Address correspondence and reprint requests to Dr. N. Bernasconi at Montreal Neurological Institute and Hospital, 3801 University Street, Montreal, Quebec, Canada H3A 2B4. E-mail: email@example.com
Summary: Purpose: We previously showed a reduction in the volume of the entorhinal cortex (EC) ipsilateral to the seizure focus in patients with intractable temporal lobe epilepsy (TLE). The purpose of this study was to examine the specificity of EC atrophy in epilepsy.
Methods: We performed volumetric measurement of the EC on high-resolution magnetic resonance imaging (MRI) in patients with TLE (n = 70), extratemporal lobe epilepsy (ETE; n = 18), and idiopathic generalized epilepsy (IGE; n = 20). EC volumes of epilepsy patients were compared with those of 48 age- and sex-matched normal controls. Within the TLE group, 63 patients were selected prospectively with hippocampal atrophy ipsilateral to the seizure focus. The remaining seven patients were chosen retrospectively based on normal volumetric MRI of the hippocampus and amygdale, as well as normal histopathologic examination of the resected tissue.
Results: Compared with normal controls, EC volume was smaller ipsilateral but not contralateral to the seizure focus in patients with TLE (p < 0.001). No difference in the EC volumes ipsilateral and contralateral to the seizure focus was seen in patients with ETE and IGE compared with normal controls. The individual analysis showed that the EC was atrophic in 73% of TLE patients with hippocampal atrophy. Three of the seven TLE patients with normal volumetric MRI of the hippocampus and amygdala and normal histopathologic examination had EC atrophy ipsilateral to the seizure focus. In no patient with ETE or IGE was the EC found to be atrophic.
Conclusions: EC atrophy ipsilateral to the seizure focus appears to be specific to mesial temporal lobe structural damage associated with TLE.
The importance of the entorhinal cortex (EC) in temporal lobe epileptogenesis is being increasingly recognized. Electrophysiological studies in animal models of temporal lobe epilepsy (TLE) have shown that the EC is able to generate spontaneous epileptiform activity independent of input from the hippocampus (1) and may have a lower threshold for seizure elicitation than does the hippocampus (2). In patients with intractable TLE, investigations with stereotactically implanted depth electrodes have shown that seizures may originate in the EC (3). Neuropathologic examination of surgically resected specimens has revealed cell loss and astrogliosis in the EC (4,5), suggesting that the EC plays a pivotal role in the neuronal circuitry necessary for temporal lobe seizure activity.
With high-resolution magnetic resonance imaging (MRI), we previously demonstrated atrophic changes of the EC ipsilateral to the seizure focus in patients with intractable TLE (6,7). Similar results were found in subsequent MRI studies (8,9). We also showed that EC atrophy in TLE may occur regardless of hippocampal atrophy (10). However, the diagnostic specificity of this finding is not clear. It is not known whether volumetric changes of the EC are present only in TLE or whether they can be encountered in other types of epilepsy as well.
To determine whether EC changes are specific to TLE, we performed volumetric measurement of the EC on high-resolution MRI in groups of patients with medically refractory nonlesional TLE, extratemporal lobe epilepsy (ETE), and idiopathic generalized epilepsy (IGE).
We studied 101 consecutive age- and sex-matched patients admitted for the investigation of their epilepsy. The patient population consisted of 63 patients with medically intractable TLE and hippocampal atrophy on volumetric MRI [29 male subjects; mean age (±SD), 36 ± 11 years; range, 16–51 years], 18 patients with ETE (eight male subjects; mean age, 32 ± 8 years; range, 17–44 years), and 20 patients with IGE (10 male subjects; mean age, 35 ± 10 years; range, 15–52 years). In addition, we reviewed retrospectively the pathology reports between 1995 and 2000 of 73 TLE patients in whom volumetric MRI images of the hippocampus and amygdala were normal. Of 24 patients who underwent surgery, 19 had a suitable hippocampal specimens for qualitative histopathologic examination (11). In seven of 19, no evidence for neuronal loss or gliosis of the hippocampus was found in the histopathologic examination. Patients were compared with 48 healthy controls (23 men; age range, 20–65 years; mean, 32 ± 11 years). Patients with expansive or structural lesion (other than hippocampal atrophy in patients with TLE) on high-resolution MRI scans, including T1- and T2-weighted, proton-density, inversion-recovery, and fluid-attenuation inversion recovery scans, were not included in this study.
Identification of the type and location of seizure onset was determined by a comprehensive evaluation including detailed history, neurologic examination, review of medical records, neuropsychological evaluation, MRI scans, and an extensive EEG investigation.
In patients with TLE and ETE, the seizure focus was determined by predominantly ipsilateral interictal epileptic abnormalities and unequivocal unilateral seizure onset recorded during prolonged video-EEG monitoring. In TLE patients, the seizure focus was right-sided in 35 patients and left-sided in 35. In 10 patients with ETE, the seizure focus was right-sided, and in eight, it was left-sided. Fourteen patients had frontal lobe epilepsy; in two patients, the epileptic focus was parietal, and in two, it was occipital. IGE was defined according to the definition of Commission on Classification and Terminology of the International League Against Epilepsy (ILAE) (12).
Volumetric measurement of the hippocampus, as previously described (13), was performed in TLE patients and showed unilateral atrophy ipsilateral to the seizure focus in 63 patients and was normal in seven.
MRI scanning and analysis
MRIs were acquired on a 1.5-Tesla Gyroscan (Philips Medical System, Eindhoven, the Netherlands) by using a T1-weighted gradient-echo sequence (repetition time, 18 ms; echo time, 10 ms; one signal average; flip angle, 30 degrees; matrix, 256 × 256; field of view, 250 mm; slice thickness, 1 mm). Approximately 170 slices were acquired with an isotropic voxel size of 1 mm3.
Analysis was performed on a Silicon Graphics workstation (Mountain View, CA, U.S.A.). Images were automatically registered in a standard, stereotaxic space (14) to adjust for differences in total brain volume and brain orientation and to facilitate the identification of boundaries by minimizing variability in slice orientation (15). This procedure uses only a linear transformation. The automatic stereotaxic transformation is as accurate as the manual procedure, but shows greater stability (15). Each image underwent automated correction for intensity nonuniformity due to radiofrequency inhomogeneity of the MR scanner and intensity standardization (16).
Volumetric analysis was performed by using an interactive software package DISPLAY, developed at the Brain Imaging Center of the Montreal Neurological Institute. This program allows simultaneous viewing of MR images in coronal, sagittal, and horizontal orientations. The segmentation of the EC was performed based on our previously published protocol (6) (Fig. 1). The total duration for the measurement for the left and right EC was ∼20 minutes per subject.
Group differences for age were assessed by using Student's t test. The gender distribution was examined with the χ2 test. The statistical significance of differences in mean volumes between right and left EC volume was assessed by using the paired t test. Group differences of EC volumes were assessed by using analysis of variance (ANOVA) followed by Tukey's post hoc comparisons. For each individual, left–right asymmetries in EC volumes were calculated as follows: (L−R)/(L+R)/2, where L and R refer to the mean left and right volumes. Group differences in left–right asymmetries were examined by using ANOVA.
For analysis of individual patients, we considered as abnormal (a) values that were 2 SD below the mean of normal controls, (b) left–right asymmetries >2 SD above or less than the mean of normal controls.
In normal controls, the mean volume ± SD of the right EC (1,271 ± 143 mm3) and that of the left EC (1,244 ± 121 mm3) were not significantly different.
The results of group comparisons are shown in Fig. 2. Compared with that in normal controls, EC volume was small ipsilateral to the seizure focus in patients with both left and right TLE (p < 0.001), and an asymmetry was found (p < 0.001). No significant difference in contralateral EC volumes was noted between TLE patients and normal controls. No significant difference was found in EC volumes ipsilateral and contralateral to the seizure focus in patients with left ETE (p = 0.6 and p = 0.9, respectively) and right ETE (p = 1.0 and p = 0.9, respectively) compared with normal controls, and no significant asymmetry (p = 0.9). No significant difference was seen between left EC (p = 0.7) and right EC (p = 0.9) volumes and no asymmetry (p = 0.7) in IGE compared with normal controls.
The individual analysis showed that in 46 (73%) of 63 TLE patients with hippocampal atrophy, the EC was atrophic. In 43 (93%) of 46, the EC was atrophic ipsilateral to the seizure focus, and in three (7%) of 46, a bilateral symmetric atrophy was seen, and EC volume was normal in 17 (27%) of 63. Of the seven TLE patients with normal hippocampal and amygdalar volumes and normal histopathologic examination, three (43%) patients had EC atrophy ipsilateral to the seizure focus. In no ETE or IGE patient was the EC found to be atrophic.
Our study, based on homogeneous groups of patients with nonlesional epilepsy, shows that the EC is atrophic only in patients with TLE, but not in those with ETE or IGE. In agreement with our previous study (6), atrophy in TLE is lateralized to the side of the epileptic focus in the majority of the patients. In a previous study based on volumetric analysis of 10 patients with ETE, of whom eight had lesions, no atrophy was found in the EC (9).
In many centers for epilepsy surgery in which volumetric MRI measurements of mesial temporal lobe structures are included in the presurgical lateralization of the seizure focus, these measurements are generally performed only for the hippocampus and sometimes for the amygdala. We previously showed that measurement of the EC can provide correct lateralization of the seizure focus in 64% of TLE patients in whom no hippocampal or amygdalar atrophy is found (10). Furthermore, as shown in the present study, EC atrophy can be the only hallmark of mesial temporal disease in a large proportion of TLE patients, in whom both MRI volumetric measurement and histopathologic examination of the hippocampus are normal. As tissue was removed by subpial aspiration in our patients, histopathology of the EC was not available. In a previous study based on the examination of the surgical specimen in four TLE patients, astrogliosis and neuronal loss were seen predominantly in layer III of the EC (17). More recently, histopathologic examination of surgical specimens in 20 TLE patients showed variable degrees of pathologic changes in all layers of the EC (5) and a significant interindividual variation in the lesion distribution within layers. The lack of EC volume loss in some of our TLE patients could be explained by a similar phenomenon: atrophic changes being restricted to a specific layer and/or being too subtle to be quantified on MRI. Furthermore, in keeping with our data, pathologic abnormalities of the EC also were observed in the absence of hippocampal sclerosis (5).
Results from our previous studies and the present findings suggest that EC measurement provides additional information on the structural integrity of the mesial temporal lobe in TLE, particularly in those cases in which MRI volumetric assessment of the hippocampus and amygdala is normal (10). Conversely, differentiating between frontal and temporal lobe epilepsy can sometimes be difficult. The findings of this study demonstrate the potential value of volumetric analysis of the EC, not only in lateralizing temporal lobe disease and localizing atrophy, but also in helping to distinguish frontal from temporal epilepsies.
Aside from its importance in the lateralization of the seizure focus, MRI is being increasingly used in the building of classifiers to predict surgical outcome (18–20). With a bayesian classifier based on MRI volumetry of the hippocampus and amygdala and proton magnetic resonance spectroscopy imaging of the temporal lobe, we correctly predicted the surgical outcomes in 92% of patients who experienced worthwhile improvement (20). However, only 63% of patients who did not experience worthwhile improvement were correctly classified. Including MR volumetry of the EC could improve the classification accuracy.
Resection size also could influence the outcome, and including postoperative resection size of individual mesial temporal structures could potentially increase the predictive value of MRI in surgical outcome. The relative contribution of the removal of various mesial temporal lobe structures to surgical outcome has not been extensively studied. A previous qualitative evaluation of surgical removal on postoperative MRI images showed that a good seizure outcome was related to a radical removal of the parahippocampal gyrus, the bulk of which is composed of the EC (21). Further studies based on quantitative MRI analysis are needed to assess the relation between seizure freedom and the extent of resection of specific mesial temporal structures, such as the hippocampus, the amygdala, and the EC.
In conclusion, we showed that atrophy of the EC is present only in patients with TLE and is not found in ETE or in IGE. Volumetric measurement of the EC is a useful adjunct to the presurgical MRI evaluation of patients with intractable seizures.
Acknowledgment: This study was funded by a grant from the Canadian Institutes of Health Research (CIHR). Neda Bernasconi was supported by the Savoy Foundation for Epilepsy, St. Jean-sur-Richelieu, Quebec, Canada.