- Email: [email protected]
Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy
YEBEH-05210; No of Pages 6 Epilepsy & Behavior xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy Linglin Yang a,1, Hong Li b,1, Lujia Zhu a, Xinfeng Yu b, Bo Jin a, Cong Chen a, Shan Wang a, Meiping Ding a, Minming Zhang b, Zhong Chen a, Shuang Wang a,⁎ a b
Epilepsy Center, Department of Neurology, Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China Departments of Radiology, Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China
a r t i c l e
i n f o
Article history: Received 16 November 2016 Revised 6 February 2017 Accepted 6 February 2017 Available online xxxx Keywords: Secondarily generalized tonic–clonic seizure Temporal lobe epilepsy Hippocampal sclerosis Shape analysis Medial thalamus Ventral pallidum
a b s t r a c t Mesial temporal lobe epilepsy (mTLE) is a common type of drug-resistant epilepsy and secondarily generalized tonic–clonic seizures (sGTCS) have devastating consequences for patients' safety and quality of life. To probe the mechanism underlying the genesis of sGTCS, we investigated the structural differences between patients with and without sGTCS in a cohort of mTLE with radiologically defined unilateral hippocampal sclerosis. We performed voxel-based morphometric analysis of cortex and vertex-wise shape analysis of subcortical structures (the basal ganglia and thalamus) on MRI of 39 patients (21 with and 18 without sGTCS). Comparisons were initially made between sGTCS and non-sGTCS groups, and subsequently made between uncontrolled-sGTCS and controlled-sGTCS subgroups. Regional atrophy of the ipsilateral ventral pallidum (cluster size = 450 voxels, corrected p = 0.047, Max voxel coordinate = 107, 120, 65), medial thalamus (cluster size = 1128 voxels, corrected p = 0.049, Max voxel coordinate = 107, 93, 67), middle frontal gyrus (cluster size = 60 voxels, corrected p b 0.05, Max voxel coordinate = −30, 49.5, 6), and contralateral posterior cingulate cortex (cluster size = 130 voxels, corrected p b 0.05, Max voxel coordinate = 16.5, −57, 27) was found in the sGTCS group relative to the non-sGTCS group. Furthermore, the uncontrolled-sGTCS subgroup showed more pronounced atrophy of the ipsilateral medial thalamus (cluster size = 1240 voxels, corrected p = 0.014, Max voxel coordinate = 107, 93, 67) than the controlled-sGTCS subgroup. These findings indicate a central role of thalamus and pallidum in the pathophysiology of sGTCS in mTLE. © 2017 Elsevier Inc. All rights reserved.
1. Introduction Mesial temporal lobe epilepsy (mTLE) is a common type of focal epilepsy, and is usually accompanied by hippocampal sclerosis. Seizures that occur in TLE with hippocampal sclerosis (TLE–HS) are typically resistant to antiepileptic drugs, which are associated with severe cognitive impairment, affective disorders, and stigmatization [1,2]. Among the seizure types associated with TLE–HS, secondarily generalized tonic– clonic seizures (sGTCS) are the most debilitating type and these in turn put the patients at additional risk of fatal injuries. Additionally, patients with TLE and sGTCS demonstrated worse surgical outcomes after anterior temporal lobectomy compared to those without sGTCS [3,4].
⁎ Corresponding author at: Epilepsy Center, Department of Neurology, Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310009, China. E-mail address: [email protected] (S. Wang). 1 These authors contributed equally.
Advancing the treatment strategies of sGTCS requires a detailed and clear understanding of the mechanism as to how the seizures are initiated, propagated, and terminated. However, the mechanism underlying the genesis of the sGTCS in mTLE has not been fully elucidated. Changes of cerebral blood flow (CBF) in the thalamus, basal ganglia, and frontoparietal cortex during the transition from partial seizures to secondarily generalization have been reported in patients with a mixed group of temporal and extra-temporal lobe epilepsy [5]. Some electrophysiological studies suggested that the basal ganglia exert an inhibitory effect against sGTCS in patients with mTLE [6–8]. Also, changes in the blood oxygen level-dependent (BOLD) signals were associated with generalized spike-wave (GSW) activity in the thalamo–cortical network. These were found in both primarily and secondarily generalized epilepsies, using simultaneous electroencephalogram and functional magnetic resonance imaging (EEG–fMRI) [9,10]. These lines of evidence suggest functional involvement of the basal ganglia-thalamo-cortical network in sGTCS. However, it remains unknown whether the structural alterations in these key nodes such as basal ganglia, thalamus, and cortex are involved in the pathophysiological mechanism of sGTCS.
http://dx.doi.org/10.1016/j.yebeh.2017.02.011 1525-5050/© 2017 Elsevier Inc. All rights reserved.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
2
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
Clinically, some patients with mTLE have sGTCS despite taking appropriate medication treatment, while some other patients have only partial seizures even though not on medications. This discrepancy in susceptibility to sGTCS might be associated with differences in the epileptogenic network, which may be reflected by morphometric analysis. In the present study, we investigated the relationship between cortical and subcortical gray matter (GM) alterations and the occurrence of sGTCS in a cohort of patients with unilateral TLE-HS. In particular, we initially applied sophisticated morphometric approaches to investigate the regional alterations in cortical volume and the surface shape of the thalamus and basal ganglia in patients with sGTCS versus those without sGTCS (non-sGTCS). Subsequently, we compared the structural alterations in patients with sGTCS that were not controlled by appropriate medical treatment (uncontrolled-sGTCS) versus those with medically controlled sGTCS (controlled-sGTCS).
2. Methods 2.1. Participants From April 2014 to June 2016, consecutive Chinese patients with unilateral TLE–HS, as confirmed by the clinical workup and imaging findings, were included in this study. Hippocampal sclerosis was defined if both hippocampal atrophy and increased T2 signals were observed on MRI [11]. All the patients were diagnosed by at least two experienced epileptologists based on history, seizure semiology, longterm scalp video-EEG monitoring, neuroimaging findings, and neuropsychological assessments. Some patients who required surgery were diagnosed as unilateral mTLE after stereo-EEG evaluation because their scalp EEG recordings did not provide adequate localizing or lateralizing information. Patients were excluded with: 1) history of brain trauma or surgery, 2) evidence of infectious origin, 3) evidence of secondary lesion that may be contributing to seizures, 4) serious psychiatric disorders, 5) ambiguous history of sGTCS, 6) seizures arising from the temporal lobe which is contralateral to the sclerotic hippocampus, or 7) bilateral mTLE. A typical sGTCS is characterized by bilateral rigid tonic extremities followed by rhythmic clonic jerks, along with complete loss of consciousness [12]. The seizure type was identified based on the seizure video, and the average seizure frequency was calculated for each patient according to the seizure diaries or family reports. Patients with sGTCS were further divided into two subgroups: 1) controlled-sGTCS subgroup, in which patients had secondarily generalized seizures only at the beginning of the disease course or after the medication was reduced, 2) uncontrolled-sGTCS subgroup, which included patients who had sGTCS events despite appropriate medical treatment with good compliance. Pharmacoresistance was defined as failure to achieve sustained freedom from seizures with adequate doses of two tolerated and appropriately chosen antiepileptic drugs [13]. The clinical variables were analyzed using two-sample independent t tests for continuous variables and Chi-square tests for categorical variables. The significance level was set at p b 0.05.
2.2. MRI protocol MR images were obtained on a 3.0T scanner (MR750, GE Healthcare, USA) with an 8-channel brain phased array coil. Highresolution coronal T2-weighted images perpendicular to the long axis of the hippocampus were acquired using spoiled gradient echo sequence with TR/TE = 5518/176 ms, flip angle = 110°, slice thickness = 2 mm, matrix = 512 × 512. Sagittal 3D T1-weighted images were acquired using brain volume imaging (BRAVO) sequence with TR/TE = 8.2/3.2 ms, TI = 450 ms, flip angle = 12°, slice thickness = 1 mm, matrix = 256 × 256.
2.3. Imaging analysis It has been reported that hemisphere lateralization of the epileptogetic zone is irrelevant for sGTCS history in patients with TLE [3]. Therefore, images of patients with right mTLE were side-flipped, and brain structures were marked as ipsilateral and contralateral to the epileptogenic zone in the present study. 2.3.1. Shape analysis of the basal ganglia and thalamus Automated segmentation and vertex-wise shape analysis of the basal ganglia and thalamus were carried out using the FSL-integrated registration and segmentation toolbox (FSL-FIRST) software (v.5.0.0; http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/first) installed on a Mac workstation. Briefly, the shape models in the FIRST were constructed from a library of manually segmented images. FIRST automatically searches for the most probable shape illustrations when given the observed intensities from the input images. To normalize the inter-individual head size differences, the segmented subcortical regions including the caudate, putamen, pallidum, and thalamus were reconstructed in the Montreal Neurological Institute space. Comparisons were initially made between non-sGTCS and sGTCS groups, and subsequently made between controlled-sGTCS and uncontrolled-sGTCS subgroups. Group differences were investigated by computing a general linear model, and the model included group (sGTCS/non-sGTCS or controlled-sGTCS/uncontrolled-sGTCS) as an independent factor, and age, gender, and intracranial volume (ICV) as covariates. Permutation testing using “randomize” with 5000 Monte Carlo simulations was used to calculate statistics of the segmented structures, corrected for multiple comparisons by dividing the p-value of 0.05. Clusterwise extent correction was applied, with a threshold of F N 2.0. Results from the surface analysis were visualized using Freesurfer's Freeview software [14]. 2.3.2. Voxel-based morphometric analysis of cortex The structural T1WI data were processed using voxel-based morphometry 8 (VBM8, http://dbm.neuro.unijena.de/vbm.html) toolbox of Statistical Parametric Mapping 8 (SPM8, http://www.fil.ion.ucl.ac. uk/spm) software running on MATLAB R2009b. In the first step, VBM automatically segmented GM, white matter (WM), and cerebrospinal fluid (CSF) in each patient. The volumes of GM, WM, and CSF voxels were determined separately for each patient and summed to calculate ICV. After segmentation, all images were rigidly aligned. Then, a “DARTEL” template of GM was created by nonlinearly aligning the GM images to a common space. The native GM images were normalized to the “DARTEL” template by applying the individual flow fields of all scans, using modulation to compensate for volume changes because of compression and/or expansion. Finally, GM images were smoothed using a Gaussian filter with full-width at half-maximum of 6 mm. Images were visually inspected at each and every processing step. Whole brain voxel-wise two-sample t tests were used to analyze the differences of the GM volume between non-sGTCS and sGTCS groups, with age, gender, and ICV as covariates. The statistical threshold was corrected for multiple comparisons on the cluster levels, which was determined by Monte Carlo simulation (5000 iterations) using AlphaSim as implemented in the REST-toolbox [15]. A combination threshold of p b 0.0001 on the voxel level and a cluster size N46 mm3 voxels was considered significant, which corresponded to a corrected p b 0.05. Then, the voxel values from the significant cluster in group comparison were extracted (data not shown). Subsequently, comparison of voxel values was made between controlled-sGTCS and uncontrolled-sGTCS, using multivariate analysis of covariance (MANCOVA) on age, gender, and ICV. To estimate the impairing effect of epilepsy, partial correlation coefficients were computed between the voxel values and epilepsy duration or frequency of partial seizures, controlling for age, gender, and ICV. To estimate the impairing effect of sGTCS, comparison of voxel values were made
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx Table 1 Demographic and clinical parameters in TLE-HS patients. Variable
sGTCS(n = 21)
Non-sGTCS(n = 18)
Sig.
Gender, M/F Age at scan (years) Age at onset (years) Epilepsy duration (years) Frequency of partial seizures (number per month) Lateralization of HS, L/R Febrile convulsions Pharmacoresistance
9/12 32.1 ± 10.3 17.5 ± 7.6 14.7 ± 9.5 2.6 ± 3.2
6/12 30.1 ± 9.1 13.4 ± 9.8 16.7 ± 11.4 3.2 ± 3.3
0.742 0.521 0.155 0.552 0.521
11/10 12 19
11/7 12 17
0.748 0.742 1.000
Chi-square was used for dichotomous variables, and two sample t-test was used for all other variables. Data were presented as mean ± SD. TLE, temporal lobe epilepsy; HS, hippocampal sclerosis; sGTCS, secondarily generalized tonic-clonic seizure.
between frequent-sGTCS (≥ 3 times per year) and infrequent-sGTCS, correcting for age, gender, and ICV. Statistically significant differences were observed at p b 0.05.
3. Results 3.1. Demographic and clinical data Thirty-nine patients with unilateral TLE–HS were enrolled. Of these, three of them were diagnosed with stereo-EEG evaluation and postoperative pathologic findings. The mean age was 31.2 ± 9.7 years and 15 of them (38.5%) were male. The age at onset was 15.6 ± 8.8 years, with a mean duration of 15.6 ± 10.4 years. Twenty patients (56.4%) had left TLE–HS. Based on the seizure type, the patients were divided into two groups: sGTCS (n = 21), and non-sGTCS (n = 18). The two groups showed no statistically significant differences in age, age at onset, gender, epilepsy duration, lateralization of epileptogenic zone, frequency of partial seizures, history of febrile convulsion, and pharmacoresistance (Table 1).
3
Further, the sGTCS group was divided into two subgroups: controlled-sGTCS (n = 8) and uncontrolled-sGTCS (n = 13). The two subgroups were statistically similar in age (p = 0.192), age at onset (p = 0.467), gender (p = 0.387), lateralization of epileptogenic zone (p = 1.000), frequency of partial seizures (p = 0.324), history of febrile convulsion (p = 0.673), and pharmacoresistance (p = 0.133). However, patients with uncontrolled-sGTCS had shorter epilepsy duration (12.0 ± 7.9 years vs. 20.6 ± 10.2 years, p = 0.042). 3.2. Shape analysis of the basal ganglia and thalamus Compared to the non-sGTCS group, the sGTCS group presented significant regional atrophy of the ipsilateral pallidum and thalamus (Fig. 1). Shape alterations were principally observed in the medial thalamus ipsilateral to the epileptogenic zone (cluster size = 1128 voxels, corrected p = 0.049, Max voxel coordinate = 107, 93, 67). Pallidal shape alterations were observed in the ventral regions (cluster size = 450 voxels, corrected p = 0.047, Max voxel coordinate = 107, 120, 65). There were no significant differences in the surface shape of the putamen and caudate between non-sGTCS and sGTCS groups at corrected statistical thresholds (p = 0.05). The uncontrolled-sGTCS subgroup had regional inward surface deflation of the ipsilateral thalamus relative to the controlled-sGTCS subgroup (Fig. 1). Thalamic alterations were observed in the medial area (cluster size = 1240 voxels, corrected p = 0.014, Max voxel coordinate = 107, 93, 67). There were no significant differences in the surface shape of the ipsilateral pallidum between the controlledsGTCS and uncontrolled-sGTCS subgroups at corrected statistical thresholds (p = 0.05). 3.3. Voxel-based morphometric analysis of cortex Compared to the non-sGTCS group, the sGTCS group presented significant regional atrophy in the ipsilateral middle frontal gyrus (MFG, cluster size = 60 voxels, corrected p b 0.05, Max voxel coordinate = −30, 49.5, 6, Fig. 2), and contralateral posterior cingulate
Fig. 1. 3D template and vertex-wise shape analysis of the ipsilateral pallidum and thalamus. sGTCS, secondary generalized tonic–clonic seizures; I, ipsilateral; GP, pallidum; TH, thalamus.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
4
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
Fig. 2. Voxel-based morphometric analysis showed regional atrophy of the ipsilateral MFG and contralateral PCC. MFG, middle frontal gyrus; PCC, posterior cingulate cortex; I, ipsilateral; C, contralateral.
cortex (PCC, cluster size = 130 voxels, corrected p b 0.05, Max voxel coordinate = 16.5, −57, 27, Fig. 2). Voxel values extracted from the significant clusters in group comparisons (the ipsilateral MFG and contralateral PCC) showed no correlation with epilepsy duration (p = 0.149 for MFG and p = 0.652 for PCC) and frequency of partial seizures (p = 0.110 for MFG and p = 0.519 for PCC), controlling for age, gender, and ICV. However, patients with frequent sGTCS showed significant atrophy in these regions compared with those with infrequent sGTCS, with smaller voxel values (the ipsilateral MFG, 0.334 vs. 0.473, p = 0.034; the contralateral PCC, 0.184vs. 0.474, p = 0.004). The uncontrolled-sGTCS subgroup showed no regional atrophy in the ipsilateral MFG and contralateral PCC when compared to the controlled-sGTCS subgroup. 4. Discussion Our study firstly revealed the structural alterations underlying the occurrence of sGTCS. In the mTLE study cohort, regional atrophy of the ipsilateral thalamus, pallidum, MFG, and contralateral PCC was associated with occurrence of sGTCS. Furthermore, the uncontrolled-sGTCS subgroup showed more pronounced atrophy in the ipsilateral thalamus than the controlled-sGTCS subgroup. These findings suggest that sGTCS in mTLE are not associated with diffused GM loss but rather are related to localized structural changes in the basal ganglia–thalamo–cortical network.
In the sGTCS group, the atrophied thalamic regions were specifically localized to the medial thalamus ipsilaterally rather than bilaterally (Fig. 1), which may indicate a difference in the epileptogenic networks of primarily and secondarily generalized tonic–clonic seizures [16–18]. Patients with uncontrolled-sGTCS showed more pronounced atrophy within the nearly identical regions compared to those with controlledsGTCS (Fig. 1), though the former had shorter epilepsy duration. There are many lines of evidence that suggest the early involvement of the thalamus during secondarily generalization in focal epilepsy [5,10,19]. The medial thalamic regions, including the medial pulvinar (PuM), mediodorsal nucleus (MDN), and centromedian nucleus (CMN) [20, 21], crucially participate in the regulation of both the cortical motor hyperexcitability and conscious state [22–25]. Specially, the dorsal midline thalamus (PuM and MDN) is strongly connected with the hippocampus in mTLE [21,26], and has been considered as a seizure synchronizer and amplifier of ictal activity (Fig. 3) [27–30]. Notably, the dorsal midline thalamus facilitates the generalization of seizures through thalamocortical connection [28], and serves as a central contributor to sGTCS in mTLE [4]. In animal models of mTLE, activation of GABA-A receptors explicitly in the dorsal midline thalamus can attenuate the excitatory thalamic output, thus suppressing sGTCS (Fig. 3) [28,29]. On the other hand, the CMN is a part of the reticular thalamus, projecting towards the dorsal midline thalamus with GABAergic neurons and decreasing both the number of generalized seizures and the duration of convulsions in animal models [31,32]. Moreover, stimulation of the
Fig. 3. Schematic illustration of the hypothesis: interaction of the hippocampus, thalamus, pallidum, and neocortex in the genesis of sGTCS in mTLE. Excitatory (inhibiting) effect is showed in red (blue), direct (indirect) effect is showed in solid (dashed) line. Medial view of the thalamus showing different nuclei refers to a published study [45]. PuM, medial pulvinar; MDN, mediodorsal nucleus; CMN, centromedian nucleus; sGTCS, secondarily generalized tonic–clonic seizures. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
CMN appears to be an efficacious treatment in patients with primarily and secondarily generalized seizures [33,34]. Atrophy of the CMN may decrease its inhibitory effect on the dorsal midline thalamus, thus facilitating the occurrence of sGTCS. However, the dorsal midline thalamus was also atrophied, which was similar to a postmortem study documenting pathologic changes in patients with mTLE, most of whom had a history of sGTCS [35]. A possible explanation for this includes hippocampal-evoked firing-induced secondarily progressive changes, for which atrophy in the dorsal midline thalamus has been correlated with epilepsy duration and severity [36,37]. Therefore, our findings strongly suggest that the medial thalamic atrophy is an important mediator for sGTCS. The ipsilateral ventral pallidum atrophy was also associated with sGTCS (Fig. 1). As a part of the basal ganglia, the pallidum exerts an inhibitory effect on the genesis of sGTCS [6–8]. Disinhibition of the ventral pallidum suppressed seizures, while inhibition of the ventral pallidum aggravated seizures in a genetic model of absence epilepsy in the rat [38]. It is conceivable that the ventral pallidum exerts a comparable modulatory influence on sGTCS. Rather than direct regulatory effect, the ventral pallidum plays an inhibitory role in sGTCS via the dorsal midline thalamus. Similar to the CMN in thalamus, the ventral pallidum projects to the dorsal midline thalamus with GABAergic neurons [22,23], and participates in the synaptic modulation of thalamic excitatory output to distant cortical regions, thus reducing the occurrence of sGTCS, which are in line with our study results (Fig. 3) [23]. Further, deep brain stimulation in the ventral pallidum may prevent sGTCS. In our study, atrophy of the ventral pallidum can be hypothesized to lead to decreased inhibitory effect on the dorsal midline thalamus, and thus contribute to generalization of hippocampal-evoked seizures. The voxel-based morphometric analysis revealed regional atrophy of the ipsilateral MFG and contralateral PCC in the sGTCS group (Fig. 2). Distant neocortex, including MFG (a part of the prefrontal cortex) and PCC, receive excitatory efferents of the dorsal midline thalamus (Fig. 3) [20,28], along with the spread of hippocampal-evoked firing [28]. In secondarily generalized epilepsy, typical GSW was involved in certain selected areas within the thalamocortical network, rather than the whole brain homogenously [19,39]. Interictal GSWrelated BOLD changes were found to be predominantly negative in the prefrontal cortex and PCC [10,19]. A study on ictal CBF changes, using single photon emission computed tomography, also found hypoperfusion in the prefrontal cortex and cingulate accompanied by hyperperfusion in the thalamus and basal ganglia [5]. Both the ipsilateral MFG and contralateral PCC showed sGTCS frequency-dependent atrophy in our study, suggesting that their involvement in sGTCS leads to secondarily progressive changes. In brief, these atrophic cortical regions are targets of the dorsal midline thalamus and components of the basal ganglia– thalamo–cortocal network (Fig. 3). In the present study, we applied automated morphometric approaches, such as vertex-wised surface shape analysis and VBM. Automated segmentation techniques have significant strengths in avoiding manual measurement errors and bias and can be applied widely in clinical practice. Nevertheless, a major shortcoming of occasional inaccuracies in structural segmentation was present. Therefore, several check-points (skull strip, segmentation, and surface reconstruction) were visually inspected with caution, and minor segmentation errors were manually corrected. VBM is one of the most frequently used quantitative volumetric analysis techniques [16,40,41]. However, it is a group-averaged morphometric analysis, and sensitive to inaccurate determination of graywhite matter interface and the arbitrary smoothing procedure [16,42]. Because of the methodological limitation of VBM in precisely assessing atrophy, we further applied the vertex analysis implemented in FIRST, which can provide useful information about the location and pattern of morphological changes of the subcortical structures on MRI [43]. The quantitative assessment of subcortical surface shape directly measures geometry without an additional smoothing procedure, and
5
can sensitively detect subcortical atrophy in previous studies [16,43]. In our study, whole-structure volumetry of the basal ganglia and thalamus showed no differences between the sGTCS and non-sGTCS groups using VBM (data not shown), while shape analysis found focal changes in the ipsilateral pallidum and thalamus. That is, the advanced analysis of regional structural morphology can detect subtle but biologically meaningful results that may be obscured by conventional wholestructure volumetry [44]. Thus, the high-sensitivity method is now applied in the evaluation of various neurological and psychiatric disorders [14,16,40,45]. Several limitations were present which should be considered. Firstly, the number of patients was relatively small (n = 39). To increase the statistical power, side-flipping of MRI data was done in the patients, which is frequently performed in quantitative studies [5,14,46]. However, it was revealed that gray matter atrophy was more intense in patients with left mTLE than those with right mTLE [14,47]. Additional studies with a larger sample size and subgroups of left and right mTLE will be essential to more firmly establish these findings. Secondly, our study was cross-sectional and thus interpretation of our results with respect to the causal relationship is limited. Future prospective studies incorporating a longitudinal design would provide a hint to disentangle the causal relations. These results, when combined with further investigations on the functional alterations associated with sGTCS, will likely provide a better understanding of the mechanisms of sGTCS, which would be reached which is helpful in developing novel treatment in the future. 5. Conclusion Localized atrophy in the ipsilateral basal ganglia, thalamus, as well as bilateral cortical regions forms the structural basis of sGTCS in mTLE. This morphologic analysis supports the basal ganglia–thalamo–cortical network as playing an important role in the genesis of sGTCS. Conflict of interest The authors declare that they have no conflict of interest. Ethical standard The study was approved by the Medical Ethics Committee of the Second Affiliated Hospital, Zhejiang University School of Medicine and written informed consent was obtained from all subjects (Study No. 2014-151). Acknowledgement This research was supported by the National Natural Science Foundation of China (No. 81271435, 81671282 and 91332202). References [1] De Boer HM, Mula M, Sander JW. The global burden and stigma of epilepsy. Epilepsy Behav 2008;12:540–6. [2] Helmstaedter C, Kurthen M, Lux S, Reuber M, Elger CE. Chronic epilepsy and cognition: a longitudinal study in temporal lobe epilepsy. Ann Neurol 2003;54:425–32. [3] Bone B, Fogarasi A, Schulz R, Gyimesi C, Kalmar Z, Kovacs N, et al. Secondarily generalized seizures in temporal lobe epilepsy. Epilepsia 2012;53:817–24. [4] Keller SS, Richardson MP, Schoene-Bake JC, O'Muircheartaigh J, Elkommos S, Kreilkamp B, et al. Thalamotemporal alteration and postoperative seizures in temporal lobe epilepsy. Ann Neurol 2015;77:1529–39. [5] Blumenfeld H, Varghese GI, Purcaro MJ. Cortical and subcortical networks in human secondarily generalized tonic–clonic seizures. Brain 2009;132:999–1012. [6] Popovic L, Vojvodic N, Ristic AJ, Bascarevic V, Sokic D, Kostic VS. Ictal dystonia and secondary generalization in temporal lobe seizures: a video-EEG study. Epilepsy Behav 2012;25:501–4. [7] Feddersen B, Remi J, Kilian M, Vercueil L, Deransart C, Depaulis A, et al. Is ictal dystonia associated with an inhibitory effect on seizure propagation in focal epilepsies? Epilepsy Res 2012;99:274–80.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
6
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
[8] Rektor I, Kuba R, Brázdil M, Halámek J, Jurák P. Ictal and peri-ictal oscillations in the human basal ganglia in temporal lobe epilepsy. Epilepsy Behav 2011;20:512–7. [9] Archer JS, Abbott DF, Waites AB, Jackson GD. fMRI “deactivation” of the posterior cingulate during generalized spike and wave. Neuroimage 2003;20:1915–22. [10] Hamandi K, Salek-Haddadi A, Laufs H, Liston A, Friston K, Fish DR, et al. EEG–fMRI of idiopathic and secondarily generalized epilepsies. Neuroimage 2006;31:1700–10. [11] Janszky J, Janszky I, Schulz R, Hoppe M, Behne F, Pannek HW, et al. Temporal lobe epilepsy with hippocampal sclerosis: predictors for long-term surgical outcome. Brain 2005;128:395–404. [12] Lhatoo S, Luders H. Textbook of epilepsy surgery. Informa Healthcare; 2008. [13] Kwan P, Arzimanoglou A, Berg AT, Brodie MJ, Allen H, Mathern G, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51:1069–77. [14] Keller SS, Richardson MP, O'Muircheartaigh J, Schoene-Bake JC, Elger C, Weber B. Morphometric MRI alterations and postoperative seizure control in refractory temporal lobe epilepsy. Hum Brain Mapp 2015;36:1637–47. [15] Song XW, Dong ZY, Long XY, Li SF, Zuo XN, Zhu CZ, et al. REST: a toolkit for restingstate functional magnetic resonance imaging data processing. PLoS One 2011; 6:e25031. [16] Kim JH, Kim JB, Seo W-K, Suh S-I, Koh S-B. Volumetric and shape analysis of thalamus in idiopathic generalized epilepsy. J Neurol 2013;260:1846–54. [17] Ciumas C, Savic I. Structural changes in patients with primary generalized tonic and clonic seizures. Neurology 2006;67:683–6. [18] Huang W, Lu G, Zhang Z, Zhong Y, Wang Z, Yuan C, et al. Gray-matter volume reduction in the thalamus and frontal lobe in epileptic patients with generalized tonic– clonic seizures. J Neuroradiol 2011;38:298–303. [19] An D, Dubeau F, Gotman J. BOLD responses related to focal spikes and widespread bilateral synchronous discharges generated in the frontal lobe. Epilepsia 2015;56: 366–74. [20] Eckert U, Metzger CD, Buchmann JE, Kaufmann J, Osoba A, Li M, et al. Preferential networks of the mediodorsal nucleus and centromedian–parafascicular complex of the thalamus—a DTI tractography study. Hum Brain Mapp 2012;33:2627–37. [21] Dinkelacker V, Valabregue R, Thivard L, Lehericy S, Baulac M, Samson S, et al. Hippocampal-thalamic wiring in medial temporal lobe epilepsy: enhanced connectivity per hippocampal voxel. Epilepsia 2015;56:1217–26. [22] Smith Y, Raju DV, Pare JF, Sidibe M. The thalamostriatal system: a highly specific network of the basal ganglia circuitry. Trends Neurosci 2004;27:520–7. [23] Root DH, Melendez RI, Zaborszky L, Napier TC. The ventral pallidum: subregionspecific functional anatomy and roles in motivated behaviors. Prog Neurobiol 2015;130:29–70. [24] Miller JW, Ferrendelli JA. The central medial nucleus: thalamic site of seizure regulation. Brain Res 1990;508:297–300. [25] Blumenfeld H. Impaired consciousness in epilepsy. Lancet Neurol 2012;11:814–26. [26] Behrens TE, Johansen-Berg H, Woolrich MW, Smith SM, Wheeler-Kingshott CA, Boulby PA, et al. Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging. Nat Neurosci 2004;54:750–7. [27] Guye M, Régis J, Tamura M, Wendling F, Gonigal AM, Chauvel P, et al. The role of corticothalamic coupling in human temporal lobe epilepsy. Brain J Neurol 2006; 129:1917–28. [28] Bertram EH, Zhang DX, Williamson JM. Multiple roles of midline dorsal thalamic nuclei in induction and spread of limbic seizures. Epilepsia 2008;49:256–68. [29] Sloan DM, Zhang DX, Iii EHB. Increased GABAergic inhibition in the midline thalamus affects signaling and seizure spread in the hippocampus–prefrontal cortex pathway. Epilepsia 2011;52:523–30.
[30] Sloan DM, Zhang DX, Iii EHB. Excitatory amplification through divergent–convergent circuits: the role of the midline thalamus in limbic seizures. Neurobiol Dis 2011;43: 435–45. [31] Nanobashvili Z, Chachua T, Nanobashvili A, Bilanishvili I, Lindvall O, Kokaia Z. Suppression of limbic motor seizures by electrical stimulation in thalamic reticular nucleus. Exp Neurol 2003;181:224–30. [32] Çavdar S, Onat FY, Çakmak YÖ, Yananli HR, Gülçebi M, Aker R. The pathways connecting the hippocampal formation, the thalamic reuniens nucleus and the thalamic reticular nucleus in the rat. J Anat 2008;212:249–56. [33] Valentín A, Navarrete EG, Chelvarajah R, Torres C, Navas M, Vico L, et al. Deep brain stimulation of the centromedian thalamic nucleus for the treatment of generalized and frontal epilepsies. Epilepsia 2013;54:1823–33. [34] Velasco M, Velasco F, Velasco AL, Jiménez F, Brito F, Márquez I. Acute and chronic electrical stimulation of the centromedian thalamic nucleus: modulation of reticulo-cortical systems and predictor factors for generalized seizure control. Arch Med Res 2000;31:304–15. [35] Sinjab B, Martinian L, Sisodiya SM, Thom M. Regional thalamic neuropathology in patients with hippocampal sclerosis and epilepsy: a postmortem study. Epilepsia 2013;54:2125–33. [36] Bernhardt BC, Bernasconi N, Kim H, Bernasconi A. Mapping thalamocortical network pathology in temporal lobe epilepsy. Neurology 2012;78:129–36. [37] Barron DS, Tandon N, Lancaster JL, Fox PT. Thalamic structural connectivity in medial temporal lobe epilepsy. Epilepsia 2014;55:e50–5. [38] Deransart C, Riban V, Lê BT, Hechler V, Marescaux C, Depaulis A. Evidence for the involvement of the pallidum in the modulation of seizures in a genetic model of absence epilepsy in the rat. Neurosci Lett 1999;265:131–4. [39] Rektor I, Zákopčan J, Tyrlíková I, Kuba R, Brazdil M, Chrastina J, et al. Secondary generalization in seizures of temporal lobe origin: ictal EEG pattern in a stereo-EEG study. Epilepsy Behav 2009;15:235–9. [40] Lu Y, Liang H, Han D, Yin M, Li Z, Cheng Y, et al. The volumetric and shape changes of the putamen and thalamus in first episode, untreated major depressive disorder. Neuroimage Clin 2016;11:658–66. [41] Dreifuss S, Vingerhoets F, Lazeyras F, Andino SG, Spinelli L, Delavelle J, et al. Volumetric measurements of subcortical nuclei in patients with temporal lobe epilepsy. Neurology 2001;57:1636–41. [42] Barron DS, Fox PM, Laird AR, Robinson JL, Fox PT. Thalamic medial dorsal nucleus atrophy in medial temporal lobe epilepsy: a VBM meta-analysis. Neuroimage Clin 2011;2:25–32. [43] Patenaude B, Smith SM, Kennedy DN, Jenkinson M. A Bayesian model of shape and appearance for subcortical brain segmentation. Neuroimage 2011;56:907–22. [44] Keller SS, O'Muircheartaigh J, Traynor C, Towgood K, Barker GJ, Richardson MP. Thalamotemporal impairment in temporal lobe epilepsy: a combined MRI analysis of structure, integrity, and connectivity. Epilepsia 2014;55:306–15. [45] Hänggi J, Bellwald D, Brugger P. Shape alterations of basal ganglia and thalamus in xenomelia. Neuroimage Clin 2016;11:760–9. [46] Yoo JY, Farooque P, Chen WC, Youngblood MW, Zaveri HP, Gerrard JL, et al. Ictal spread of medial temporal lobe seizures with and without secondary generalization: an intracranial electroencephalography analysis. Epilepsia 2014;55:289–95. [47] Coan A, Appenzeller S, Bonilha L, Li L, Cendes F. Seizure frequency and lateralization affect progression of atrophy in temporal lobe epilepsy. Neurology 2009;73:834–42.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy Linglin Yang a,1, Hong Li b,1, Lujia Zhu a, Xinfeng Yu b, Bo Jin a, Cong Chen a, Shan Wang a, Meiping Ding a, Minming Zhang b, Zhong Chen a, Shuang Wang a,⁎ a b
Epilepsy Center, Department of Neurology, Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China Departments of Radiology, Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China
a r t i c l e
i n f o
Article history: Received 16 November 2016 Revised 6 February 2017 Accepted 6 February 2017 Available online xxxx Keywords: Secondarily generalized tonic–clonic seizure Temporal lobe epilepsy Hippocampal sclerosis Shape analysis Medial thalamus Ventral pallidum
a b s t r a c t Mesial temporal lobe epilepsy (mTLE) is a common type of drug-resistant epilepsy and secondarily generalized tonic–clonic seizures (sGTCS) have devastating consequences for patients' safety and quality of life. To probe the mechanism underlying the genesis of sGTCS, we investigated the structural differences between patients with and without sGTCS in a cohort of mTLE with radiologically defined unilateral hippocampal sclerosis. We performed voxel-based morphometric analysis of cortex and vertex-wise shape analysis of subcortical structures (the basal ganglia and thalamus) on MRI of 39 patients (21 with and 18 without sGTCS). Comparisons were initially made between sGTCS and non-sGTCS groups, and subsequently made between uncontrolled-sGTCS and controlled-sGTCS subgroups. Regional atrophy of the ipsilateral ventral pallidum (cluster size = 450 voxels, corrected p = 0.047, Max voxel coordinate = 107, 120, 65), medial thalamus (cluster size = 1128 voxels, corrected p = 0.049, Max voxel coordinate = 107, 93, 67), middle frontal gyrus (cluster size = 60 voxels, corrected p b 0.05, Max voxel coordinate = −30, 49.5, 6), and contralateral posterior cingulate cortex (cluster size = 130 voxels, corrected p b 0.05, Max voxel coordinate = 16.5, −57, 27) was found in the sGTCS group relative to the non-sGTCS group. Furthermore, the uncontrolled-sGTCS subgroup showed more pronounced atrophy of the ipsilateral medial thalamus (cluster size = 1240 voxels, corrected p = 0.014, Max voxel coordinate = 107, 93, 67) than the controlled-sGTCS subgroup. These findings indicate a central role of thalamus and pallidum in the pathophysiology of sGTCS in mTLE. © 2017 Elsevier Inc. All rights reserved.
1. Introduction Mesial temporal lobe epilepsy (mTLE) is a common type of focal epilepsy, and is usually accompanied by hippocampal sclerosis. Seizures that occur in TLE with hippocampal sclerosis (TLE–HS) are typically resistant to antiepileptic drugs, which are associated with severe cognitive impairment, affective disorders, and stigmatization [1,2]. Among the seizure types associated with TLE–HS, secondarily generalized tonic– clonic seizures (sGTCS) are the most debilitating type and these in turn put the patients at additional risk of fatal injuries. Additionally, patients with TLE and sGTCS demonstrated worse surgical outcomes after anterior temporal lobectomy compared to those without sGTCS [3,4].
⁎ Corresponding author at: Epilepsy Center, Department of Neurology, Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310009, China. E-mail address: [email protected] (S. Wang). 1 These authors contributed equally.
Advancing the treatment strategies of sGTCS requires a detailed and clear understanding of the mechanism as to how the seizures are initiated, propagated, and terminated. However, the mechanism underlying the genesis of the sGTCS in mTLE has not been fully elucidated. Changes of cerebral blood flow (CBF) in the thalamus, basal ganglia, and frontoparietal cortex during the transition from partial seizures to secondarily generalization have been reported in patients with a mixed group of temporal and extra-temporal lobe epilepsy [5]. Some electrophysiological studies suggested that the basal ganglia exert an inhibitory effect against sGTCS in patients with mTLE [6–8]. Also, changes in the blood oxygen level-dependent (BOLD) signals were associated with generalized spike-wave (GSW) activity in the thalamo–cortical network. These were found in both primarily and secondarily generalized epilepsies, using simultaneous electroencephalogram and functional magnetic resonance imaging (EEG–fMRI) [9,10]. These lines of evidence suggest functional involvement of the basal ganglia-thalamo-cortical network in sGTCS. However, it remains unknown whether the structural alterations in these key nodes such as basal ganglia, thalamus, and cortex are involved in the pathophysiological mechanism of sGTCS.
http://dx.doi.org/10.1016/j.yebeh.2017.02.011 1525-5050/© 2017 Elsevier Inc. All rights reserved.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
2
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
Clinically, some patients with mTLE have sGTCS despite taking appropriate medication treatment, while some other patients have only partial seizures even though not on medications. This discrepancy in susceptibility to sGTCS might be associated with differences in the epileptogenic network, which may be reflected by morphometric analysis. In the present study, we investigated the relationship between cortical and subcortical gray matter (GM) alterations and the occurrence of sGTCS in a cohort of patients with unilateral TLE-HS. In particular, we initially applied sophisticated morphometric approaches to investigate the regional alterations in cortical volume and the surface shape of the thalamus and basal ganglia in patients with sGTCS versus those without sGTCS (non-sGTCS). Subsequently, we compared the structural alterations in patients with sGTCS that were not controlled by appropriate medical treatment (uncontrolled-sGTCS) versus those with medically controlled sGTCS (controlled-sGTCS).
2. Methods 2.1. Participants From April 2014 to June 2016, consecutive Chinese patients with unilateral TLE–HS, as confirmed by the clinical workup and imaging findings, were included in this study. Hippocampal sclerosis was defined if both hippocampal atrophy and increased T2 signals were observed on MRI [11]. All the patients were diagnosed by at least two experienced epileptologists based on history, seizure semiology, longterm scalp video-EEG monitoring, neuroimaging findings, and neuropsychological assessments. Some patients who required surgery were diagnosed as unilateral mTLE after stereo-EEG evaluation because their scalp EEG recordings did not provide adequate localizing or lateralizing information. Patients were excluded with: 1) history of brain trauma or surgery, 2) evidence of infectious origin, 3) evidence of secondary lesion that may be contributing to seizures, 4) serious psychiatric disorders, 5) ambiguous history of sGTCS, 6) seizures arising from the temporal lobe which is contralateral to the sclerotic hippocampus, or 7) bilateral mTLE. A typical sGTCS is characterized by bilateral rigid tonic extremities followed by rhythmic clonic jerks, along with complete loss of consciousness [12]. The seizure type was identified based on the seizure video, and the average seizure frequency was calculated for each patient according to the seizure diaries or family reports. Patients with sGTCS were further divided into two subgroups: 1) controlled-sGTCS subgroup, in which patients had secondarily generalized seizures only at the beginning of the disease course or after the medication was reduced, 2) uncontrolled-sGTCS subgroup, which included patients who had sGTCS events despite appropriate medical treatment with good compliance. Pharmacoresistance was defined as failure to achieve sustained freedom from seizures with adequate doses of two tolerated and appropriately chosen antiepileptic drugs [13]. The clinical variables were analyzed using two-sample independent t tests for continuous variables and Chi-square tests for categorical variables. The significance level was set at p b 0.05.
2.2. MRI protocol MR images were obtained on a 3.0T scanner (MR750, GE Healthcare, USA) with an 8-channel brain phased array coil. Highresolution coronal T2-weighted images perpendicular to the long axis of the hippocampus were acquired using spoiled gradient echo sequence with TR/TE = 5518/176 ms, flip angle = 110°, slice thickness = 2 mm, matrix = 512 × 512. Sagittal 3D T1-weighted images were acquired using brain volume imaging (BRAVO) sequence with TR/TE = 8.2/3.2 ms, TI = 450 ms, flip angle = 12°, slice thickness = 1 mm, matrix = 256 × 256.
2.3. Imaging analysis It has been reported that hemisphere lateralization of the epileptogetic zone is irrelevant for sGTCS history in patients with TLE [3]. Therefore, images of patients with right mTLE were side-flipped, and brain structures were marked as ipsilateral and contralateral to the epileptogenic zone in the present study. 2.3.1. Shape analysis of the basal ganglia and thalamus Automated segmentation and vertex-wise shape analysis of the basal ganglia and thalamus were carried out using the FSL-integrated registration and segmentation toolbox (FSL-FIRST) software (v.5.0.0; http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/first) installed on a Mac workstation. Briefly, the shape models in the FIRST were constructed from a library of manually segmented images. FIRST automatically searches for the most probable shape illustrations when given the observed intensities from the input images. To normalize the inter-individual head size differences, the segmented subcortical regions including the caudate, putamen, pallidum, and thalamus were reconstructed in the Montreal Neurological Institute space. Comparisons were initially made between non-sGTCS and sGTCS groups, and subsequently made between controlled-sGTCS and uncontrolled-sGTCS subgroups. Group differences were investigated by computing a general linear model, and the model included group (sGTCS/non-sGTCS or controlled-sGTCS/uncontrolled-sGTCS) as an independent factor, and age, gender, and intracranial volume (ICV) as covariates. Permutation testing using “randomize” with 5000 Monte Carlo simulations was used to calculate statistics of the segmented structures, corrected for multiple comparisons by dividing the p-value of 0.05. Clusterwise extent correction was applied, with a threshold of F N 2.0. Results from the surface analysis were visualized using Freesurfer's Freeview software [14]. 2.3.2. Voxel-based morphometric analysis of cortex The structural T1WI data were processed using voxel-based morphometry 8 (VBM8, http://dbm.neuro.unijena.de/vbm.html) toolbox of Statistical Parametric Mapping 8 (SPM8, http://www.fil.ion.ucl.ac. uk/spm) software running on MATLAB R2009b. In the first step, VBM automatically segmented GM, white matter (WM), and cerebrospinal fluid (CSF) in each patient. The volumes of GM, WM, and CSF voxels were determined separately for each patient and summed to calculate ICV. After segmentation, all images were rigidly aligned. Then, a “DARTEL” template of GM was created by nonlinearly aligning the GM images to a common space. The native GM images were normalized to the “DARTEL” template by applying the individual flow fields of all scans, using modulation to compensate for volume changes because of compression and/or expansion. Finally, GM images were smoothed using a Gaussian filter with full-width at half-maximum of 6 mm. Images were visually inspected at each and every processing step. Whole brain voxel-wise two-sample t tests were used to analyze the differences of the GM volume between non-sGTCS and sGTCS groups, with age, gender, and ICV as covariates. The statistical threshold was corrected for multiple comparisons on the cluster levels, which was determined by Monte Carlo simulation (5000 iterations) using AlphaSim as implemented in the REST-toolbox [15]. A combination threshold of p b 0.0001 on the voxel level and a cluster size N46 mm3 voxels was considered significant, which corresponded to a corrected p b 0.05. Then, the voxel values from the significant cluster in group comparison were extracted (data not shown). Subsequently, comparison of voxel values was made between controlled-sGTCS and uncontrolled-sGTCS, using multivariate analysis of covariance (MANCOVA) on age, gender, and ICV. To estimate the impairing effect of epilepsy, partial correlation coefficients were computed between the voxel values and epilepsy duration or frequency of partial seizures, controlling for age, gender, and ICV. To estimate the impairing effect of sGTCS, comparison of voxel values were made
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx Table 1 Demographic and clinical parameters in TLE-HS patients. Variable
sGTCS(n = 21)
Non-sGTCS(n = 18)
Sig.
Gender, M/F Age at scan (years) Age at onset (years) Epilepsy duration (years) Frequency of partial seizures (number per month) Lateralization of HS, L/R Febrile convulsions Pharmacoresistance
9/12 32.1 ± 10.3 17.5 ± 7.6 14.7 ± 9.5 2.6 ± 3.2
6/12 30.1 ± 9.1 13.4 ± 9.8 16.7 ± 11.4 3.2 ± 3.3
0.742 0.521 0.155 0.552 0.521
11/10 12 19
11/7 12 17
0.748 0.742 1.000
Chi-square was used for dichotomous variables, and two sample t-test was used for all other variables. Data were presented as mean ± SD. TLE, temporal lobe epilepsy; HS, hippocampal sclerosis; sGTCS, secondarily generalized tonic-clonic seizure.
between frequent-sGTCS (≥ 3 times per year) and infrequent-sGTCS, correcting for age, gender, and ICV. Statistically significant differences were observed at p b 0.05.
3. Results 3.1. Demographic and clinical data Thirty-nine patients with unilateral TLE–HS were enrolled. Of these, three of them were diagnosed with stereo-EEG evaluation and postoperative pathologic findings. The mean age was 31.2 ± 9.7 years and 15 of them (38.5%) were male. The age at onset was 15.6 ± 8.8 years, with a mean duration of 15.6 ± 10.4 years. Twenty patients (56.4%) had left TLE–HS. Based on the seizure type, the patients were divided into two groups: sGTCS (n = 21), and non-sGTCS (n = 18). The two groups showed no statistically significant differences in age, age at onset, gender, epilepsy duration, lateralization of epileptogenic zone, frequency of partial seizures, history of febrile convulsion, and pharmacoresistance (Table 1).
3
Further, the sGTCS group was divided into two subgroups: controlled-sGTCS (n = 8) and uncontrolled-sGTCS (n = 13). The two subgroups were statistically similar in age (p = 0.192), age at onset (p = 0.467), gender (p = 0.387), lateralization of epileptogenic zone (p = 1.000), frequency of partial seizures (p = 0.324), history of febrile convulsion (p = 0.673), and pharmacoresistance (p = 0.133). However, patients with uncontrolled-sGTCS had shorter epilepsy duration (12.0 ± 7.9 years vs. 20.6 ± 10.2 years, p = 0.042). 3.2. Shape analysis of the basal ganglia and thalamus Compared to the non-sGTCS group, the sGTCS group presented significant regional atrophy of the ipsilateral pallidum and thalamus (Fig. 1). Shape alterations were principally observed in the medial thalamus ipsilateral to the epileptogenic zone (cluster size = 1128 voxels, corrected p = 0.049, Max voxel coordinate = 107, 93, 67). Pallidal shape alterations were observed in the ventral regions (cluster size = 450 voxels, corrected p = 0.047, Max voxel coordinate = 107, 120, 65). There were no significant differences in the surface shape of the putamen and caudate between non-sGTCS and sGTCS groups at corrected statistical thresholds (p = 0.05). The uncontrolled-sGTCS subgroup had regional inward surface deflation of the ipsilateral thalamus relative to the controlled-sGTCS subgroup (Fig. 1). Thalamic alterations were observed in the medial area (cluster size = 1240 voxels, corrected p = 0.014, Max voxel coordinate = 107, 93, 67). There were no significant differences in the surface shape of the ipsilateral pallidum between the controlledsGTCS and uncontrolled-sGTCS subgroups at corrected statistical thresholds (p = 0.05). 3.3. Voxel-based morphometric analysis of cortex Compared to the non-sGTCS group, the sGTCS group presented significant regional atrophy in the ipsilateral middle frontal gyrus (MFG, cluster size = 60 voxels, corrected p b 0.05, Max voxel coordinate = −30, 49.5, 6, Fig. 2), and contralateral posterior cingulate
Fig. 1. 3D template and vertex-wise shape analysis of the ipsilateral pallidum and thalamus. sGTCS, secondary generalized tonic–clonic seizures; I, ipsilateral; GP, pallidum; TH, thalamus.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
4
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
Fig. 2. Voxel-based morphometric analysis showed regional atrophy of the ipsilateral MFG and contralateral PCC. MFG, middle frontal gyrus; PCC, posterior cingulate cortex; I, ipsilateral; C, contralateral.
cortex (PCC, cluster size = 130 voxels, corrected p b 0.05, Max voxel coordinate = 16.5, −57, 27, Fig. 2). Voxel values extracted from the significant clusters in group comparisons (the ipsilateral MFG and contralateral PCC) showed no correlation with epilepsy duration (p = 0.149 for MFG and p = 0.652 for PCC) and frequency of partial seizures (p = 0.110 for MFG and p = 0.519 for PCC), controlling for age, gender, and ICV. However, patients with frequent sGTCS showed significant atrophy in these regions compared with those with infrequent sGTCS, with smaller voxel values (the ipsilateral MFG, 0.334 vs. 0.473, p = 0.034; the contralateral PCC, 0.184vs. 0.474, p = 0.004). The uncontrolled-sGTCS subgroup showed no regional atrophy in the ipsilateral MFG and contralateral PCC when compared to the controlled-sGTCS subgroup. 4. Discussion Our study firstly revealed the structural alterations underlying the occurrence of sGTCS. In the mTLE study cohort, regional atrophy of the ipsilateral thalamus, pallidum, MFG, and contralateral PCC was associated with occurrence of sGTCS. Furthermore, the uncontrolled-sGTCS subgroup showed more pronounced atrophy in the ipsilateral thalamus than the controlled-sGTCS subgroup. These findings suggest that sGTCS in mTLE are not associated with diffused GM loss but rather are related to localized structural changes in the basal ganglia–thalamo–cortical network.
In the sGTCS group, the atrophied thalamic regions were specifically localized to the medial thalamus ipsilaterally rather than bilaterally (Fig. 1), which may indicate a difference in the epileptogenic networks of primarily and secondarily generalized tonic–clonic seizures [16–18]. Patients with uncontrolled-sGTCS showed more pronounced atrophy within the nearly identical regions compared to those with controlledsGTCS (Fig. 1), though the former had shorter epilepsy duration. There are many lines of evidence that suggest the early involvement of the thalamus during secondarily generalization in focal epilepsy [5,10,19]. The medial thalamic regions, including the medial pulvinar (PuM), mediodorsal nucleus (MDN), and centromedian nucleus (CMN) [20, 21], crucially participate in the regulation of both the cortical motor hyperexcitability and conscious state [22–25]. Specially, the dorsal midline thalamus (PuM and MDN) is strongly connected with the hippocampus in mTLE [21,26], and has been considered as a seizure synchronizer and amplifier of ictal activity (Fig. 3) [27–30]. Notably, the dorsal midline thalamus facilitates the generalization of seizures through thalamocortical connection [28], and serves as a central contributor to sGTCS in mTLE [4]. In animal models of mTLE, activation of GABA-A receptors explicitly in the dorsal midline thalamus can attenuate the excitatory thalamic output, thus suppressing sGTCS (Fig. 3) [28,29]. On the other hand, the CMN is a part of the reticular thalamus, projecting towards the dorsal midline thalamus with GABAergic neurons and decreasing both the number of generalized seizures and the duration of convulsions in animal models [31,32]. Moreover, stimulation of the
Fig. 3. Schematic illustration of the hypothesis: interaction of the hippocampus, thalamus, pallidum, and neocortex in the genesis of sGTCS in mTLE. Excitatory (inhibiting) effect is showed in red (blue), direct (indirect) effect is showed in solid (dashed) line. Medial view of the thalamus showing different nuclei refers to a published study [45]. PuM, medial pulvinar; MDN, mediodorsal nucleus; CMN, centromedian nucleus; sGTCS, secondarily generalized tonic–clonic seizures. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
CMN appears to be an efficacious treatment in patients with primarily and secondarily generalized seizures [33,34]. Atrophy of the CMN may decrease its inhibitory effect on the dorsal midline thalamus, thus facilitating the occurrence of sGTCS. However, the dorsal midline thalamus was also atrophied, which was similar to a postmortem study documenting pathologic changes in patients with mTLE, most of whom had a history of sGTCS [35]. A possible explanation for this includes hippocampal-evoked firing-induced secondarily progressive changes, for which atrophy in the dorsal midline thalamus has been correlated with epilepsy duration and severity [36,37]. Therefore, our findings strongly suggest that the medial thalamic atrophy is an important mediator for sGTCS. The ipsilateral ventral pallidum atrophy was also associated with sGTCS (Fig. 1). As a part of the basal ganglia, the pallidum exerts an inhibitory effect on the genesis of sGTCS [6–8]. Disinhibition of the ventral pallidum suppressed seizures, while inhibition of the ventral pallidum aggravated seizures in a genetic model of absence epilepsy in the rat [38]. It is conceivable that the ventral pallidum exerts a comparable modulatory influence on sGTCS. Rather than direct regulatory effect, the ventral pallidum plays an inhibitory role in sGTCS via the dorsal midline thalamus. Similar to the CMN in thalamus, the ventral pallidum projects to the dorsal midline thalamus with GABAergic neurons [22,23], and participates in the synaptic modulation of thalamic excitatory output to distant cortical regions, thus reducing the occurrence of sGTCS, which are in line with our study results (Fig. 3) [23]. Further, deep brain stimulation in the ventral pallidum may prevent sGTCS. In our study, atrophy of the ventral pallidum can be hypothesized to lead to decreased inhibitory effect on the dorsal midline thalamus, and thus contribute to generalization of hippocampal-evoked seizures. The voxel-based morphometric analysis revealed regional atrophy of the ipsilateral MFG and contralateral PCC in the sGTCS group (Fig. 2). Distant neocortex, including MFG (a part of the prefrontal cortex) and PCC, receive excitatory efferents of the dorsal midline thalamus (Fig. 3) [20,28], along with the spread of hippocampal-evoked firing [28]. In secondarily generalized epilepsy, typical GSW was involved in certain selected areas within the thalamocortical network, rather than the whole brain homogenously [19,39]. Interictal GSWrelated BOLD changes were found to be predominantly negative in the prefrontal cortex and PCC [10,19]. A study on ictal CBF changes, using single photon emission computed tomography, also found hypoperfusion in the prefrontal cortex and cingulate accompanied by hyperperfusion in the thalamus and basal ganglia [5]. Both the ipsilateral MFG and contralateral PCC showed sGTCS frequency-dependent atrophy in our study, suggesting that their involvement in sGTCS leads to secondarily progressive changes. In brief, these atrophic cortical regions are targets of the dorsal midline thalamus and components of the basal ganglia– thalamo–cortocal network (Fig. 3). In the present study, we applied automated morphometric approaches, such as vertex-wised surface shape analysis and VBM. Automated segmentation techniques have significant strengths in avoiding manual measurement errors and bias and can be applied widely in clinical practice. Nevertheless, a major shortcoming of occasional inaccuracies in structural segmentation was present. Therefore, several check-points (skull strip, segmentation, and surface reconstruction) were visually inspected with caution, and minor segmentation errors were manually corrected. VBM is one of the most frequently used quantitative volumetric analysis techniques [16,40,41]. However, it is a group-averaged morphometric analysis, and sensitive to inaccurate determination of graywhite matter interface and the arbitrary smoothing procedure [16,42]. Because of the methodological limitation of VBM in precisely assessing atrophy, we further applied the vertex analysis implemented in FIRST, which can provide useful information about the location and pattern of morphological changes of the subcortical structures on MRI [43]. The quantitative assessment of subcortical surface shape directly measures geometry without an additional smoothing procedure, and
5
can sensitively detect subcortical atrophy in previous studies [16,43]. In our study, whole-structure volumetry of the basal ganglia and thalamus showed no differences between the sGTCS and non-sGTCS groups using VBM (data not shown), while shape analysis found focal changes in the ipsilateral pallidum and thalamus. That is, the advanced analysis of regional structural morphology can detect subtle but biologically meaningful results that may be obscured by conventional wholestructure volumetry [44]. Thus, the high-sensitivity method is now applied in the evaluation of various neurological and psychiatric disorders [14,16,40,45]. Several limitations were present which should be considered. Firstly, the number of patients was relatively small (n = 39). To increase the statistical power, side-flipping of MRI data was done in the patients, which is frequently performed in quantitative studies [5,14,46]. However, it was revealed that gray matter atrophy was more intense in patients with left mTLE than those with right mTLE [14,47]. Additional studies with a larger sample size and subgroups of left and right mTLE will be essential to more firmly establish these findings. Secondly, our study was cross-sectional and thus interpretation of our results with respect to the causal relationship is limited. Future prospective studies incorporating a longitudinal design would provide a hint to disentangle the causal relations. These results, when combined with further investigations on the functional alterations associated with sGTCS, will likely provide a better understanding of the mechanisms of sGTCS, which would be reached which is helpful in developing novel treatment in the future. 5. Conclusion Localized atrophy in the ipsilateral basal ganglia, thalamus, as well as bilateral cortical regions forms the structural basis of sGTCS in mTLE. This morphologic analysis supports the basal ganglia–thalamo–cortical network as playing an important role in the genesis of sGTCS. Conflict of interest The authors declare that they have no conflict of interest. Ethical standard The study was approved by the Medical Ethics Committee of the Second Affiliated Hospital, Zhejiang University School of Medicine and written informed consent was obtained from all subjects (Study No. 2014-151). Acknowledgement This research was supported by the National Natural Science Foundation of China (No. 81271435, 81671282 and 91332202). References [1] De Boer HM, Mula M, Sander JW. The global burden and stigma of epilepsy. Epilepsy Behav 2008;12:540–6. [2] Helmstaedter C, Kurthen M, Lux S, Reuber M, Elger CE. Chronic epilepsy and cognition: a longitudinal study in temporal lobe epilepsy. Ann Neurol 2003;54:425–32. [3] Bone B, Fogarasi A, Schulz R, Gyimesi C, Kalmar Z, Kovacs N, et al. Secondarily generalized seizures in temporal lobe epilepsy. Epilepsia 2012;53:817–24. [4] Keller SS, Richardson MP, Schoene-Bake JC, O'Muircheartaigh J, Elkommos S, Kreilkamp B, et al. Thalamotemporal alteration and postoperative seizures in temporal lobe epilepsy. Ann Neurol 2015;77:1529–39. [5] Blumenfeld H, Varghese GI, Purcaro MJ. Cortical and subcortical networks in human secondarily generalized tonic–clonic seizures. Brain 2009;132:999–1012. [6] Popovic L, Vojvodic N, Ristic AJ, Bascarevic V, Sokic D, Kostic VS. Ictal dystonia and secondary generalization in temporal lobe seizures: a video-EEG study. Epilepsy Behav 2012;25:501–4. [7] Feddersen B, Remi J, Kilian M, Vercueil L, Deransart C, Depaulis A, et al. Is ictal dystonia associated with an inhibitory effect on seizure propagation in focal epilepsies? Epilepsy Res 2012;99:274–80.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011
6
L. Yang et al. / Epilepsy & Behavior xxx (2017) xxx–xxx
[8] Rektor I, Kuba R, Brázdil M, Halámek J, Jurák P. Ictal and peri-ictal oscillations in the human basal ganglia in temporal lobe epilepsy. Epilepsy Behav 2011;20:512–7. [9] Archer JS, Abbott DF, Waites AB, Jackson GD. fMRI “deactivation” of the posterior cingulate during generalized spike and wave. Neuroimage 2003;20:1915–22. [10] Hamandi K, Salek-Haddadi A, Laufs H, Liston A, Friston K, Fish DR, et al. EEG–fMRI of idiopathic and secondarily generalized epilepsies. Neuroimage 2006;31:1700–10. [11] Janszky J, Janszky I, Schulz R, Hoppe M, Behne F, Pannek HW, et al. Temporal lobe epilepsy with hippocampal sclerosis: predictors for long-term surgical outcome. Brain 2005;128:395–404. [12] Lhatoo S, Luders H. Textbook of epilepsy surgery. Informa Healthcare; 2008. [13] Kwan P, Arzimanoglou A, Berg AT, Brodie MJ, Allen H, Mathern G, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51:1069–77. [14] Keller SS, Richardson MP, O'Muircheartaigh J, Schoene-Bake JC, Elger C, Weber B. Morphometric MRI alterations and postoperative seizure control in refractory temporal lobe epilepsy. Hum Brain Mapp 2015;36:1637–47. [15] Song XW, Dong ZY, Long XY, Li SF, Zuo XN, Zhu CZ, et al. REST: a toolkit for restingstate functional magnetic resonance imaging data processing. PLoS One 2011; 6:e25031. [16] Kim JH, Kim JB, Seo W-K, Suh S-I, Koh S-B. Volumetric and shape analysis of thalamus in idiopathic generalized epilepsy. J Neurol 2013;260:1846–54. [17] Ciumas C, Savic I. Structural changes in patients with primary generalized tonic and clonic seizures. Neurology 2006;67:683–6. [18] Huang W, Lu G, Zhang Z, Zhong Y, Wang Z, Yuan C, et al. Gray-matter volume reduction in the thalamus and frontal lobe in epileptic patients with generalized tonic– clonic seizures. J Neuroradiol 2011;38:298–303. [19] An D, Dubeau F, Gotman J. BOLD responses related to focal spikes and widespread bilateral synchronous discharges generated in the frontal lobe. Epilepsia 2015;56: 366–74. [20] Eckert U, Metzger CD, Buchmann JE, Kaufmann J, Osoba A, Li M, et al. Preferential networks of the mediodorsal nucleus and centromedian–parafascicular complex of the thalamus—a DTI tractography study. Hum Brain Mapp 2012;33:2627–37. [21] Dinkelacker V, Valabregue R, Thivard L, Lehericy S, Baulac M, Samson S, et al. Hippocampal-thalamic wiring in medial temporal lobe epilepsy: enhanced connectivity per hippocampal voxel. Epilepsia 2015;56:1217–26. [22] Smith Y, Raju DV, Pare JF, Sidibe M. The thalamostriatal system: a highly specific network of the basal ganglia circuitry. Trends Neurosci 2004;27:520–7. [23] Root DH, Melendez RI, Zaborszky L, Napier TC. The ventral pallidum: subregionspecific functional anatomy and roles in motivated behaviors. Prog Neurobiol 2015;130:29–70. [24] Miller JW, Ferrendelli JA. The central medial nucleus: thalamic site of seizure regulation. Brain Res 1990;508:297–300. [25] Blumenfeld H. Impaired consciousness in epilepsy. Lancet Neurol 2012;11:814–26. [26] Behrens TE, Johansen-Berg H, Woolrich MW, Smith SM, Wheeler-Kingshott CA, Boulby PA, et al. Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging. Nat Neurosci 2004;54:750–7. [27] Guye M, Régis J, Tamura M, Wendling F, Gonigal AM, Chauvel P, et al. The role of corticothalamic coupling in human temporal lobe epilepsy. Brain J Neurol 2006; 129:1917–28. [28] Bertram EH, Zhang DX, Williamson JM. Multiple roles of midline dorsal thalamic nuclei in induction and spread of limbic seizures. Epilepsia 2008;49:256–68. [29] Sloan DM, Zhang DX, Iii EHB. Increased GABAergic inhibition in the midline thalamus affects signaling and seizure spread in the hippocampus–prefrontal cortex pathway. Epilepsia 2011;52:523–30.
[30] Sloan DM, Zhang DX, Iii EHB. Excitatory amplification through divergent–convergent circuits: the role of the midline thalamus in limbic seizures. Neurobiol Dis 2011;43: 435–45. [31] Nanobashvili Z, Chachua T, Nanobashvili A, Bilanishvili I, Lindvall O, Kokaia Z. Suppression of limbic motor seizures by electrical stimulation in thalamic reticular nucleus. Exp Neurol 2003;181:224–30. [32] Çavdar S, Onat FY, Çakmak YÖ, Yananli HR, Gülçebi M, Aker R. The pathways connecting the hippocampal formation, the thalamic reuniens nucleus and the thalamic reticular nucleus in the rat. J Anat 2008;212:249–56. [33] Valentín A, Navarrete EG, Chelvarajah R, Torres C, Navas M, Vico L, et al. Deep brain stimulation of the centromedian thalamic nucleus for the treatment of generalized and frontal epilepsies. Epilepsia 2013;54:1823–33. [34] Velasco M, Velasco F, Velasco AL, Jiménez F, Brito F, Márquez I. Acute and chronic electrical stimulation of the centromedian thalamic nucleus: modulation of reticulo-cortical systems and predictor factors for generalized seizure control. Arch Med Res 2000;31:304–15. [35] Sinjab B, Martinian L, Sisodiya SM, Thom M. Regional thalamic neuropathology in patients with hippocampal sclerosis and epilepsy: a postmortem study. Epilepsia 2013;54:2125–33. [36] Bernhardt BC, Bernasconi N, Kim H, Bernasconi A. Mapping thalamocortical network pathology in temporal lobe epilepsy. Neurology 2012;78:129–36. [37] Barron DS, Tandon N, Lancaster JL, Fox PT. Thalamic structural connectivity in medial temporal lobe epilepsy. Epilepsia 2014;55:e50–5. [38] Deransart C, Riban V, Lê BT, Hechler V, Marescaux C, Depaulis A. Evidence for the involvement of the pallidum in the modulation of seizures in a genetic model of absence epilepsy in the rat. Neurosci Lett 1999;265:131–4. [39] Rektor I, Zákopčan J, Tyrlíková I, Kuba R, Brazdil M, Chrastina J, et al. Secondary generalization in seizures of temporal lobe origin: ictal EEG pattern in a stereo-EEG study. Epilepsy Behav 2009;15:235–9. [40] Lu Y, Liang H, Han D, Yin M, Li Z, Cheng Y, et al. The volumetric and shape changes of the putamen and thalamus in first episode, untreated major depressive disorder. Neuroimage Clin 2016;11:658–66. [41] Dreifuss S, Vingerhoets F, Lazeyras F, Andino SG, Spinelli L, Delavelle J, et al. Volumetric measurements of subcortical nuclei in patients with temporal lobe epilepsy. Neurology 2001;57:1636–41. [42] Barron DS, Fox PM, Laird AR, Robinson JL, Fox PT. Thalamic medial dorsal nucleus atrophy in medial temporal lobe epilepsy: a VBM meta-analysis. Neuroimage Clin 2011;2:25–32. [43] Patenaude B, Smith SM, Kennedy DN, Jenkinson M. A Bayesian model of shape and appearance for subcortical brain segmentation. Neuroimage 2011;56:907–22. [44] Keller SS, O'Muircheartaigh J, Traynor C, Towgood K, Barker GJ, Richardson MP. Thalamotemporal impairment in temporal lobe epilepsy: a combined MRI analysis of structure, integrity, and connectivity. Epilepsia 2014;55:306–15. [45] Hänggi J, Bellwald D, Brugger P. Shape alterations of basal ganglia and thalamus in xenomelia. Neuroimage Clin 2016;11:760–9. [46] Yoo JY, Farooque P, Chen WC, Youngblood MW, Zaveri HP, Gerrard JL, et al. Ictal spread of medial temporal lobe seizures with and without secondary generalization: an intracranial electroencephalography analysis. Epilepsia 2014;55:289–95. [47] Coan A, Appenzeller S, Bonilha L, Li L, Cendes F. Seizure frequency and lateralization affect progression of atrophy in temporal lobe epilepsy. Neurology 2009;73:834–42.
Please cite this article as: Yang L, et al, Localized shape abnormalities in the thalamus and pallidum are associated with secondarily generalized seizures in mesial temporal lobe epilepsy, Epilepsy Behav (2017), http://dx.doi.org/10.1016/j.yebeh.2017.02.011