp/intracranial EEG recordings in a smaller study group (Pacia and Ebersole, 1997), but they did not provide etiological or postsurgical data for further confirmation of their hypothesis on seizure onset. In a further study, the same working group (Assaf and Ebersole, 1999) investigated whether visual or quantitative ictal scalp EEG recordings predict surgical outcome after ATLE of patients with MTLE suggestive of unilateral HS. In line with their previous studies, patients with theta rhythm as ictal onset pattern on visual scalp EEG had a significant better
Fig. 3. Relation between initial seizure frequency (x/s) and disease duration (years) at presurgical evaluation.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
6
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx
postsurgical outcome than those with delta rhythms, but HS could not be confirmed in all histopathological specimens. The different results between these studies and our work may be due to the different methods to prove mesiotemporal seizure-generation (findings in intracranial EEG recordings subsequent to non-invasive EEG recordings; combination of clinical features strongly suggestive for MTLE or by MRI demonstrating a tumor or by direct referring to surgery), heterogeneous etiologies of TLE and different surgical procedure. In line with our results, in a recent study including 152 seizures of 42 people with MTLE-HS, rhythmic theta/alpha activity was observed in about half of the patients (53.9%) and in about two third of those patients who were seizure-free after surgery (Sirin et al., 2013). However, attenuation of background activity, cessation of interictal epileptiform discharges and repetitive spiking were seen more frequently in the non-seizure-free group (59.1% of seizures). In contrast, Dericioglu and colleagues (Dericioglu et al., 2008) analyzed 71 ictal scalp EEG recordings of 25 patients and reported that at seizure-onset, cessation of interictal epileptiform discharges was the most frequent pattern (35.2%) followed by ipsilateral regional temporal rhythmic delta/theta activity (24%). They did not clearly differentiate between theta- or delta-rhythms so that comparison of data with ours is hampered. The differences between the results of Dericioglu (Dericioglu et al., 2008) and others could also be explained by different criteria of determination of seizure-onset and may reflect the fact that visual analysis of EEG is, at least to some extent, subjective and influenced by the experience and views of the reporting physician. In addition, we have shown some intraindividual variability of RA in about 24% of the patients, which is likely to contribute to the different proportions of seizures within the theta-/alpha- or delta-bands. 4.2. Other EEG features Contralateral seizure propagation especially with short latencies was considered to be a negative predictor for postsurgical seizure outcome in several studies of people with MTLE (Lieb et al., 1986; Adam et al., 1994; Schulz et al., 2000; Lee et al., 2006). In our patients, we have not found such correlations, maybe due to the strict inclusion criteria, resulting into a homogeneous patient group. In addition, previous studies have suggested that localized scalp ictal EEG pattern after the onset of grossly observable ictal behavior may be inadequate for localization and that only EEG patterns appearing before impaired consciousness are valid (Engel et al., 1981; Quesney and Gloor, 1985). Here we could not find associations between clinical and EEG onset. However, the ‘‘true” seizure-onset in people with MTLE can only be detected with intracranial depths electrodes which is an inherent problem of all studies investigating scalp EEG and cannot be overcome. Furthermore, the determination of the ‘‘true” clinical seizure onset can also be very challenging and may be influenced by the individual views of the reporting physician. Interictal sharp-slow wave focus was ipsilateral to resection side in the majority of our patients, and exclusively contralateral to the resection side in only 3% of our patients. Again, our results did not confirm previous findings of a poorer postsurgical seizure outcome when interictal sharp-slow wave focus was bilateral or contralateral to the resection side (Steinhoff et al., 1995; Schulz et al., 2000). 4.3. Determining factors of the EEG pattern at seizure-onset Surprisingly, our data suggest that in unilateral MTLE, ictal rhythms at seizure-onset on scalp EEG gets faster with increasing epilepsy duration, possibly via time-dependent alterations of epileptogenic networks, but does not predict postsurgical seizure outcome. A previous study has described a correlation between the degree of HS and the EEG frequency at seizure-onset (Vossler
et al., 1998). In our patients, only the degree of cell loss in the hippocampal subfield CA4 was weakly associated with the EEG-frequency at seizure-onset (a greater cell loss was associated with faster frequencies of RA at seizure-onset), possibly via alterations of the hippocampal networks that determine seizure activity. The exact underlying mechanisms of this observation, however, remain unclear. No other feature appeared to influence the pattern at seizure-onset. 4.4. Study limitations There are a number of limitations to our study. We have strictly selected a subgroup of people with unilateral MTLE-HS in a tertiary referral epilepsy center and excluded patients with inconsistent MRI and EEG findings or artefacts at ictal onset. Thus, our results are difficult to apply to people with MTLE caused by other etiologies than HS, to patients with MTLE of unknown origin and to those who display inconsistent findings or other difficulties (e.g. bilateral EEG pattern, contralateral EEG onset after clinical seizure-onset). We aimed at estimating the predictive value of ictal scalp EEG features for postoperative seizure outcome in a homogeneous uniform group of people with chronic unilateral MTLE-HS which reflects a common entity in tertiary epilepsy centers, diagnosed and treated with state of the art procedures. All patients were diagnosed by 3 Tesla-MRI, which enables physicians to improve ‘‘noninvasive” diagnosis of MTLE-HS as compared to 1.5 Tesla-MRI (Winston et al., 2013) and, in fact, MRI diagnosis of MTLE-HS matched in all operated histopathological specimen. Furthermore, only people undergoing SAHE were included, as we intended to relate the value of preoperative scalp EEG more precisely on mesio-temporal structures (which is not possible in the case of anterior temporal lobectomies). This conservative approach provided a relatively ‘‘pure” and homogeneous population, strengthening the validity of our results for this particular patient group. Interpretation of ictal EEG and simultaneous video files is commonly influenced at least to some extent by subjective interpretation of the reporting physician. Our EEG files were re-evaluated for the study by two experienced senior consultants in epileptology at a tertiary epilepsy center with profound expertise in epilepsy surgery, but inter-rater reliability was not systematically assessed. In conclusion, we have shown that in unilateral MTLE-HS, the ictal pattern at seizure-onset on scalp EEG gets faster with increasing epilepsy duration, possibly via time-dependent alterations of epileptogenic networks, but does not predict postsurgical seizure outcome. Conflict of interest M.P.M. received payments for congress participation, travel expenses, lecture and manuscript preparation from UCB and EISAI. C.B. and P.N. report no conflict of interest. C.E.E. received honoraria for consultancy, expert testimony and lectures from UCB Pharma, Desitin and Pfizer. He is employee of the Life and Brain Institute Bonn. R.S. has received speaker fees from Cyberonics, EISAI, Novartis and UCB. Study sponsorship or funding None. References Adam C, Saint-Hilaire JM, Richer F. Temporal and spatial characteristics of intracerebral seizure propagation: predictive value in surgery for temporal lobe epilepsy. Epilepsia 1994;35:1065–72.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035
M.P. Malter et al. / Clinical Neurophysiology xxx (2015) xxx–xxx Assaf BA, Ebersole JS. Visual and quantitative ictal EEG predictors of outcome after temporal lobectomy. Epilepsia 1999;40:52–61. Bien CG, Raabe AL, Schramm J, Becker A, Urbach H, Elger CE. Trends in presurgical evaluation and surgical treatment of epilepsy at one centre from 1988–2009. J Neurol Neurosurg Psychiatry 2013;84:54–61. Blumcke I, Pauli E, Clusmann H, Schramm J, Becker A, Elger C, et al. A new clinicopathological classification system for mesial temporal sclerosis. Acta Neuropathol 2007;113:235–44. Blumcke I, Thom M, Aronica E, Armstrong DD, Bartolomei F, Bernasconi A, et al. International consensus classification of hippocampal sclerosis in temporal lobe epilepsy: a Task Force report from the ILAE Commission on Diagnostic Methods. Epilepsia 2013;54:1315–29. Chin PS, Berg AT, Spencer SS, Lee ML, Shinnar S, Sperling MR, et al. Patient-perceived impact of resective epilepsy surgery. Neurology 2006;66:1882–7. Dericioglu N, Oguz KK, Ergun EL, Tezer FI, Saygi S. Ictal/interictal EEG patterns and functional neuroimaging findings in subcortical band heterotopia: report of three cases and review of the literature. Clin EEG Neurosci 2008;39:43–9. Dupont S, Tanguy ML, Clemenceau S, Adam C, Hazemann P, Baulac M. Long-term prognosis and psychosocial outcomes after surgery for MTLE. Epilepsia 2006;47:2115–24. Ebersole JS, Pacia SV. Localization of temporal lobe foci by ictal EEG patterns. Epilepsia 1996;37:386–99. Engel Jr J, Rausch R, Lieb JP, Kuhl DE, Crandall PH. Correlation of criteria used for localizing epileptic foci in patients considered for surgical therapy of epilepsy. Ann Neurol 1981;9:215–24. Helmstaedter C, Richter S, Roske S, Oltmanns F, Schramm J, Lehmann TN. Differential effects of temporal pole resection with amygdalohippocampectomy versus selective amygdalohippocampectomy on material-specific memory in patients with mesial temporal lobe epilepsy. Epilepsia 2008;49:88–97. Jeong SW, Lee SK, Hong KS, Kim KK, Chung CK, Kim H. Prognostic factors for the surgery for mesial temporal lobe epilepsy: longitudinal analysis. Epilepsia 2005;46:1273–9. Jutila L, Immonen A, Mervaala E, Partanen J, Partanen K, Puranen M, et al. Long term outcome of temporal lobe epilepsy surgery: analyses of 140 consecutive patients. J Neurol Neurosurg Psychiatry 2002;73:486–94. Kilpatrick C, Cook M, Matkovic Z, O’Brien T, Kaye A, Murphy M. Seizure frequency and duration of epilepsy are not risk factors for postoperative seizure outcome in patients with hippocampal sclerosis. Epilepsia 1999;40: 899–903. Kim WJ, Park SC, Lee SJ, Lee JH, Kim JY, Lee BI, et al. The prognosis for control of seizures with medications in patients with MRI evidence for mesial temporal sclerosis. Epilepsia 1999;40:290–3. Kumlien E, Doss RC, Gates JR. Treatment outcome in patients with mesial temporal sclerosis. Seizure 2002;11:413–7. Lee SA, Yim SB, Lim YM, Kang JK, Lee JK. Factors predicting seizure outcome of anterior temporal lobectomy for patients with mesial temporal sclerosis. Seizure 2006;15:397–404. Lieb JP, Engel Jr J, Babb TL. Interhemispheric propagation time of human hippocampal seizures. I. Relationship to surgical outcome. Epilepsia 1986;27:286–93. Malter MP, Tschampa HJ, Helmstaedter C, Urbach H, von Lehe M, Becker A, et al. Outcome after epilepsy surgery in patients with MRI features of bilateral ammon’s horn sclerosis. Epilepsy Res 2013;105:150–7. Morino M, Uda T, Naito K, Yoshimura M, Ishibashi K, Goto T, et al. Comparison of neuropsychological outcomes after selective amygdalohippocampectomy versus anterior temporal lobectomy. Epilepsy Behav 2006;9:95–100. Pacia SV, Ebersole JS. Intracranial EEG substrates of scalp ictal patterns from temporal lobe foci. Epilepsia 1997;38:642–54. Paglioli E, Palmini A, Portuguez M, Paglioli E, Azambuja N, da Costa JC, et al. Seizure and memory outcome following temporal lobe surgery: selective compared
7
with nonselective approaches for hippocampal sclerosis. J Neurosurg 2006;104:70–8. Quesney LF, Gloor P. Localization of epileptic foci. Electroencephalogr Clin Neurophysiol Suppl 1985;37:165–200. Risinger MW, Engel Jr J, Van Ness PC, Henry TR, Crandall PH. Ictal localization of temporal lobe seizures with scalp/sphenoidal recordings. Neurology 1989;39:1288–93. Schulz R, Luders HO, Hoppe M, Tuxhorn I, May T, Ebner A. Interictal EEG and ictal scalp EEG propagation are highly predictive of surgical outcome in mesial temporal lobe epilepsy. Epilepsia 2000;41:564–70. Semah F, Picot MC, Adam C, Broglin D, Arzimanoglou A, Bazin B, et al. Is the underlying cause of epilepsy a major prognostic factor for recurrence? Neurology 1998;51:1256–62. Serles W, Caramanos Z, Lindinger G, Pataraia E, Baumgartner C. Combining ictal surface-electroencephalography and seizure semiology improves patient lateralization in temporal lobe epilepsy. Epilepsia 2000;41:1567–73. Sirin NG, Gurses C, Bebek N, Dirican A, Baykan B, Gokyigit A. A quadruple examination of ictal EEG patterns in mesial temporal lobe epilepsy with hippocampal sclerosis: onset, propagation, later significant pattern, and termination. J Clin Neurophysiol 2013;30:329–38. Spencer S, Huh L. Outcomes of epilepsy surgery in adults and children. Lancet Neurol 2008;7:525–37. Steinhoff BJ, So NK, Lim S, Luders HO. Ictal scalp EEG in temporal lobe epilepsy with unitemporal versus bitemporal interictal epileptiform discharges. Neurology 1995;45:889–96. Thom M, Liagkouras I, Elliot KJ, Martinian L, Harkness W, McEvoy A, et al. Reliability of patterns of hippocampal sclerosis as predictors of postsurgical outcome. Epilepsia 2010a;51:1801–8. Thom M, Mathern GW, Cross JH, Bertram EH. Mesial temporal lobe epilepsy: how do we improve surgical outcome? Ann Neurol 2010b;68:424–34. Urbach H, Hattingen J, von Oertzen J, Luyken C, Clusmann H, Kral T, et al. MR imaging in the presurgical workup of patients with drug-resistant epilepsy. Am J Neuroradiol 2004;25:919–26. von Rhein B, Nelles M, Urbach H, Von Lehe M, Schramm J, Helmstaedter C. Neuropsychological outcome after selective amygdalohippocampectomy: subtemporal versus transsylvian approach. J Neurol Neurosurg Psychiatry 2012;83:887–93. Vossler DG, Kraemer DL, Knowlton RC, Kjos BO, Rostad SW, Wyler AR, et al. Temporal ictal electroencephalographic frequency correlates with hippocampal atrophy and sclerosis. Ann Neurol 1998;43:756–62. Walczak TS, Radtke RA, Lewis DV. Accuracy and interobserver reliability of scalp ictal EEG. Neurology 1992;42:2279–85. Wendling AS, Hirsch E, Wisniewski I, Davanture C, Ofer I, Zentner J, et al. Selective amygdalohippocampectomy versus standard temporal lobectomy in patients with mesial temporal lobe epilepsy and unilateral hippocampal sclerosis. Epilepsy Res 2013;104:94–104. Wiebe S, Blume WT, Girvin JP, Eliasziw MEffectiveness, Efficiency of Surgery for Temporal Lobe Epilepsy Study G. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med 2001;345:311–8. Wieser HG, Blume WT, Fish D, Goldensohn E, Hufnagel A, King D, et al. ILAE Commission Report. Proposal for a new classification of outcome with respect to epileptic seizures following epilepsy surgery. Epilepsia 2001;42:282–6. Winston GP, Micallef C, Kendell BE, Bartlett PA, Williams EJ, Burdett JL, et al. The value of repeat neuroimaging for epilepsy at a tertiary referral centre: 16 years of experience. Epilepsy Res 2013;105:349–55. Wyler AR, Dohan FC, Schweitzer FB, Barry AD. A grading system for mesial temporal pathology (hippocampal sclerosis) from anterior temporal lobectomy. J Epilepsy 1992;5:220–5. Yasuda CL, Tedeschi H, Oliveira EL, Ribas GC, Costa AL, Cardoso TA, et al. Comparison of short-term outcome between surgical and clinical treatment in temporal lobe epilepsy: a prospective study. Seizure 2006;15:35–40.
Please cite this article in press as: Malter MP et al. Features of scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal sclerosis: Determining factors and predictive value for epilepsy surgery. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.06.035