- Email: [email protected]
Subtle pathological changes in neocortical temporal lobe epilepsy
Epilepsy & Behavior 71 (2017) 17–22
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Subtle pathological changes in neocortical temporal lobe epilepsy☆ Juan G. Ochoa a,⁎, Diana Hentgarden a, Audrey Paulzak c, Melissa Ogden a, Richard Pryson d, Markus Lamle e, Walter G. Rusyniak b a
Department of Neurology, University of South Alabama, United States Department of Neurosurgery, University of South Alabama, United States Department of Neurosurgery, University of Rochester, United States d Department of Pathology, Cleveland Clinic Foundation, United States e Department of Radiology, University of South Alabama, United States b c
a r t i c l e
i n f o
Article history: Received 20 July 2016 Revised 7 December 2016 Accepted 7 January 2017 Available online xxxx Keywords: TLE Epilepsy surgery Parahippocampus FCD Neocortical temporal lobe epilepsy HFO
a b s t r a c t This was a prospective observational study to correlate the clinical symptoms, electrophysiology, imaging, and surgical pathology of patients with temporal lobe epilepsy (TLE) without hippocampal sclerosis. We selected consecutive patients with TLE and normal MRI undergoing temporal lobe resection between April and September 2015. Clinical features, imaging, and functional data were reviewed. Intracranial monitoring and language mapping were performed when it was required according to our team recommendation. Prior to hippocampal resection, intraoperative electrocorticography was performed using depth electrodes in the amygdala and the hippocampus. The resected hippocampus was sent for pathological analysis. Results: Five patients with diagnosis with non-lesional TLE were included. We did not find distinctive clinical features that could be a characteristic of non-lesional TLE. The mean follow-up was 13.2 months (11–15 months); 80% of patients achieved Engel Class I outcome. There was no distinctive electrographic findings in these patients. Histopathologic analysis was negative for mesial temporal sclerosis. A second blinded independent neuropathologist with expertise in epilepsy found ILAE type I focal cortical dysplasia in the parahippocampal gyrus in all patients. A third independent neuropathologist reported changes in layer 2 with larger pyramidal neurons in 4 cases but concluded that none of these cases met the diagnostic criteria of FCD. Subtle pathological changes could be associated with a parahippocampal epileptic zone and should be investigated in patients with MRI-negative TLE. This study also highlights the lack of interobserver reliability for the diagnosis of mild cortical dysplasia. Finally, selective amygdalo-hippocampectomy or laser ablation of the hippocampus may not control intractable epilepsy in this specific population. © 2017 Elsevier Inc. All rights reserved.
1. Introduction Temporal lobe epilepsy (TLE) is the most common form of epilepsy of focal origin. It is commonly associated with hippocampal sclerosis (HS) and other pathologies such as tumors and malformations of cortical development (MCD). In a study of 243 patients with TLE who underwent surgery published in 2009, only 5% had negative pathology [1]. These patients may have poor localization of their seizure focus and may be considered nonsurgical candidates. Lesional neocortical TLE (nTLE) cases are often not reported in the literature because they may be less likely to be admitted for video-EEG monitoring [2] Nonetheless, nTLE has started has started to be a recognized as a different entity from mesial TLE, in which histopathological analysis
☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ⁎ Corresponding author at: Department of Neurology, 2451 Fillingim St, 10-E, Mobile, AL 36617-2293, United States. E-mail address: [email protected] (J.G. Ochoa).
http://dx.doi.org/10.1016/j.yebeh.2017.01.001 1525-5050/© 2017 Elsevier Inc. All rights reserved.
reveals low grade or no sclerosis of hippocampal resections. In a small case series published in 2009, MCD was found in the parahippocampal gyrus of patients with only subtle but distinctive abnormalities on MRI [3]. In MRI-negative patients, the localization is more challenging, usually requiring intracranial monitoring for localization of the epileptogenic lesion. From the clinical presentation, including semiology and scalp EEG, TLE with HS and nTLE are difficult to differentiate. A study in 2010 compared the clinical features of patients with parahippocampal inferior temporal lesions and HS, finding that hypermotor and bilateral motor symptoms were more common in patients with lesions in the posterior parahippocampal gyrus group compared to patients with HS [4]. Another study reported typical TLE seizure semiology with preservation of memory and normal MRI; interestingly, hippocampal pathology reported in few of these patients was normal and MCD was found in the lateral temporal cortex [5]. Although TLE with HS is a thoroughly studied entity, the incidence, characterization, etiology, and pathophysiology of nTLE is not well documented in the literature. This study aimed to further investigate and analyze nTLE, including clinical symptoms, scalp EEG, intra-operative EEG,
18
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
high-resolution MRI, surgical histopathology, and surgical outcome of patients with medically-intractable nTLE who underwent anterior temporal lobectomy. 2. Materials and methods 2.1. Subjects and patient selection We selected consecutive patients evaluated at the University of South Alabama Comprehensive Epilepsy Program between April and September 2015. We targeted patients with medically-intractable epilepsy localized to either left or right temporal lobe by noninvasive methods and normal hippocampus on 3-T MRI, using T1, FLAIR, and T2 sequences with thin coronal cuts through the hippocampus. Diagnosis of TLE was reached through the normal epilepsy workup, including: seizure semiology, scalp EEG, brain PET, 3 T-MRI, EEG source imaging, Wada test, and neuropsychological evaluation. We only included patients with TLE without evidence of HS (nTLE) on 3T-MRI images who were scheduled for epilepsy surgery. 2.2. MRI protocol Magnetic resonance imaging scanning was performed using a Philips Ingenia 3.0 Tesla MRI scanner (Philips Medical Systems, Best, The Netherlands) with a 16-channel head coil. The scans were performed using the non-contrast standard clinical seizure protocol to acquire coronal FLAIR, T1, and T2 weighted sequences to screen for possible underlying mesial temporal sclerosis or focal cortical dysplasias. Coronal FLAIR TE was 110, TR was 9000, TI was 2600, slice thickness was 3.0 mm, interslice gap 1.0 mm, matrix size 252 × 239, and the FOV = 200 mm. Coronal T2w turbo spin echo (TSE) sequence, echo time (TE) = 82, repetition time (TR) = 3325, TSE factor 21, single echo, slice thickness 2.5 mm without interslice gap, flip angle = 90° with 120° refocussing pulse, matrix size 400 × 301, and the FOV = 200 mm. For the coronal T1w 3D (TFE SENSE) sequence, TFE factor was 180, TE was 4.6, TR was 9.9, slice thickness was 0.8 mm without interslice gap, 0.8 mm isotropic resolution, matrix size 300 × 180, multishot single echo technique, flip angle = 8°, and the FOV = 240 mm Axial T1. Turbo spin echo T2w sequences as well as axial FLAIR images were also obtained. 2.3. Electroencephalographic evaluation and source analysis The scalp EEG was recorded with a 10–20 lead placement, using a Neuvo amplifier (Compumedics), which acquires raw data at 10 kHz and then applies a software second-order Infinite Impulse Response (IIR) Butterworth low-pass filter at 40% of the sampling frequency, recorded at a sampling rate of 500 Hz. Ictal patterns were reviewed using a common average montage and standard filter setting (HFF 70 Hz, LFF 1 Hz, notch filter 60 Hz). Analysis of EEG source of interictal and ictal activity was performed with Curry 7 software. The EEG source was obtained using a moving dipole model and SLORETA, co-registered with the patient's own MRI. 2.4. Intracranial electroencephalographic recording Patients who required intracranial monitoring for confirmation of the seizure-onset localization, as determined in our epilepsy surgery conference, were implanted with intracranial electrodes using subdural grids and/or strips covering the cortical areas that were suspected as potential epileptic zones based on non-invasive data. The seizure onset zone was determined by the presence of an electrographic seizure, which was defined as a sustained rhythmic change in the EEG background at a frequency of N 2 Hz, not explained by level of arousal or artifacts, and clearly distinguished from background EEG and interictal activity, and correlated with the patient's typical clinical behavior [6].
The intracranial EEG recording setting was similar to the scalp EEG with the exception of the use of a wider filter band of 1–200 Hz to visualize high-frequency oscillations (HFO) associated within the ictal zone. The HFO power map distribution was displayed onto an image of threedimensional (3D) reconstructed brain images using the patient's own co-registered MRI and computed tomography (CT) with the implanted electrodes (CURRY 7). Language mapping using cortical electrical stimulation through the implanted electrodes was performed in patients with language dominance ipsilateral to the epileptic zone (as determined by earlier WADA testing). 2.5. Surgery and electrocorticography (ECoG) After intracranial electroencephalographic (ECoG) analysis was performed and the anatomic and cortical epileptogenic zone was identified, patients were scheduled for surgical resection (in all cases an anterior temporal lobectomy). The extent of the resection was planned pre-operatively based on localization of the ictal zones according to the available functional and imaging data. First, the lateral neocortex of the temporal lobe was resected, the temporal horn of the lateral ventricle was entered, and the hippocampus and amygdala were identified. Before any resection of the mesial temporal lobe structures, all patients underwent electrocorticography analysis using depth electrode placement under direct visualization and stereotactic placement in the amygdala, anterior, middle and posterior hippocampus, to assess the presence of ictal and interictal epileptic activity. The patients were anesthetized with sevoflurane during the recording at the lowest level of anesthesia possible to allow epileptic activity but avoiding movement or consciousness. After ECoG data were recorded for at least 20 min or until a seizure pattern was observed, the patient underwent hippocampal resection as posterior as the tectal plate. 2.6. Postoperative follow-up The presence of seizure recurrence was closely monitored in both neurology and neurosurgery outpatient clinics, and surgical success was evaluated using Engel's classification. All antiepileptic drugs (AEDs) were continued in the post-operative period. 2.7. Histopathological analysis procedures Each surgical specimen was comprised of multiple sections of the amygdala, hippocampus, and parahippocampal cortex which roughly correlated to the intra-operative depth electrode placement and recordings. The division of the tissue by sectors around the depth electrode recording areas was exclusive for the study patients. The specimens were fixed with 10% neutral buffered formalin. They received routine processing by our university pathology lab and then were sent for analysis to a well-known outside neuropathologist. All the histopathology samples without evidence of mesial temporal sclerosis were sent to an independent neuropathologist with extensive experience in analyzing epilepsy specimens since there is poor reliability for diagnosis of mild FCD among neuropathologists with low experience in epilepsy [7,8]. A typical sample slide of each case was sent to a third independent neuropathologist for review. Neither outside neuropathologist had access to clinical data — including MRI, electroencephalographic findings, or surgical outcome. 3. Results 3.1. Demographic characteristics The study included five consecutive patients with TLE without evidence of mesial temporal lesions on MRI. Only subjects scheduled for epilepsy surgery between April and September 2015 were included in the study. Three patients with TLE operated within this period were
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
19
Table 1 Demographics, epilepsy history, and neuropsychology. #
Age
Sex
Epilepsy duration (years)
Seizure onset (age)
Seizures per month
Neuropsychology
WADA
1
63
F
29
34
30
Neuropsychology: slightly more efficient dominant side without clear focal deficit.
2
39
F
9
30
5
Neuropsychology: not performed
3
32
F
13
19
4
Neuropsychology: global impairment with stronger verbal versus visual memory
4
60
M
49
11
2
Neuropsychology: memory impairment without clear lateralization
5
46
M
9
37
6
Neuropsychology: memory impairment with stronger functioning of non-dominant vs dominant side
Language: Left Memory (L) 5/10 (R) 6/10 Language: Left Memory (L) 9/10 (R) 4/10 Language: Left Memory (L) 8/10 (R) 5/10 Language: Left Memory (L) 3/10 (R) 4/10 Language: Left Memory (L) 1/10 (R) 7/10
excluded because of evidence of mesial sclerosis on MRI. Among the included patients, three were females and two were males. The age range was 32 to 63 years old (mean age 48.0 years) at the time of the first evaluation. Most patients had onset of seizures during adulthood except one patient in whom seizures started at age 11. There was no clear precipitating event, no head injury, and no family history of epilepsy in this cohort. Most patients had a long history of seizures with a mean duration of 21.8 years, ranging from 9 to 49 years. Seizure frequency ranged from daily to twice a month. Most patients had tried more than three AEDs at the time of evaluation without good control of seizures (medically intractable) (see Table 1).
3.2. Seizure semiology Clinical features observed in the five patients are summarized in the Table 2. Three patients exhibited auras that consisted of olfactory sensations, epigastric burning, subjective feelings of confusion, or being present in a strange place. One patient exhibited visual auras of flashing lights. Automatisms were present only in three out of five patients, most commonly lip smacking and chewing movements. Hand automatisms were observed only in one patient. Verbal automatisms consisted of repetitive verbalization in one patient. Dialepsis was observed uniformly among all five patients. Motor behavior was contralateral in 80% of the patients. Secondary seizure generalization was noted in three patients. Four patients exhibited contralateral
PET
Right temporal hypometabolism
Subtle right mesial temporal hypometabolism
Left temporal hypometabolism
head deviation whereas one patient had ipsilateral head deviation (see Table 2). 3.3. Video-electroencephalographic studies and source analysis All patients underwent scalp video-EEG monitoring for at least 3 days in our epilepsy monitoring unit. During this time, AEDs were completely withdrawn (see Table 3). 3.4. Pathology results All postsurgical amgydalar, hippocampal, and parahippocampal tissue samples were initially sent to a clinical neuropathologist. The pathology report in all patients showed no evidence of microscopic abnormality in the amygdala, hippocampus, or parahippocampal cortex. In the light of no histological abnormalities to explain the etiology of TLE, the protocol was modified to send all specimens from the enrolled patients to an independent neuropathologist with exposure to large number of epileptic temporal lobe specimens. The histological analysis reported significant microscopic abnormalities in all subjects. Interestingly these abnormalities were present, not in the hippocampus, but in the parahippocampal gyrus. Concordant with the MRI findings, none of the five subjects had evidence of hippocampal sclerosis. Patient two had mild focal perivascular chronic inflammatory changes and gliosis. All the subjects had microscopic abnormalities in layer two of the parahippocampal
Table 2 Seizure semiology and localization. Patient
Aura
Automatisms
Dialepsis
Motor
Other
Seizure laterality
1
Olfactory: rotten eggs
No
Yes
Right arm shaking, right head deviation
Left temporal
2
No
Yes
Abdominal: epigastric burning
Behavior arrest
Right temporal
4
No
Left head deviation Secondary generalization Left head and eye deviation Left leg shaking Secondary generalization Left hand shaking and holding
Right temporal
3
Verbal Chewing Lip smacking Chewing Hand No
Blank staring Unable to speak Behavior arrest
Left temporal
5
Psychic: feeling in a strange place Visual: flashing lights
Lip smacking Chewing
Blank staring Behavior arrest Blank staring
Yes
Yes Yes
Right head and eye deviation Secondary generalization
Left temporal
20
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
Table 3 Scalp and intracranial EEG findings. #
Focal slowing
Interictal spike maxima
Ictal EEG onset
Scalp EEG source analysis
Intracranial EEG
1
Left anterior temporal Right temporal
T1
Rhythmic theta at T1 Right temporal rhythmic theta
Ictal source involving the left anterior lateral cortex and the mesial cortex Right anterior temporal
Right temporal rhythmic theta
Right anterior temporal
Left temporal rhythmic delta Left temporal rhythmic theta
Left inferior temporal
IC Video EEG: ictal HFO and spike onset in the left parahippocampal region. Hippocampal ECoG: brief anterior hippocampal seizure IC Video EEG: none Hippocampal ECoG: right middle and anterior hippocampal fast repetitive spike bursts lasting about 3 s. IC Video EEG: none Hippocampal ECoG: frequent amygdala spikes and seizure onset in the anterior and middle hippocampus with rapid spread to the entire hippocampus IC Video EEG: ictal HFO in the left parahippocampal region Hippocampal ECoG: left anterior hippocampal seizures lasting about 40 s IC Video EEG: ictal HFO and spike onset in the left parahippocampal region Hippocampal ECoG: left amygdala spikes, no seizures
2
3
None
F8- T2 Less frequent F7-T1 F8-T6
4
None
T3-T5
5
None
F7-T1
Left mesial temporal
Fig. 1. H&E stained surgical specimens of the parahippocampal gyrus depicting subtle neuronal changes and comparing the diagnoses of two independent reviewers.
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
gyrus. Four of the five patients had changes consistent with had focal cortical dysplasia ILAE type Ib and one patient had focal cortical dysplasia ILAE Type Ic according to the second neuropathologist. The third reviewer concurred with the parahippocampal layer two abnormalities in 4 patients but he felt that these changes were not specific and not consistent with FCD (see Fig. 1). 3.5. Postoperative follow-up and outcome All patients were closely followed in neurology and neurosurgery clinics after surgery. No recurrence of seizures was reported in four out of five patients (Engel's class IA). The mean follow-up period was 13.2 months (11–15 months). Patient two reported ill-described nocturnal events, different from her previous seizures, after one-year seizure-free postoperatively (Engel class I). Repeated video EEG monitoring for three days with medication withdrawal failed to capture the reported event but demonstrated intermittent contralateral temporal spikes on EEG. 4. Discussion Our cases demonstrated the presence of mild focal cortical abnormalities in layer two of the parahippocampal gyrus but there was no concordance on the interpretation of these findings. Focal cortical dysplasia as defined by the ILAE [9] could not be definitely demonstrated in our patients with typical clinical characteristics of “non-lesional” TLE. None of these abnormalities were mentioned in the original pathology report, highlighting the difficulty of recognizing subtle cortical changes, particularly when the neuropathologist has low access to epilepsy surgery specimens [7,8,10]. The poor interrater reliability particularly for FCD type I has been recognized before [8], and the significance of the reported subtle changes is unknown. There were no clear uniform clinical features that could be correlated with parahippocampal cortical dysplasia. All of our patients had the typical clinical characteristics of TLE including prominent ipsilateral focal memory dysfunction, which was unilateral in three out five patients. There were some atypical features of TLE such as visual symptoms in one patient and frequent motor signs and secondary generalization that could be explained by abrupt AED withdrawal during the evaluation. Most of our patients had a typical scalp ictal EEG pattern of mesial temporal epilepsy with the exception of one patient who had focal delta slowing at the onset of the seizure, which has been associated with nTLE [11]. Electroencephalographic source imaging provided sublobar focal ictal localization consistent with the intracranial recording in two out of five patients; whereas EEG source imaging in the other three patients localized diffusely to the anterior temporal area. The ictal source localization was not clear enough to separate hippocampal vs parahippocampal onset. Focal HFO at the onset of the seizure in the parahippocampal region was found in all patients with intracranial EEG recording, consistent with previous reports of HFO activity between 80 and 120 Hz as a marker for the ictal zone [12–15]. Intra-operative hippocampal recording with depth electrodes captured spontaneous seizures in three patients and interictal spikes in all five patients. The spikes were likely induced by sevoflurane but not the seizures since the dose was decreased and seizures are not typically associated with this anesthetic agent [16]. Unfortunately, we did not record simultaneous parahippocampal and hippocampal electrical activity and we cannot determine if the origin of the seizure was the hippocampus or the parahippocampal gyrus. The normal hippocampal histopathology after an average duration of 21 years of epilepsy supports the localization of the epileptic zone in the parahippocampal gyrus with rapid propagation to the hippocampus, and from there to the anterior temporal network, giving an electrographic and clinical manifestation of mesial TLE. This theory explains why nTLE is often misdiagnosed as regular TLE.
21
5. Clinical significance 1. These findings support the need for more attention to subtle cortical abnormalities and training of neuropathologists to diagnose mild cortical dysplasia. 2. Our results also explain the previously documented higher failure rate of tailored amygdalo-hippocampectomy and a potential failure for laser ablation therapy due to lack of lateral cortical and parahippocampal resection [17]. 3. Although anterior temporal lobectomy is a well-documented surgical cure (70–90%) for well-selected candidates with TLE, it carries a large morbidity as there is frequent short-term memory dysfunction following surgery [18]. Use of less invasive techniques such as cortical neurostimulation in order to preserve an apparently normal hippocampus should be considered. 6. Conclusions Subtle cortical abnormalities should be suspected in patients with nTLE. Noninvasive ictal onset localization is very difficult. EEG source imaging provides fairly good localization but not enough to differentiate hippocampal vs parahippocampal onset. The improvement of noninvasive techniques may eventually allow more precise localization and less-invasive treatments. Finally, mild cortical abnormalities in the parahippocampal gyrus are not commonly reported in routine pathology analysis. Compliance with ethical standards Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent Our institutional IRB waived informed consent to observe these patients and analyze the tissue for pathological abnormalities associated with the clinical syndrome and electrophysiological findings. Conflict of interest The authors declare that they have no conflict of interest. References [1] Tassi L, Meroni A, Deleo F, Villani F, Mai R, Russo GL, et al. Temporal lobe epilepsy: neuropathological and clinical correlations in 243 surgically treated patients. Epileptic Disord 2009;11(4):281–92. [2] Bercovici E, Kumar BS, Mirsattari SM. Neocortical temporal lobe epilepsy. Epilepsy Res Treat 2012;2012:103160. [3] Pillay N, Fabinyi GCA, Myles TS, Fitt GJ, Berkovic SF, Jackson GD. Parahippocampal epilepsy with subtle dysplasia: a cause of ‘imaging negative’ partial epilepsy. Epilepsia 2009;50(12):2611–8. [4] Mirandola L, Badawy RA, Saunders AM, McIntosh A, Berkovic SF, Jackson GD. Clinical features of seizures associated with parahippocampal/inferior temporal lesions compared to those with hippocampal sclerosis. Epilepsia 2010;51(9):1906–9. [5] Suresh S, Sweet J, Fastenau PS, Lüders H, Landazuri P, Miller J. Temporal lobe epilepsy in patients with nonlesional MRI and normal memory: an SEEG study. J Neurosurg 2015:1–7. [6] Spencer SS, Guimaraes P, Katz A, Kim J, Spencer D. Morphological patterns of seizures recorded intracranially. Epilepsia 1992;33(3):537–45. [7] Coras R, de Boer OJ, Armstrong D, Becker A, Jacques TS, Miyata H, et al. Good interobserver and intraobserver agreement in the evaluation of the new ILAE classification of focal cortical dysplasias. Epilepsia 2012;53(8):1341–8. [8] Chamberlain WA, Cohen ML, Gyure KA, Kleinschmidt-DeMasters BK, Perry A, Powell SZ, et al. Interobserver and intraobserver reproducibility in focal cortical dysplasia (malformations of cortical development). Epilepsia 2009;50(12):2593–8. [9] Blümcke I, Thom M, Aronica E, Armstrong DD, Vinters HV, Palmini A, et al. The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification
22
[10] [11] [12] [13] [14]
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22 proposed by an ad hoc Task Force of the ILAE Diagnostic Methods commission 1. Epilepsia 2011;52(1):158–74. Thom M. Neuropathological findings in epilepsy. Curr Diagn Pathol 2004;10(2): 93–105. Gil-Nagel A, Risinger MW. Ictal semiology in hippocampal versus extrahippocampal temporal lobe epilepsy. Brain J Neurol 1997;120(Pt 1):183–92. Fisher RS, Webber WR, Lesser RP, Arroyo S, Uematsu S. High-frequency EEG activity at the start of seizures. J Clin Neurophysiol 1992;9(3):441–8. Zijlmans M, Jiruska P, Zelmann R, Leijten FSS, Jefferys JGR, Gotman J. High-frequency oscillations as a new biomarker in epilepsy. Ann Neurol 2012;71(2):169–78. Weiss SA, Lemesiou A, Connors R, Banks GP, McKhann GM, Goodman RR, et al. Seizure localization using ictal phase-locked high gamma: a retrospective surgical outcome study. Neurology 2015;84(23):2320–8.
[15] Ochi A, Otsubo H, Donner EJ, Elliott I, Iwata R, Funaki T, et al. Dynamic changes of ictal high-frequency oscillations in neocortical epilepsy: using multiple band frequency analysis. Epilepsia 2007;48(2):286–96. [16] Hisada K, Morioka T, Fukui K, Nishio S, Kuruma T, Irita K, et al. Effects of sevoflurane and isoflurane on electrocorticographic activities in patients with temporal lobe epilepsy. J Neurosurg Anesthesiol 2001;13(4):333–7. [17] Grote A, Witt J-A, Surges R, von Lehe M, Pieper M, Elger CE, et al. A second chancereoperation in patients with failed surgery for intractable epilepsy: long-term outcome, neuropsychology and complications. J Neurol Neurosurg Psychiatry 2015. [18] Helmstaedter C, Elger CE. Cognitive consequences of two-thirds anterior temporal lobectomy on verbal memory in 144 patients: a three-month follow-up study. Epilepsia 1996;37(2):171–80.
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Subtle pathological changes in neocortical temporal lobe epilepsy☆ Juan G. Ochoa a,⁎, Diana Hentgarden a, Audrey Paulzak c, Melissa Ogden a, Richard Pryson d, Markus Lamle e, Walter G. Rusyniak b a
Department of Neurology, University of South Alabama, United States Department of Neurosurgery, University of South Alabama, United States Department of Neurosurgery, University of Rochester, United States d Department of Pathology, Cleveland Clinic Foundation, United States e Department of Radiology, University of South Alabama, United States b c
a r t i c l e
i n f o
Article history: Received 20 July 2016 Revised 7 December 2016 Accepted 7 January 2017 Available online xxxx Keywords: TLE Epilepsy surgery Parahippocampus FCD Neocortical temporal lobe epilepsy HFO
a b s t r a c t This was a prospective observational study to correlate the clinical symptoms, electrophysiology, imaging, and surgical pathology of patients with temporal lobe epilepsy (TLE) without hippocampal sclerosis. We selected consecutive patients with TLE and normal MRI undergoing temporal lobe resection between April and September 2015. Clinical features, imaging, and functional data were reviewed. Intracranial monitoring and language mapping were performed when it was required according to our team recommendation. Prior to hippocampal resection, intraoperative electrocorticography was performed using depth electrodes in the amygdala and the hippocampus. The resected hippocampus was sent for pathological analysis. Results: Five patients with diagnosis with non-lesional TLE were included. We did not find distinctive clinical features that could be a characteristic of non-lesional TLE. The mean follow-up was 13.2 months (11–15 months); 80% of patients achieved Engel Class I outcome. There was no distinctive electrographic findings in these patients. Histopathologic analysis was negative for mesial temporal sclerosis. A second blinded independent neuropathologist with expertise in epilepsy found ILAE type I focal cortical dysplasia in the parahippocampal gyrus in all patients. A third independent neuropathologist reported changes in layer 2 with larger pyramidal neurons in 4 cases but concluded that none of these cases met the diagnostic criteria of FCD. Subtle pathological changes could be associated with a parahippocampal epileptic zone and should be investigated in patients with MRI-negative TLE. This study also highlights the lack of interobserver reliability for the diagnosis of mild cortical dysplasia. Finally, selective amygdalo-hippocampectomy or laser ablation of the hippocampus may not control intractable epilepsy in this specific population. © 2017 Elsevier Inc. All rights reserved.
1. Introduction Temporal lobe epilepsy (TLE) is the most common form of epilepsy of focal origin. It is commonly associated with hippocampal sclerosis (HS) and other pathologies such as tumors and malformations of cortical development (MCD). In a study of 243 patients with TLE who underwent surgery published in 2009, only 5% had negative pathology [1]. These patients may have poor localization of their seizure focus and may be considered nonsurgical candidates. Lesional neocortical TLE (nTLE) cases are often not reported in the literature because they may be less likely to be admitted for video-EEG monitoring [2] Nonetheless, nTLE has started has started to be a recognized as a different entity from mesial TLE, in which histopathological analysis
☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ⁎ Corresponding author at: Department of Neurology, 2451 Fillingim St, 10-E, Mobile, AL 36617-2293, United States. E-mail address: [email protected] (J.G. Ochoa).
http://dx.doi.org/10.1016/j.yebeh.2017.01.001 1525-5050/© 2017 Elsevier Inc. All rights reserved.
reveals low grade or no sclerosis of hippocampal resections. In a small case series published in 2009, MCD was found in the parahippocampal gyrus of patients with only subtle but distinctive abnormalities on MRI [3]. In MRI-negative patients, the localization is more challenging, usually requiring intracranial monitoring for localization of the epileptogenic lesion. From the clinical presentation, including semiology and scalp EEG, TLE with HS and nTLE are difficult to differentiate. A study in 2010 compared the clinical features of patients with parahippocampal inferior temporal lesions and HS, finding that hypermotor and bilateral motor symptoms were more common in patients with lesions in the posterior parahippocampal gyrus group compared to patients with HS [4]. Another study reported typical TLE seizure semiology with preservation of memory and normal MRI; interestingly, hippocampal pathology reported in few of these patients was normal and MCD was found in the lateral temporal cortex [5]. Although TLE with HS is a thoroughly studied entity, the incidence, characterization, etiology, and pathophysiology of nTLE is not well documented in the literature. This study aimed to further investigate and analyze nTLE, including clinical symptoms, scalp EEG, intra-operative EEG,
18
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
high-resolution MRI, surgical histopathology, and surgical outcome of patients with medically-intractable nTLE who underwent anterior temporal lobectomy. 2. Materials and methods 2.1. Subjects and patient selection We selected consecutive patients evaluated at the University of South Alabama Comprehensive Epilepsy Program between April and September 2015. We targeted patients with medically-intractable epilepsy localized to either left or right temporal lobe by noninvasive methods and normal hippocampus on 3-T MRI, using T1, FLAIR, and T2 sequences with thin coronal cuts through the hippocampus. Diagnosis of TLE was reached through the normal epilepsy workup, including: seizure semiology, scalp EEG, brain PET, 3 T-MRI, EEG source imaging, Wada test, and neuropsychological evaluation. We only included patients with TLE without evidence of HS (nTLE) on 3T-MRI images who were scheduled for epilepsy surgery. 2.2. MRI protocol Magnetic resonance imaging scanning was performed using a Philips Ingenia 3.0 Tesla MRI scanner (Philips Medical Systems, Best, The Netherlands) with a 16-channel head coil. The scans were performed using the non-contrast standard clinical seizure protocol to acquire coronal FLAIR, T1, and T2 weighted sequences to screen for possible underlying mesial temporal sclerosis or focal cortical dysplasias. Coronal FLAIR TE was 110, TR was 9000, TI was 2600, slice thickness was 3.0 mm, interslice gap 1.0 mm, matrix size 252 × 239, and the FOV = 200 mm. Coronal T2w turbo spin echo (TSE) sequence, echo time (TE) = 82, repetition time (TR) = 3325, TSE factor 21, single echo, slice thickness 2.5 mm without interslice gap, flip angle = 90° with 120° refocussing pulse, matrix size 400 × 301, and the FOV = 200 mm. For the coronal T1w 3D (TFE SENSE) sequence, TFE factor was 180, TE was 4.6, TR was 9.9, slice thickness was 0.8 mm without interslice gap, 0.8 mm isotropic resolution, matrix size 300 × 180, multishot single echo technique, flip angle = 8°, and the FOV = 240 mm Axial T1. Turbo spin echo T2w sequences as well as axial FLAIR images were also obtained. 2.3. Electroencephalographic evaluation and source analysis The scalp EEG was recorded with a 10–20 lead placement, using a Neuvo amplifier (Compumedics), which acquires raw data at 10 kHz and then applies a software second-order Infinite Impulse Response (IIR) Butterworth low-pass filter at 40% of the sampling frequency, recorded at a sampling rate of 500 Hz. Ictal patterns were reviewed using a common average montage and standard filter setting (HFF 70 Hz, LFF 1 Hz, notch filter 60 Hz). Analysis of EEG source of interictal and ictal activity was performed with Curry 7 software. The EEG source was obtained using a moving dipole model and SLORETA, co-registered with the patient's own MRI. 2.4. Intracranial electroencephalographic recording Patients who required intracranial monitoring for confirmation of the seizure-onset localization, as determined in our epilepsy surgery conference, were implanted with intracranial electrodes using subdural grids and/or strips covering the cortical areas that were suspected as potential epileptic zones based on non-invasive data. The seizure onset zone was determined by the presence of an electrographic seizure, which was defined as a sustained rhythmic change in the EEG background at a frequency of N 2 Hz, not explained by level of arousal or artifacts, and clearly distinguished from background EEG and interictal activity, and correlated with the patient's typical clinical behavior [6].
The intracranial EEG recording setting was similar to the scalp EEG with the exception of the use of a wider filter band of 1–200 Hz to visualize high-frequency oscillations (HFO) associated within the ictal zone. The HFO power map distribution was displayed onto an image of threedimensional (3D) reconstructed brain images using the patient's own co-registered MRI and computed tomography (CT) with the implanted electrodes (CURRY 7). Language mapping using cortical electrical stimulation through the implanted electrodes was performed in patients with language dominance ipsilateral to the epileptic zone (as determined by earlier WADA testing). 2.5. Surgery and electrocorticography (ECoG) After intracranial electroencephalographic (ECoG) analysis was performed and the anatomic and cortical epileptogenic zone was identified, patients were scheduled for surgical resection (in all cases an anterior temporal lobectomy). The extent of the resection was planned pre-operatively based on localization of the ictal zones according to the available functional and imaging data. First, the lateral neocortex of the temporal lobe was resected, the temporal horn of the lateral ventricle was entered, and the hippocampus and amygdala were identified. Before any resection of the mesial temporal lobe structures, all patients underwent electrocorticography analysis using depth electrode placement under direct visualization and stereotactic placement in the amygdala, anterior, middle and posterior hippocampus, to assess the presence of ictal and interictal epileptic activity. The patients were anesthetized with sevoflurane during the recording at the lowest level of anesthesia possible to allow epileptic activity but avoiding movement or consciousness. After ECoG data were recorded for at least 20 min or until a seizure pattern was observed, the patient underwent hippocampal resection as posterior as the tectal plate. 2.6. Postoperative follow-up The presence of seizure recurrence was closely monitored in both neurology and neurosurgery outpatient clinics, and surgical success was evaluated using Engel's classification. All antiepileptic drugs (AEDs) were continued in the post-operative period. 2.7. Histopathological analysis procedures Each surgical specimen was comprised of multiple sections of the amygdala, hippocampus, and parahippocampal cortex which roughly correlated to the intra-operative depth electrode placement and recordings. The division of the tissue by sectors around the depth electrode recording areas was exclusive for the study patients. The specimens were fixed with 10% neutral buffered formalin. They received routine processing by our university pathology lab and then were sent for analysis to a well-known outside neuropathologist. All the histopathology samples without evidence of mesial temporal sclerosis were sent to an independent neuropathologist with extensive experience in analyzing epilepsy specimens since there is poor reliability for diagnosis of mild FCD among neuropathologists with low experience in epilepsy [7,8]. A typical sample slide of each case was sent to a third independent neuropathologist for review. Neither outside neuropathologist had access to clinical data — including MRI, electroencephalographic findings, or surgical outcome. 3. Results 3.1. Demographic characteristics The study included five consecutive patients with TLE without evidence of mesial temporal lesions on MRI. Only subjects scheduled for epilepsy surgery between April and September 2015 were included in the study. Three patients with TLE operated within this period were
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
19
Table 1 Demographics, epilepsy history, and neuropsychology. #
Age
Sex
Epilepsy duration (years)
Seizure onset (age)
Seizures per month
Neuropsychology
WADA
1
63
F
29
34
30
Neuropsychology: slightly more efficient dominant side without clear focal deficit.
2
39
F
9
30
5
Neuropsychology: not performed
3
32
F
13
19
4
Neuropsychology: global impairment with stronger verbal versus visual memory
4
60
M
49
11
2
Neuropsychology: memory impairment without clear lateralization
5
46
M
9
37
6
Neuropsychology: memory impairment with stronger functioning of non-dominant vs dominant side
Language: Left Memory (L) 5/10 (R) 6/10 Language: Left Memory (L) 9/10 (R) 4/10 Language: Left Memory (L) 8/10 (R) 5/10 Language: Left Memory (L) 3/10 (R) 4/10 Language: Left Memory (L) 1/10 (R) 7/10
excluded because of evidence of mesial sclerosis on MRI. Among the included patients, three were females and two were males. The age range was 32 to 63 years old (mean age 48.0 years) at the time of the first evaluation. Most patients had onset of seizures during adulthood except one patient in whom seizures started at age 11. There was no clear precipitating event, no head injury, and no family history of epilepsy in this cohort. Most patients had a long history of seizures with a mean duration of 21.8 years, ranging from 9 to 49 years. Seizure frequency ranged from daily to twice a month. Most patients had tried more than three AEDs at the time of evaluation without good control of seizures (medically intractable) (see Table 1).
3.2. Seizure semiology Clinical features observed in the five patients are summarized in the Table 2. Three patients exhibited auras that consisted of olfactory sensations, epigastric burning, subjective feelings of confusion, or being present in a strange place. One patient exhibited visual auras of flashing lights. Automatisms were present only in three out of five patients, most commonly lip smacking and chewing movements. Hand automatisms were observed only in one patient. Verbal automatisms consisted of repetitive verbalization in one patient. Dialepsis was observed uniformly among all five patients. Motor behavior was contralateral in 80% of the patients. Secondary seizure generalization was noted in three patients. Four patients exhibited contralateral
PET
Right temporal hypometabolism
Subtle right mesial temporal hypometabolism
Left temporal hypometabolism
head deviation whereas one patient had ipsilateral head deviation (see Table 2). 3.3. Video-electroencephalographic studies and source analysis All patients underwent scalp video-EEG monitoring for at least 3 days in our epilepsy monitoring unit. During this time, AEDs were completely withdrawn (see Table 3). 3.4. Pathology results All postsurgical amgydalar, hippocampal, and parahippocampal tissue samples were initially sent to a clinical neuropathologist. The pathology report in all patients showed no evidence of microscopic abnormality in the amygdala, hippocampus, or parahippocampal cortex. In the light of no histological abnormalities to explain the etiology of TLE, the protocol was modified to send all specimens from the enrolled patients to an independent neuropathologist with exposure to large number of epileptic temporal lobe specimens. The histological analysis reported significant microscopic abnormalities in all subjects. Interestingly these abnormalities were present, not in the hippocampus, but in the parahippocampal gyrus. Concordant with the MRI findings, none of the five subjects had evidence of hippocampal sclerosis. Patient two had mild focal perivascular chronic inflammatory changes and gliosis. All the subjects had microscopic abnormalities in layer two of the parahippocampal
Table 2 Seizure semiology and localization. Patient
Aura
Automatisms
Dialepsis
Motor
Other
Seizure laterality
1
Olfactory: rotten eggs
No
Yes
Right arm shaking, right head deviation
Left temporal
2
No
Yes
Abdominal: epigastric burning
Behavior arrest
Right temporal
4
No
Left head deviation Secondary generalization Left head and eye deviation Left leg shaking Secondary generalization Left hand shaking and holding
Right temporal
3
Verbal Chewing Lip smacking Chewing Hand No
Blank staring Unable to speak Behavior arrest
Left temporal
5
Psychic: feeling in a strange place Visual: flashing lights
Lip smacking Chewing
Blank staring Behavior arrest Blank staring
Yes
Yes Yes
Right head and eye deviation Secondary generalization
Left temporal
20
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
Table 3 Scalp and intracranial EEG findings. #
Focal slowing
Interictal spike maxima
Ictal EEG onset
Scalp EEG source analysis
Intracranial EEG
1
Left anterior temporal Right temporal
T1
Rhythmic theta at T1 Right temporal rhythmic theta
Ictal source involving the left anterior lateral cortex and the mesial cortex Right anterior temporal
Right temporal rhythmic theta
Right anterior temporal
Left temporal rhythmic delta Left temporal rhythmic theta
Left inferior temporal
IC Video EEG: ictal HFO and spike onset in the left parahippocampal region. Hippocampal ECoG: brief anterior hippocampal seizure IC Video EEG: none Hippocampal ECoG: right middle and anterior hippocampal fast repetitive spike bursts lasting about 3 s. IC Video EEG: none Hippocampal ECoG: frequent amygdala spikes and seizure onset in the anterior and middle hippocampus with rapid spread to the entire hippocampus IC Video EEG: ictal HFO in the left parahippocampal region Hippocampal ECoG: left anterior hippocampal seizures lasting about 40 s IC Video EEG: ictal HFO and spike onset in the left parahippocampal region Hippocampal ECoG: left amygdala spikes, no seizures
2
3
None
F8- T2 Less frequent F7-T1 F8-T6
4
None
T3-T5
5
None
F7-T1
Left mesial temporal
Fig. 1. H&E stained surgical specimens of the parahippocampal gyrus depicting subtle neuronal changes and comparing the diagnoses of two independent reviewers.
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22
gyrus. Four of the five patients had changes consistent with had focal cortical dysplasia ILAE type Ib and one patient had focal cortical dysplasia ILAE Type Ic according to the second neuropathologist. The third reviewer concurred with the parahippocampal layer two abnormalities in 4 patients but he felt that these changes were not specific and not consistent with FCD (see Fig. 1). 3.5. Postoperative follow-up and outcome All patients were closely followed in neurology and neurosurgery clinics after surgery. No recurrence of seizures was reported in four out of five patients (Engel's class IA). The mean follow-up period was 13.2 months (11–15 months). Patient two reported ill-described nocturnal events, different from her previous seizures, after one-year seizure-free postoperatively (Engel class I). Repeated video EEG monitoring for three days with medication withdrawal failed to capture the reported event but demonstrated intermittent contralateral temporal spikes on EEG. 4. Discussion Our cases demonstrated the presence of mild focal cortical abnormalities in layer two of the parahippocampal gyrus but there was no concordance on the interpretation of these findings. Focal cortical dysplasia as defined by the ILAE [9] could not be definitely demonstrated in our patients with typical clinical characteristics of “non-lesional” TLE. None of these abnormalities were mentioned in the original pathology report, highlighting the difficulty of recognizing subtle cortical changes, particularly when the neuropathologist has low access to epilepsy surgery specimens [7,8,10]. The poor interrater reliability particularly for FCD type I has been recognized before [8], and the significance of the reported subtle changes is unknown. There were no clear uniform clinical features that could be correlated with parahippocampal cortical dysplasia. All of our patients had the typical clinical characteristics of TLE including prominent ipsilateral focal memory dysfunction, which was unilateral in three out five patients. There were some atypical features of TLE such as visual symptoms in one patient and frequent motor signs and secondary generalization that could be explained by abrupt AED withdrawal during the evaluation. Most of our patients had a typical scalp ictal EEG pattern of mesial temporal epilepsy with the exception of one patient who had focal delta slowing at the onset of the seizure, which has been associated with nTLE [11]. Electroencephalographic source imaging provided sublobar focal ictal localization consistent with the intracranial recording in two out of five patients; whereas EEG source imaging in the other three patients localized diffusely to the anterior temporal area. The ictal source localization was not clear enough to separate hippocampal vs parahippocampal onset. Focal HFO at the onset of the seizure in the parahippocampal region was found in all patients with intracranial EEG recording, consistent with previous reports of HFO activity between 80 and 120 Hz as a marker for the ictal zone [12–15]. Intra-operative hippocampal recording with depth electrodes captured spontaneous seizures in three patients and interictal spikes in all five patients. The spikes were likely induced by sevoflurane but not the seizures since the dose was decreased and seizures are not typically associated with this anesthetic agent [16]. Unfortunately, we did not record simultaneous parahippocampal and hippocampal electrical activity and we cannot determine if the origin of the seizure was the hippocampus or the parahippocampal gyrus. The normal hippocampal histopathology after an average duration of 21 years of epilepsy supports the localization of the epileptic zone in the parahippocampal gyrus with rapid propagation to the hippocampus, and from there to the anterior temporal network, giving an electrographic and clinical manifestation of mesial TLE. This theory explains why nTLE is often misdiagnosed as regular TLE.
21
5. Clinical significance 1. These findings support the need for more attention to subtle cortical abnormalities and training of neuropathologists to diagnose mild cortical dysplasia. 2. Our results also explain the previously documented higher failure rate of tailored amygdalo-hippocampectomy and a potential failure for laser ablation therapy due to lack of lateral cortical and parahippocampal resection [17]. 3. Although anterior temporal lobectomy is a well-documented surgical cure (70–90%) for well-selected candidates with TLE, it carries a large morbidity as there is frequent short-term memory dysfunction following surgery [18]. Use of less invasive techniques such as cortical neurostimulation in order to preserve an apparently normal hippocampus should be considered. 6. Conclusions Subtle cortical abnormalities should be suspected in patients with nTLE. Noninvasive ictal onset localization is very difficult. EEG source imaging provides fairly good localization but not enough to differentiate hippocampal vs parahippocampal onset. The improvement of noninvasive techniques may eventually allow more precise localization and less-invasive treatments. Finally, mild cortical abnormalities in the parahippocampal gyrus are not commonly reported in routine pathology analysis. Compliance with ethical standards Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent Our institutional IRB waived informed consent to observe these patients and analyze the tissue for pathological abnormalities associated with the clinical syndrome and electrophysiological findings. Conflict of interest The authors declare that they have no conflict of interest. References [1] Tassi L, Meroni A, Deleo F, Villani F, Mai R, Russo GL, et al. Temporal lobe epilepsy: neuropathological and clinical correlations in 243 surgically treated patients. Epileptic Disord 2009;11(4):281–92. [2] Bercovici E, Kumar BS, Mirsattari SM. Neocortical temporal lobe epilepsy. Epilepsy Res Treat 2012;2012:103160. [3] Pillay N, Fabinyi GCA, Myles TS, Fitt GJ, Berkovic SF, Jackson GD. Parahippocampal epilepsy with subtle dysplasia: a cause of ‘imaging negative’ partial epilepsy. Epilepsia 2009;50(12):2611–8. [4] Mirandola L, Badawy RA, Saunders AM, McIntosh A, Berkovic SF, Jackson GD. Clinical features of seizures associated with parahippocampal/inferior temporal lesions compared to those with hippocampal sclerosis. Epilepsia 2010;51(9):1906–9. [5] Suresh S, Sweet J, Fastenau PS, Lüders H, Landazuri P, Miller J. Temporal lobe epilepsy in patients with nonlesional MRI and normal memory: an SEEG study. J Neurosurg 2015:1–7. [6] Spencer SS, Guimaraes P, Katz A, Kim J, Spencer D. Morphological patterns of seizures recorded intracranially. Epilepsia 1992;33(3):537–45. [7] Coras R, de Boer OJ, Armstrong D, Becker A, Jacques TS, Miyata H, et al. Good interobserver and intraobserver agreement in the evaluation of the new ILAE classification of focal cortical dysplasias. Epilepsia 2012;53(8):1341–8. [8] Chamberlain WA, Cohen ML, Gyure KA, Kleinschmidt-DeMasters BK, Perry A, Powell SZ, et al. Interobserver and intraobserver reproducibility in focal cortical dysplasia (malformations of cortical development). Epilepsia 2009;50(12):2593–8. [9] Blümcke I, Thom M, Aronica E, Armstrong DD, Vinters HV, Palmini A, et al. The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification
22
[10] [11] [12] [13] [14]
J.G. Ochoa et al. / Epilepsy & Behavior 71 (2017) 17–22 proposed by an ad hoc Task Force of the ILAE Diagnostic Methods commission 1. Epilepsia 2011;52(1):158–74. Thom M. Neuropathological findings in epilepsy. Curr Diagn Pathol 2004;10(2): 93–105. Gil-Nagel A, Risinger MW. Ictal semiology in hippocampal versus extrahippocampal temporal lobe epilepsy. Brain J Neurol 1997;120(Pt 1):183–92. Fisher RS, Webber WR, Lesser RP, Arroyo S, Uematsu S. High-frequency EEG activity at the start of seizures. J Clin Neurophysiol 1992;9(3):441–8. Zijlmans M, Jiruska P, Zelmann R, Leijten FSS, Jefferys JGR, Gotman J. High-frequency oscillations as a new biomarker in epilepsy. Ann Neurol 2012;71(2):169–78. Weiss SA, Lemesiou A, Connors R, Banks GP, McKhann GM, Goodman RR, et al. Seizure localization using ictal phase-locked high gamma: a retrospective surgical outcome study. Neurology 2015;84(23):2320–8.
[15] Ochi A, Otsubo H, Donner EJ, Elliott I, Iwata R, Funaki T, et al. Dynamic changes of ictal high-frequency oscillations in neocortical epilepsy: using multiple band frequency analysis. Epilepsia 2007;48(2):286–96. [16] Hisada K, Morioka T, Fukui K, Nishio S, Kuruma T, Irita K, et al. Effects of sevoflurane and isoflurane on electrocorticographic activities in patients with temporal lobe epilepsy. J Neurosurg Anesthesiol 2001;13(4):333–7. [17] Grote A, Witt J-A, Surges R, von Lehe M, Pieper M, Elger CE, et al. A second chancereoperation in patients with failed surgery for intractable epilepsy: long-term outcome, neuropsychology and complications. J Neurol Neurosurg Psychiatry 2015. [18] Helmstaedter C, Elger CE. Cognitive consequences of two-thirds anterior temporal lobectomy on verbal memory in 144 patients: a three-month follow-up study. Epilepsia 1996;37(2):171–80.