Address correspondence to Dr. Angelo Labate, Cattedra ed U.O. di Neurologia, Università degli Studi “Magna Græcia,” Campus Universitario Germaneto, Viale Europa, 88100 Catanzaro, Italy. E-mail: firstname.lastname@example.org
Purpose: In refractory mesial temporal lobe epilepsy (MTLE) extrahippocampal and neocortical abnormalities have been described in patients with or without mesial temporal sclerosis (MTS). Recently we observed gray matter reductions in regions outside the hippocampus in benign MTLE with or without MTS. Cortical thickness has been proposed as a viable methodologic alternative for assessment of neuropathologic changes in extratemporal regions. Herein, we aimed to use this technique to describe cortical abnormalities in patients with benign TLE.
Methods: Whole-brain cortical thickness analysis (FreeSurfer) was performed in 32 unrelated patients with benign TLE [16 patients with signs of MTS on magnetic resonance imaging (MRI), pMTLE; 16 without, nMTLE] and 44 healthy controls.
Key Findings: In the pMTLE group, the most significant thinning was found in the sensorimotor cortex bilaterally but was more extensive in the left hemisphere (false discovery rate, p < 0.05). Other areas were localized in the occipital cortex, left supramarginal gyrus, left superior parietal gyrus, left paracentral sulcus, left inferior/middle/superior frontal gyrus, left inferior frontal sulcus, right cingulate cortex, right superior frontal gyrus, right inferior parietal gyrus, right fusiform gyrus, and cuneus/precuneus. In the nMTLE, a similar neurodegenerative pattern was detected, although not surviving correction for multiple comparisons. Direct comparison between pMTLE and nMTLE did not reveal significant changes.
Significance: Patients with either benign pMTLE or nMTLE showed comparable cortical thinning, mainly confined to the sensorimotor cortex. This finding that is not appreciated on routine MRI supports the hypothesis that similar to refractory MTLE, even in benign MTLE, pathology in neocortical regions maybe implicated in the pathophysiology of this syndrome.
Temporal lobe epilepsy (TLE) is the most common type of focal epilepsy in adulthood, and it either may or may not be associated with mesial temporal sclerosis (MTS) (Meencke & Veith, 1991; Babb & Brown, 1997). In the past, MTS was widely considered a biomarker of pharmacoresistance (Meencke & Veith, 1991; Babb & Brown, 1997); however, a couple of recent studies using conventional magnetic resonance imaging (MRI) or postprocessing techniques illustrated that pathology in TLE broadens beyond the mesiotemporal structures (e.g., thalamus, amygdala and parahippocampal region, pre- and postcentral regions) and does not necessarily mean drug-resistance (Labate et al., 2006; Keller & Roberts, 2008; Labate et al., 2008). To address these issues, the morphometric MRI analysis of the brain has become a commonly used approach to investigating neuroanatomic correlates of neurologic disorders such as epilepsy. Among these new morphometric MRI methods, voxel-based morphometry (VBM) that gives probabilistic information about gray matter (GM) volume has been used extensively (Ashburner & Friston, 2000). In fact, very recently when looking at potential differences between patients with either refractory or benign mesial temporal lobe epilepsy (MTLE) with (pMTLE) or without MRI signs of MTS (nMTLE), we observed the same patterns of GM atrophy regardless of the seizure severity or the radiologic presence of MTS, together with subtle abnormalities within primary motor cortex although at a borderline of significance (Labate et al., 2006, 2010). Moreover, the involvement of neocortical regions, such as sensorimotor cortex in the pathophysiology of TLE, has been previously reported by other authors (Henry et al., 1993; Bonilha et al., 2004, 2005), but the significance of these findings is still a matter of active debate.
Estimation of cortical thickness has been proposed as a viable methodologic alternative to volumetric measurements for assessment of subtle cortical changes in neurologic diseases (Fischl & Dale, 2000; Rosas et al., 2002). This method accurately measures the thickness of the cerebral cortex across the entire brain and generates cross-subject statistics in a coordinate system based on cortical anatomy (Fischl & Dale, 2000). Two different available approaches have been applied to investigate neurodegenerative processes in patients with epilepsy showing overlapping results (Fischl & Dale, 2000; Rosas et al., 2002; Bernhardt et al., 2009, 2010). Compared to VBM that provides a mixed measure of cortical GM including cortical surface area and/or cortical folding, cortical thickness has the advantage of providing a quantitative value that represents a physical property of the brain that can be measured in an individual person.
In this study, we used cortical thickness measurements to identify subtle neuropathologic changes beyond hippocampus in a peculiar group of patients with benign and very mild MTLE with or without visual evidence of MTS.
Materials and Methods
Demographic features of our population are summarized in Table 1. Data and evaluation procedures on our MTLE patients have been reported in greater detail elsewhere (Gambardella et al., 2003; Labate et al., 2006). All patients and controls were matched for age and sex and gave informed consent to participate in this study. All patients had been free of major seizures for at least 1 year before the scanning. In each patient, the diagnosis of mild MTLE was made on the basis of a range of clinical seizure semiology, typical temporal auras, and electroencephalography (EEG) and MRI criteria that are considered to be reliable indicators of “benign” MTLE (Commission on Classification Terminology of the International League Against Epilepsy, 1989; Aguglia et al., 1998; Labate et al., 2006). Neurologic examinations were unremarkable in both groups. None of our patients had mental retardation. As summarized in Table 1, the study group was composed of 16 consecutive patients with benign nMTLE (11 female; mean age 38.4 ± 14.9 years) and 16 with benign pMTLE (nine female; mean age 29.1 ± 11.6 years). Forty-four healthy volunteers were recruited as the control group. The control population consisted of 21 female and 23 male subjects (mean age 35.3 ± 9.8 years) without any history of neurologic disorders or personal or familial history of loss of consciousness. MRI allowed the diagnosis of MTS based on a characteristic pattern of abnormalities (Jackson et al., 1993). The standard MRI studies of the whole group have been visually inspected to detect signs of MTS. Furthermore, we complemented the investigation by calculating the gray matter volume of regions of interest (ROIs) in the bilateral hippocampus as we previously described elsewhere (Labate et al., 2008).
Table 1. Clinical, EEG, and MRI features of mild TLE patients with and without MTS and controls
p-Value (t or χ2)
nMTLE, mesial temporal lobe epilepsy without MRI signs of MTS; pMTLE, mesial temporal lobe epilepsy with MRI signs of MTS; FC, febrile convulsions; EEG, electroencephalography; MTS, mesiotemporal sclerosis.
11 Female (68%)
9 Female (56%)
21 Female (48%)
38.4 ± 14.9
29.1 ± 11.6
35.3 ± 9.8
Age at onset (y)
24.9 ± 8.6
16.5 ± 11.8
13.5 ± 12.3
12.6 ± 8.7
Family history of FC/epilepsy, n (%)
Antecedent FCs, n (%)
Magnetic resonance imaging
Brain MRI was performed according to our routine protocol by a 1.5-T unit (Signa NV/I; GE Medical Systems, Milwaukee, WI, U.S.A). Structural MRI data were acquired using a three-dimensional (3D) T1-weighted spoiled gradient echo (SPGR) sequence with the following parameters: TR = 15.2 ms; TE = 6.7 ms; flip angle 15 degrees; matrix size 256 × 256; FOV = 24 cm; slice thickness = 1.2 mm. Subjects were positioned to lie comfortably in the scanner with a forehead-restraining strap and various foam pads to ensure head fixation. The image protocol was identical for all subjects (patients and controls) studied. The image files in DICOM format were transferred to a Linux workstation for morphologic analyses.
Cortical thickness analysis
MRI-based quantification of cortical thickness was performed using FreeSurfer (v. 4.05) software package (http://surfer.nmr.mgh.harvard.edu). This method has been previously described in detail (Dale et al., 1999; Fischl et al., 1999; Fischl & Dale, 2000). The procedure involves segmentation of white matter, tessellation of the gray/white matter junction, inflation of the folded surface, tessellation patterns, and automatic correction of topologic defects in the resulting main fold. Cortical thickness measurements were obtained by reconstructing representations of the gray–white matter boundary and the cortical surface. The distance between these two surfaces was calculated individually at each point across the cortical mantle. This method uses both intensity and continuity information from the entire 3D MRI volume in segmentation and deformation procedures to construct representations of cortical thickness. The maps are created using spatial intensity gradients across tissue classes and, therefore, are not simply reliant on absolute signal intensity. The entire cortex in each individual subject was then visually inspected, and any inaccuracies in Talairach-transformed, skull stripped, and segmentation were manually corrected and reinspected. Thickness measurements can be mapped onto the “inflated” surface of each participant’s reconstructed brain, thereby allowing visualization without interference from cortical folding. Maps were smoothed using a circularly symmetric Gaussian kernel across the brain/cortical surface with a standard deviation of 12.6 mm and averaged across participants using a nonrigid high-dimensional spherical averaging method to align cortical folding patterns. This procedure provides accurate matching of morphologically homologous cortical locations among participants on the basis of each individual’s anatomy while minimizing metric distortions, resulting in a mean measure of cortical thickness for each group at each point on the reconstructed surface. This spherical morphing procedure was used to construct the cortical thickness difference brain maps.
For each hemisphere, differences in cortical thickness between controls and pMTLE (reduced thickness in pMTLE), controls and nMTLE (reduced thickness in nMTLE), and pMTLE with nMTLE were tested by computing a general linear model (GLM) of the effects of “diagnosis” on cortical thickness at each vertex including age and gender as a covariate of no-interest. A false discovery rate (FDR) of p ≤ 0.05 was applied to correct for multiple comparisons. In addition, definition of the ROIs was performed by the detection of contiguous regions of statistical significance in the maps described earlier. These areas of regional thinning were used to create ROIs on a standard brain that were mapped back to each individual subject using spherical morphing to find homologous regions across subjects. A mean thickness score over each location was calculated for each subject. Because patients presented different lateralization of seizure focus and since that previous studies demonstrated the different impact of this variable on the neurodegenerative patterns of patients with TLE (Bonilha et al., 2007), we decided to delineate the different contribution of left and right seizure focus on the detected cortical thinning. pMTLE patients were divided into those with a left-sided (n = 6) or a right-sided (n = 6) seizure focus. The side of the focus was concordant with the side of hippocampal atrophy in patients with pMTLE. The results of this exploratory analysis have been reported on the Data S1.
Statistical analyses were performed with the Statistical Package for Social Science software (SPSS, version 12.0, Chicago, IL, U.S.A.) for Windows. For a continuous variable, mean, standard deviation (SD), and range were reported, and an unpaired t-test was used to assess differences among groups. Categorical variables are expressed as frequencies and percentages, and the differences among group distributions were assessed using the chi-square test. All statistical analyses had a two-tailed alpha level of p < 0.05 for defining significance. Spearman rank-order correlation coefficients were computed to assess the degree of relationship between cortical thickness and clinical variables.
The patients in the nMTLE group was older than those in the pMTLE group (respectively 38.4 ± 14.9 years vs. 29.1 ± 11.6 years), although a significant difference was not detected (post hoc Tukey test, p = 0.052). Mean age at seizure onset was, respectively, 24.9 ± 8.6 years in the nMTLE group and 16.5 ± 11.8 years in the pMTLE (p = 0.03). The nMTLE group showed a duration of epilepsy of 13.5 + 12.3 years, whereas this number was 12.6 + 8.7 years in the pMTLE. Family history of epilepsy or febrile convulsion (FCs) was not different between patients with nMTLE (5 of 16, 31%) and pMTLE (4 of 16, 25%). None of our patients had prolonged or complicated FCs, head trauma with loss of consciousness, or cerebral infections prior to seizure onset. The interictal epileptiform EEG abnormalities over the temporal regions were seen in nine patients (56%) with nMTLE (five right, three bilateral, and one left) and 12 patients (75%) with pMTLE (six right and six left). In the pMTLE group the interictal epileptiform EEG abnormalities were highly concordant to the side of MTS. The whole group of pMTLE patients had MRI evidence of unilateral MTS (nine right and seven left). ROI analysis confirmed the ipsilateral hippocampal atrophy that was detected in routine MRI scans. In all patients, MRI did not detect any mass lesion such as tumor, cortical dysgenesis, vascular lesion, malformation, or posttraumatic scars.
Group differences in cortical thickness
Figure 1 shows regions of cortical thinning in patients with nMTLE and pMTLE with respect to controls. In the pMTLE versus controls contrast, there was significantly thinning in the sensorimotor cortex, principally in postcentral gyrus (Fig. 1). The effect was bilateral, although considerably more extensive, in the left hemisphere. Other foci included the bilateral calcarine sulcus and occipital pole, the left supramarginal gyrus, the left superior parietal gyrus, the left paracentral sulcus, the left inferior/middle/superior frontal gyrus, the left inferior frontal sulcus, the right cingulate cortex, the right superior frontal gyrus, the right inferior parietal gyrus, the right fusiform gyrus, and the cuneus/precuneus. There were no suprathreshold peaks indicating greater thickness in the pMTLE patients. In the nMTLE versus controls, no significant data of cortical thickness measurement survived correction for multiple comparisons. However, when the threshold for significance was lowered to p ≤ 0.05 without correction for multiple comparisons, nMTLE had a more pronounced thinning in the bilateral postcentral gyrus and lateral occipital cortex, the left cingulate cortex, the left middle frontal gyrus, the left inferior frontal sulcus, the right calcarine sulcus, the right inferior parietal gyrus, and the right superior frontal gyrus. There were no suprathreshold peaks indicating greater thickness in the nMTLE. Direct comparison between pMTLE and nMTLE did not reveal significant changes, even at an uncorrected threshold.
We performed regression analyses in order to investigate whether clinical factors might influence changes in GM morphology of patients with MTLE. No significant correlations were detected, although a marginally significant relationship was detected between the thickness of the left postcentral gyrus with disease duration (r = −0.42; p = 0.09) in the pMTLE patients.
To our knowledge this is the first study that performed cortical thickness measurements in benign MTLE with either negative or positive MRI and found mainly cortical extratemporal abnormalities. Significant cortical thinning was found within the sensorimotor cortex, principally in the postcentral gyrus bilaterally in patients with benign MTLE regardless of MRI signs of MTS as well as the side of MTS. In detail, patients with pMTLE demonstrated abnormal thinning of several cortical regions, mainly in the left and right precentral and postcentral gyrus, superior/middle and inferior prefrontal cortex, supramarginal gyrus, and occipital lobe compared to those controls. Similar but less evident results were found in the nMTLE group. These findings were unsurprising because other authors previously observed analogous results in patients with MTLE, although in a population with severe MTLE (Bernhardt et al., 2008; McDonald et al., 2008; Bernhardt et al., 2009, 2010). In fact, the participation of cortical areas and specially the sensorimotor region in the epileptogenic network in intractable forms of MTLE could give rise to more severe and progressive damage due to duration of epilepsy as well as the presence of frequent and repetitive motor seizure (McDonald et al., 2008; Bernhardt et al., 2009, 2010). Likewise, Mueller et al. (2008) described the same pattern of atrophy over the sensorimotor cortex, even in refractory MTLE patients without evidence of MTS, suggesting that these patients have extensive extratemporal cortical thinning but with a lesser distribution than patients with MTLE and MTS often show.
The involvement of the sensorimotor cortex was also observed in our previous works using VBM, albeit at a lower uncorrected statistical threshold (Labate et al., 2008, 2010). In the latter VBM study we compared patients with refractory versus mild MTLE that showed the same GM abnormalities with a decrease in thalami, hippocampus, and sensorimotor cortex compared with healthy controls (Labate et al., 2010). Of course, the atrophy observed in the sensorimotor cortex was difficult to explain considering that the population studied had a mild epilepsy characterized by very rare complex partial seizures and even rarer secondary generalized seizures. Therefore, we assumed that cortical abnormalities in this peculiar population exist but are likely to be too subtle to be detected with VBM. However, the results from the current study using a different neuroanatomic trait (cortical thickness) reinforce and confirm our VBM results, suggesting that cortical thinning is also observed in subjects with rare motor phenomenon and negative MRI. It is possible that cortical abnormalities may have been underestimated using VBM, since this method provides a mixed measure of cortical gray matter including cortical surface area and/or cortical folding, as well as cortical thickness. The present findings of our cortical thickness study corroborate the observation of Bernhardt et al., who demonstrated progressive neocortical atrophy in intractable MTLE patients likely representing seizure-induced damage areas (Bernhardt et al., 2009, 2010). On the other hand, Mueller et al. (2008) observed a different pattern of thinning in patients with MTLE with and without MTS. Although our neocortical pattern is strongly similar to both, some differences have been detected in the involvement of the temporal cortex. In fact, conversely to Mueller et al., we did not detect cortical thinning differences in crucial areas such as temporal pole or entorhinal cortex.
Therefore, the nature of the neocortical thinning in both benign and severe MTLE remains unclear, even if a number of hypothesis have been proposed for the severe form. Indeed, many authors described thinning of the sensorimotor cortex in MTLE supported by VBM studies indicating an association between gray matter loss in the motor cortex with disease duration and frequent motor seizures. Furthermore, another potential variable could result from the chronic exposure to antiepileptic drugs as shown in an animal model (Bittigau et al., 2002). However, the results of our study cannot be explained with the same hypotheses, especially the firing of repetitive seizures because the population recruited was different, with unusual clinical features such as mild epilepsy, rare motor seizures, low doses of single antiepileptic drugs, and presence or not of MTS. Therefore, in our benign pMTLE or nMTLE patients we can suppose that the sensorimotor cortex is primarily caught up in the epileptogenic network underlying the disease. Experimental studies have already illustrated that the hippocampus is associated with sensorimotor processes, especially innate processes involving control of motor responses to sensory stimuli, and hippocampal mechanisms can directly influence locomotor activity (Bast & Feldon, 2003). This hypothesis is corroborated by growing animal and imaging models indicating the significant thalamocortical involvement in the genesis of temporal lobe epilepsy (Jones et al., 1979; Bertram et al., 1998; Bernasconi et al., 2004; Guye et al., 2006; Lin et al., 2007; Labate et al., 2008, 2010). In keeping with that, considering the reciprocal connections of the thalamus with the sensorimotor cortex already reported by Jones et al. as far back as 1979, it is reasonable to consider a picture in which temporal seizure activity would propagate throughout the thalamus to frontocentral regions. Secondly, very subtle microdysplastic tissue together with heterotopic neurons may cause cortical thinning over the frontal and sensorimotor cortex. This other hypothesis is suggested by surgical studies of patients with refractory MTLE as well as imaging studies mainly of VBM that captured signs of widespread cortical atrophy (Keller et al., 2002; Bonilha et al., 2005). Similarly, we also observed in our previous work on VBM, cortical atrophy over the central regions but without reaching statistical significance (Labate et al., 2008, 2010).
In conclusion, to our knowledge this is the first study that performed cortical thickness in patients with benign MTLE with or without MTS. We have clearly found evidence for significant structural abnormalities localized to the sensorimotor cortex. Our findings of sensorimotor atrophy further indicate both structural impairment of these regions and functional impairment (Van Paesschen et al., 2003). Together with functional abnormalities in thalamocortical circuit (Jones et al., 1979; DeCarli et al., 1998) and the structural abnormalities seen on VBM (Labate et al., 2008, 2010), the current findings construct a larger pathophysiologic circuit in the pathogenesis of MTLE, even though the seizures are well controlled. Further prospective studies in the same population would be helpful.
None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.