Extratemporal White Matter Abnormalities in Mesial Temporal Lobe Epilepsy Demonstrated with Diffusion Tensor Imaging


Address correspondence and reprint requests to Dr. D.W. Gross at Division of Neurology, Department of Medicine, 2E3.19 Walter C. Mackenzie Health Sciences Centre, Edmonton, Alberta, T6G 2B7 Canada. E-mail: donald.gross@ualberta.ca


Summary: Purpose: Recent studies have demonstrated bilateral white matter abnormalities in temporal lobe epilepsy (TLE) patients with unilateral mesial temporal sclerosis (MTS). The purpose of this project was to determine whether abnormalities of water diffusion are seen in extratemporal white matter of patients with TLE and pathologically confirmed MTS and to determine whether these findings are associated with worse surgical outcome.

Methods: Eleven patients with TLE and unilateral MTS confirmed in surgical specimens and 14 controls were studied by using cerebrospinal fluid–suppressed diffusion tensor imaging (DTI) and T2 relaxometry.

Results: Hippocampal T2 signal for patients was significantly elevated both ipsilateral (p < 0.001) and contralateral (p = 0.006) to MTS. DTI demonstrated reduced fractional anisotropy of the genu of the corpus callosum (p = 0.003) and external capsule (p = 0.02) and elevated mean diffusivity of the genu (p = 0.005), splenium (p = 0.03), and external capsule (p < 0.001). For both the genu and external capsule, parallel diffusion of patients was not different from that of controls (genu, p = 0.81; external capsule, p = 0.45), whereas perpendicular diffusion was elevated (genu, p = 0.001; external capsule, p < 0.001). With mean postsurgical follow-up of 18.5 months, eight of 11 patients were entirely seizure free and the remaining three had all experienced a worthwhile reduction in seizure frequency.

Conclusions: Our findings suggest that although patients with TLE and MTS have extensive bilateral and extratemporal pathology, these findings may not be associated with a worse postsurgical outcome.


Temporal lobe epilepsy (TLE) is one of the most common focal epilepsy syndromes (1) with unilateral or strikingly asymmetrical mesial temporal sclerosis (MTS) being observed in the majority of cases (2,3). Recent evidence of bilateral white matter abnormalities in patients with unilateral MTS (4,5) suggest that patients presumed to have unilateral TLE may have extensive bilateral pathology. As diffuse disease is presumed to be associated with poor outcome, these findings could be of considerable significance for potential surgical candidates (5).

Diffusion tensor imaging (DTI) is a magnetic resonance imaging (MRI) technique that can indirectly evaluate the integrity of white matter tracts by measuring water diffusion and its directionality in three dimensions. Fractional anisotropy (FA) reflects the magnitude of directional water mobility. In normal fiber tracts, water diffusion is directional (high FA), whereas in degenerated tracts, FA decreases substantially (6). Further, correlation of DTI findings with histology has demonstrated that decreased diffusion parallel to fiber tracts (λ1) corresponds to axonal degeneration, whereas increased perpendicular diffusion (λ23) corresponds to myelin breakdown (7).

Bilateral symmetrical reductions of FA in the fornix and cingulum have been reported in patients with TLE and unilateral MTS (4). Although Arfanakis et al. (8) reported reduced FA of the external capsule and splenium of the corpus callosum in patients with TLE, the authors do not specify whether MTS was present. The purpose of this study was to use DTI to study extratemporal white matter in patients with TLE and confirmed mesial temporal pathology and to determine whether extratemporal pathology is associated with worse surgical outcome.


The protocol was approved by the University of Alberta Health Research Ethics Board, and informed consent was obtained from participants. Eleven patients with medically intractable TLE and clinical MRI evidence of unilateral MTS and 14 controls were studied. The mean age of patients and controls was 38 years and 32 years (range, 24–54 years), respectively, with no significant difference between groups (p = 0.18). Clinical information for patients is summarized in Table 1. By using a Siemens Sonata 1.5-T scanner, coronal T2 relaxometry was used to quantify MTS (9), and axial fluid-attenuated inversion recovery (FLAIR) DTI was used to evaluate white matter integrity. Sequence parameters have been previously reported (4).

Table 1. Clinical information and hippocampal T2 for patients
SubjectAge (yr)SexMTSDuration (yr)Onset (yr)Follow-up (m)OutcomeT2-ipsi (ms)T2-contra (ms)
  1. Duration, duration of epilepsy; onset, onset of epilepsy; follow-up, postsurgical follow-up; MTS, mesial temporal sclerosis; Mean hippocampal T2 for control subjects (collapsed right and left) was 114 ± 3 ms.

 120FLeft19 112Seizure free146119
 236FRight35 128Seizure free139117
 344MLeft43 111Engel class IIIa131121
 420MLeft17 323Seizure free132119
 559FLeft471223Seizure free160118
 653FLeft143928Seizure free139129
 735FRight171816Seizure free133127
 823MLeft17 623Seizure free141119
 938FRight37 116Engel class IIb130111
1048FRight47 113Engel class IIb133114
1137FLeft122510Seizure free126116

Regions of interest (ROIs) outlining each hippocampus were manually drawn in three consecutive slices of the monoexponential T2 maps to give an average T2 value. As no significant difference was observed between the right and left hippocampi of control subjects (T2 right, 114 ± 3 ms; T2 left, 115 ± 3 ms; p = 0.28), right and left values for controls were collapsed before comparison with patients. ROIs were manually drawn on FA maps outlining structures on one axial slice by a single blinded researcher and then copied to mean diffusivity and eigenvalues maps. The structures that were evaluated include the genu and splenium of the corpus callosum, the external capsule, and the anterior and posterior limbs of the internal capsule (Fig. 1). The apparent diffusion coefficient (ADC) parallel to the fiber tracts (ADC) is given by the largest principal eigenvalue (ADC1), and ADC perpendicular (ADC), by the mean of the two smaller principal eigenvalues (ADC= (λ23)/2). For paired structures, side-to-side comparisons were made for control and patient groups by using a paired t test (for patients, comparisons were made both between right and left and ipsilateral and contralateral to MTS). As no difference was observed (p > 0.3 for all structures in both groups), subsequent analysis was performed by using collapsed data (i.e., a single mean value per patient for each structure).

Figure 1.

Imaging findings: Axial non–diffusion-weighted CSF suppressed b0 (a, d), fractional anisotropy (FA) (b, e) and mean diffusivity (c, f) maps for a representative control and patient. All analyzed structures [genu and splenium of the corpus callosum, external capsule (EC) and anterior and posterior limbs of the internal capsule (IC)] are seen on a single axial slice for each subject with regions of interest being drawn on FA maps (b) and then copied to mean diffusivity and eigenvalues maps. Although statistical analysis demonstrated reduced FA of the genu of the corpus callosum and external capsule of patients, no subjective difference between patients and controls was apparent on visual inspection.

Statistical analysis

Between-group comparisons were made by using an unpaired t test with p < 0.05 being considered significant. Correlations between MRI parameters and age, duration of disease, and age at seizure onset were evaluated by using Pearson's correlation coefficient, with analysis being performed separately on patient and control groups.

Intrarater reliability

ROIs were drawn twice on eight subjects by the blinded investigator. The mean absolute difference in measurement of FA was demonstrated to be as follows: genu, 0.02; splenium, 0.03; external capsule, <0.01; anterior limb of internal capsule, 0.01; and posterior limb of the internal capsule, <0.01.


All patients underwent anterior temporal lobe resection with pathological confirmation of MTS in all cases. Eight of 11 patients are entirely seizure free (mean follow-up, 18.5 months), with the remaining three patients all experiencing a worthwhile reduction in seizure frequency (two patients are Engel class IIb, and one is class IIIa) (10).

T2 relaxometry

Hippocampal T2 values were significantly elevated in patients ipsilateral (T2, 137 ± 10 ms; p < 0.001) and contralateral (T2, 119 ± 5 ms; p = 0.006) to MTS as compared with controls (T2, 114 ± 3 ms). In all patients, T2 values were highest ipsilateral to MTS, with the difference in T2 between ipsilateral and contralateral hippocampi being significant (p < 0.001; Table 1).


FA was significantly reduced in the genu of the corpus callosum and external capsule, with no significant difference of FA being observed in the remaining structures (Fig. 2). Trace ADC was elevated in the genu, splenium, and the external capsule (Fig. 2). For the two structures with reduced FA, ADC for patients was not different from that of controls, whereas ADC was elevated (Fig. 2).

Figure 2.

Diffusion parameters: Box and whisker plots for fractional anisotropy (a), mean diffusivity (b), parallel diffusivity (c), and perpendicular diffusivity (d) (*p < 0.05; §p < 0.01). Reduced fractional anisotropy is observed in the genu of the corpus callosum and external capsule, whereas elevated mean diffusivity is observed in the genu and splenium of the corpus callosum and the external capsule of patients. For the two structures with reduced fractional anisotropy (the genu and external capsule), no difference in parallel diffusion is observed, whereas perpendicular diffusion of both structures was significantly increased. This finding is consistent with myelin degradation of the affected white matter tracts.


No significant correlation was observed between disease duration, age at seizure onset, or age and FA of the genu or the external capsule.


Reduced FA and increased ADC are consistent with myelin degradation of the genu of the corpus callosum and external capsule (7). Our findings of decreased FA in the external capsule are consistent with those of Arfanakis et al. (8). We, however, observed reductions of FA in the genu of the corpus callosum, whereas Arfanakis et al. reported decreased FA of the splenium. As well, whereas Arfanakis et al. (8) did not observe significant mean diffusivity abnormalities, we observed elevation in mean diffusivity of the genu and splenium of the corpus callosum and the external capsule. One possible explanation for the difference in results between the two studies is that all patients in our study had pathological confirmation of MTS, whereas Arfanakis et al. do not specify whether their patients had MTS. As mesial temporal pathology may have been different, it would not be unexpected if extratemporal pathology also were different. Thivard et al. (11) recently reported reduced FA of the corpus callosum (although they do not specify what portion of the callosum was abnormal) and cingulum in patients with TLE and unilateral MTS. Although a different method was used in their study (statistical parametric mapping), which could explain some of the differences in results, these findings provide further support for extratemporal white matter pathology in patients with unilateral MTS.

It has been hypothesized that bilateral hippocampal and extratemporal pathology may be associated with poor surgical prognosis (5,12), implying that patients who have a good surgical outcome have disease that is confined to the ipsilateral mesial temporal structures. Whereas the patients in our study were initially chosen based on standard clinical MRI evidence of unilateral MTS, subsequent quantitative measures demonstrated evidence of both significant contralateral mesial temporal and bilateral extratemporal pathology. Despite this, the surgical outcome in our series has been consistent with what is expected in patients with TLE and unilateral MTS (13). Quigg et al. (14) reported that although contralateral MTS was common in patients with TLE, the presence of contralateral pathology did not preclude excellent outcome. In a series of patients reported by Townsend et al. (15), of patients with TLE and MTS, nine of 11 patients with bilateral hippocampal T2 signal abnormalities and eight of nine with bilateral temporal lobe white matter T2 abnormalities were seizure free after surgery. Based on our observations, in combination with those of Quigg et al. (14) and Townsend et al. (15), we suspect that bilateral temporal and extratemporal pathology in patients with clinical MRI evidence of unilateral MTS is common and may not be predictive of poor surgical outcome.


Acknowledgment:  Operating and salary support were provided by the Savoy Foundation and University of Alberta Hospital Foundation (D.W.G.), Alberta Heritage Foundation for Medical Research and Canadian Institutes of Health Research (C.B.), and Promep (L.C.). MRI infrastructure was from the Canada Foundation for Innovation, Alberta Science and Research Authority, Alberta Heritage Foundation for Medical Research and the University of Alberta Hospital Foundation.