If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Neurology, Epilepsy Center, Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, ChinaDepartment of Neurology, Linhai Second People's Hospital, Taizhou, China
We firstly investigated the WM integrity according to drug response in FCD.
•
The WM microstructural abnormalities were different based on AED responsiveness.
•
Drug resistant epilepsy had a more extensive perilesional WM abnormalities pattern.
•
These findings may help to unravel the underlying neuroanatomical character of FCD.
Abstract
Purpose
We aimed to investigate the differences of white matter (WM) between the focal cortical dysplasia (FCD) patients with drug-resistant epilepsy and those with drug-responsive epilepsy.
Methods
Thirty epileptic patients with MRI-identified or histologically proven FCD were consecutively enrolled. Fractional anisotropy (FA) and mean diffusivity (MD) of the ipsilateral perilesional WM and contralateral homotopic WM layer masks were computed and corrected by the FA/MD of the corresponding hemispheric WM. The difference was evaluated using paired t-tests. The FA, MD and volumes of hemispheric WM and corpus callosum were also calculated.
Results
Patients with drug-resistant epilepsy showed significantly decreased FA and increased MD among ipsilateral perilesional WM layer 1 and 2, while patients with drug-responsive epilepsy showed decreased FA in only ipsilateral perilesional WM layer l, compared to remaining ipsilateral perilesional WM layers and contralateral layers 1 through 6. The integrity and volumes of the hemispheric WM and corpus callosum were similar between the two groups.
Conclusion
We demonstrated that the WM microstructural alterations differed between epileptic patients with FCD according to their antiepileptic drug responses. More extensive perilesional WM abnormality is observed in patients with drug-resistant epilepsy related to FCD.
As a common form of malformation of cortical development, focal cortical dysplasia (FCD) is one of the major etiologies of drug-resistant focal epilepsy and a common pathological finding in those who underwent resective epilepsy surgery [
]. It has been proven that drug-resistant focal epilepsy is associated with a pronounced reduction in quality of life, education, and employment prospects. Although the incidence of drug-resistance is high in epilepsy caused by FCD, a minority of those epileptic patients are reported to be responsive to antiepileptic drugs (AEDs) for a prolonged period of time or even life-long [
As gray matter's development is closely linked to its white matter, malformation of cortical development may also involves the underlying white matter [
]. Recently, several studies have attempted to understand the pathogenesis of FCD by assessing the underlying white matter (WM) using diffusion tensor imaging (DTI). It has been reported that epileptic patients with FCD exhibit WM microstructural abnormalities not only in the perilesional WM but also in the remote WM [
]. These findings emphasize that the WM microstructural abnormalities extend beyond MRI visible lesions. However, the currently literature on analysis of WM microstructure in epileptic patients related to FCD were usually based on the patients who have already developed drug-resistant epilepsy. Thus, there is limited knowledge in whether the WM microstructural abnormalities may vary in epileptic patients related to FCD with different AED responsiveness. By analyzing the WM microstructure of epileptic patients related to FCD with different AED responsive patterns, one could provide new insights into the neuroanatomical substrates of AEDs treatment response in epileptic patients related to FCD.
Herein, the study aimed to investigate the WM microstructural alterations in epileptic patients with different AEDs treatment response patterns. We hypothesized that the patients with drug-resistant epilepsy related to FCD would have more extensive WM abnormalities compared to the patients with drug-responsive epilepsy.
2. Material and Methods
2.1 Patient selection
This retrospective study was approved by the institutional review boards of the Second Affiliated Hospital of Zhejiang University(SAHZU). Written informed consent was obtained from all subjects. We reviewed the consecutive medical charts of patients from March 2012 to June 2019 from SAHZU. Only epileptic patients with MRI identified or histologically confirmed FCD, who had at least one year of follow-up after initiating either medical or surgical treatment, were included. Those without DTI and 3-dimensional (3D) T1 MRI sequences were excluded. In addition, patients with pathologically confirmed FCD type III were excluded. Age, sex, age at epilepsy onset, family history, perinatal adverse events, and febrile seizures were collected.
2.2 AED regimen and classification of AED treatment outcome
An AED regimen was defined as a trial of either a single drug (monotherapy) or a combination of 2 or more drugs [
]. Any changes in AED after initiation of therapy, such as substituting, adding drugs, or stopping the portion from the original drug combination, were defined as the end of one AED regimen. For example, switching to another AED because of severe adverse effects of the initial AED was considered a second regimen. If the type of drugs remained the same, changes in the dose of a single drug or any drugs in polytherapy were not regarded as reflective of the switching regimen.
The treatment outcomes of AEDs were assessed every three months, and the outcomes were classified into drug responsiveness and drug resistance. Drug-responsiveness was defined when a patient was seizure-free for at least one year with AED regimens (either monotherapy or combination of two or more drugs) at their last follow-up. Those who did not meet the drug-responsive criteria were defined as drug resistance. For patients who underwent epilepsy surgery, the last visit before undergoing the epilepsy surgery was recorded as the last follow-up visit and categorized into the drug resistance group. The surgical outcome was determined according to the Engel's classification scheme [
]. Surgical patients were classified as seizure free if they maintained with an Engel for score of Class I at their last follow-up. Postoperative clinical information was obtained from clinical visits and follow-up phone calls.
2.3 MRI protocols
MRI scans from SAHZU were performed on a 3T scanner (MR750, GE Healthcare), including a 3-dimensional (3D) T1 sagittal brain volume imaging sequence and diffusion tensor images. Diffusion tensor images were acquired using a spin-echo echo-planar imaging sequence (TR/TE = 8000/80.8 ms, flip angle = 90°, slice thickness = 2 mm, no gap, 67 slices, 30 volumes with noncollinear diffusion-weighted gradient directions [b = 1000 s/mm2] and five additional volumes without diffusion weighting, matrix = 128×128, axial plane resolution= 2×2 mm2, interpolated to 1 ×1 mm2). Detailed parameters were described in our previous publications [
Alterations in the hippocampal-thalamic pathway underlying secondarily generalized tonic-clonic seizures in mesial temporal lobe epilepsy: A diffusion tensor imaging study.
In nonsurgical patients, the diagnosis of FCD was made during a patient management conference based on multimodal data, including video-EEG, MRI, PET-MRI co-registration, MRI morphometric analysis programs (MAP), and clinical features. The MRI features of FCD include: cortical thickening, blurring of gray-white matter junction, hyperintense signal on T2 or fluid-attenuated inversion recovery (FLAIR) sequences, the “transmantle sign” and abnormal sulcal or gyral pattern. In surgical patients, FCD was diagnosed and classified according to the ILAE guidelines [
The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE Diagnostic Methods Commission.
A voxel-based MRI MAP was used to reveal subtle FCD lesions. MAP was carried out in SPM12 (Wellcome Department of Cognitive Neurology, London, UK) and in MATLAB 2015a (MathWorks, Natick, Massachusetts) [
]. Brain positron emission tomography (PET) images were acquired by a PET/CT scanner (Biograph mCT, Siemens Medical Solutions) at 40 min after intravenous injection of 18F-fluorodeoxyglucose, and the images were then coregistered to the corresponding MRI. MAP and PET/MRI co-registration were performed by an epileptologist (Shan Wang), who was unblinded and knew the patient's clinical and imaging information (Fig. 1). The detailed information was published previously [
(18)F-FDG PET and high-resolution MRI co-registration for pre-surgical evaluation of patients with conventional MRI-negative refractory extra-temporal lobe epilepsy.
Fig. 1Example of one patient with FCD. A shows 3-dimensional T1-weighted images. B shows T2-weighted fluid-attenuated inversion recovery (FLAIR) images on the same or closest slice. C shows co-registered MAP junction file. D shows co-registered PET finding. The white arrow shows the location of the lesion (FCD type IIb).
The FCD lesions were manually labeled by an experienced user (B.J.) based on the T1-weighted 3D MR images (including axial, coronal, and sagittal) aided by FLAIR and other T2-weighted images. Lesion segmentation was performed using the ITKSNAP software (v.3.4.02) [
2.5 Structural and diffusional image processing and analysis
One study reported the volumes of the corpus callosum were significantly lower in antiepileptic drug responders compared to nonresponders in patients with newly diagnosed focal epilepsy of unknown etiology[20]. Thus, we further analyzed the volume and integrity of corpus callosum in the following study. The T1-weighted 3D MRI images were processed with the FreeSurfer software (v.5.3.01) [
]. Hemispheric WM and corpus callosum masks were created by FreeSurfer, and segmentation results for each subject were visually reviewed independently by an experienced user (B.J.) for segmentation errors before further analysis. In the case of inaccuracies, manual editing was performed.
Functional Magnetic Resonance Imaging of the Brain (FMRIB) Diffusion Toolbox in FMRIB Software Library (FSL, v5.0.9) was used to process diffusion images. Affine registration of each diffusion‐weighted image to the base nondiffusion (b0) image was performed. To account for head motion, correcting for eddy currents distortion was performed. The gradient directions were also reoriented according to the transformational result, as previously suggested. The FSL' dtifit' command was used to reconstruct fractional anisotropy (FA) and mean diffusivity (MD) [
Each perilesional WM layer mask was dilated away from the FCD lesion by one voxel, for a total of six WM layers. The innermost layer, closest to the lesion, was labeled as layer 1 (L1), and the outermost layer was labeled as layer 6 (L6) (Fig. 2). Before creating the next layer, the lesion and the previous perilesional WM layers were merged to create a new seed to prevent the overlapping of voxels between layers [
]. Contralateral homotopic lesions were from side-flipped original FCD masks, which were created by FSL. Each contralateral homotopic WM layer mask was dilated away from the homotopic lesion by one voxel, for a total of six WM layers, which was the same as the previous steps in perilesional WM layer mask. Ipsilateral hemispheric WM was segmented by subtracting the FCD lesions from ipsilateral white matter masks. To minimize the inclusion of abnormal tissue in ipsilateral hemispheric WM masks, the FCD lesions were dilated successively three times prior to subtraction (as the following results showed that the abnormalities of perilesional NAWM extended to ipsilateral layers 2).
Fig. 2Lesion and perilesional white matter (WM) layer masks. (A) The white arrow represents lesion (FCD type IIb). (B) The red areas represent lesions. The blue, green, yellow, pink, white, and orange layers represent perilesional WM layer masks. The innermost layer adjoining lesion is layer 1, and the outermost layer away from the lesion is layer 6. (C) Dilated lesion outline (yellow), ipsilateral (red), and contralateral WM (blue) mask.
To correct for volume differences due to different head sizes, the volume of WM and FCD lesions were normalized by the total brain volume of each individual, respectively: normalized WM volume (or FCD lesion volume or corpus callosum) = WM volume (or FCD lesion volume or corpus callosum) × mean intracranial volume of the overall cohort/individual intracranial volume[19]. The corpus callosum, ipsilateral and contralateral hemispheric WM mask, perilesional WM layer mask and contralateral homotopic WM layer mask were individually applied to FA and MD maps, which were linearly aligned to their T1-weighed image. Corrected FA and MD were expressed as the ratio of raw FA/MD to mean FA/MD of the same side hemispheric WM.
2.6 Statistical analysis
Descriptive statistics were used for each variable to compare patients with drug-responsive to those with drug-resistant. If continuous variables (age, age at onset, duration of epilepsy) were normally distributed, 2-sample t tests was used. If not, Mann-Whitney U test was used. Fisher's exact test was used for categorical variables (male, febrile seizures, family history, head trauma, CNS infection, perinatal adverse events, lesion lateralization, BOSD, lesion location). The paired t-test was used to determine differences in the corrected FA and MD of perilesional WM and contralateral homotopic WM in both groups, drug-responsive and drug-resistant epilepsy. Analysis of covariance (ANCOVA) with gender and age as covariants was used to compare group differences in image parameters, such as the volume (FCD lesion, corpus callosum, and hemispheric WM), FA and MD (perilesional WM, contralateral homotopic WM, corpus callosum and hemispheric WM), between drug-responsive and drug-resistant epilepsy. Statistical significance was set at p <0.05.
3. Results
3.1 Patient characteristics
A total of 30 patients with both DTI and 3D T1 MRI sequence were enrolled, including 10 patients (33.3%) with MRI-identified FCD and 20 patients (66.7%) with histologically proven FCD (7 were FCD type IIa, 13 were FCD type IIb). According to the patient management conference, all 30 patients were MRI-positive, and we postulated that all patients with MRI-identified FCD types were highly indicative of FCD type II. Out of the 30 patients, 14 patients (46.7%) were male and 13 patients had the bottom of the sulcus dysplasia (BOSD). The mean age at seizure onset was 8.5 ± 6.7 years old, the mean age at the time of evaluation was 17.1 ± 7.6 years old, and the mean follow-up time was 27.8 ± 9.6 months. The FCD lesions were located in the frontal lobe in 19 patients, posterior quadrant in 4 patients, insular and opercular region in 4 patients, temporal lobe in 3 patients.
3.2 Outcome of AEDs treatment
At the time of the last follow-up visit, all 10 patients with MRI-identified FCD were found to have drug-responsive epilepsy: three patients achieved seizure freedom after the first AED regimen; four patients had experienced seizure fluctuance due to AEDs withdrawal but then achieved seizure freedom after re-initiation of another AED regimen; and three patients achieved seizure freedom after the second AED regimen. Surgical resection was performed in all 20 patients with drug-resistant epilepsy. Among them, 17 patients (85%) remained seizure-free with Engel Class I, and 3 (15%) had an Engel score of Class II. The demographic characteristics did not differ between in the drug-responsive group (n = 10) and the drug-resistant group (n = 20, Table 1).
Table 1Comparison of clinical data in drug-responsive epilepsy versus drug-resistant epilepsy.
Clinical characteristics
Drug-responsiveness (n = 10)
Drug-resistance (n = 20)
P value
Male, N (%)
4 (40.0%)
10 (50.0%)
0.709
Age at onset, Y
8.7 ± 5.2
8.4 ± 5.3
0.919
Age, Y
17.2 ± 7.7
17.0 ± 7.7
0.947
Duration of epilepsy, M
106.8 ± 80.4
112.7 ± 80.5
0.853
Febrile seizures
0
1(5.0%)
Family history
0
1 (5.0%)
Head trauma
0
0
CNS infection
0
1 (5.0%)
Perinatal adverse events
0
1(5.0%))
Lesion lateralization (left)
4 (40.0%)
9 (45.0%)
1.000
BOSD
4 (40.0%)
9 (45.0%)
1.000
Lesion location
Frontal lobe
7 (70.0%)
11 (55.0%)
Temporal lobe
2 (20.0%)
1(5.0%)
Insular lobe
1 (10.0%)
3(15.0%)
Posterior quadrant
0 (0%)
5(25.0%)
CNS, central nervous system; BOSD, bottom-of-sulcus cortical dysplasia; *Significantly different estimates. P-value is for Wilcoxon rank-sum, chi-square, or Fisher's exact tests as appropriate;
In patients with drug-resistant epilepsy, the corrected FA and MD abnormalities in the ipsilateral perilesional WM were observed to be extensive with the involvement of not only layer 1 but also layer 2 (p<0.05) compared to the remaining ipsilateral perilesional WM layers. When compared to the contralateral homotopic WM layer 1 through 6, ipsilateral perilesional WM layer 1 and layer 2 had significantly lower corrected FA and higher MD (p<0.05, Fig. 3A and 3B). In patients with drug-responsive epilepsy, no abnormality in FA and MD was observed in ipsilateral perilesional WM layer 1 to 6. When compared to the contralateral homotopic WM layer 1 through 6, only ipsilateral perilesional WM layer 1 showed lower corrected FA (p<0.05, Fig. 3C and 3D). Interestingly, ANCOVA showed no statistical difference in the integrity of ipsilateral perilesional WM and contralateral homotopic WM between drug-responsive and drug-resistant groups.
Fig. 3The corrected fractional anisotropy (FA) and mean diffusivity (MD) value of the perilesional white matter (WM) and contralateral homotopic region in patients with drug-resistant epilepsy and drug-responsive epilepsy. (A) The corrected FA value of drug-resistant epilepsy. (B) The corrected MD value of drug-resistant epilepsy. (C) The corrected FA value of drug-responsive epilepsy. (D) The corrected MD value of drug-responsive epilepsy. +p<0.05 represents the difference among ipsilateral layers. *p<0.05 represents the ipsilateral layers compared to the contralateral layers.
The volumes of the FCD lesions in patients with drug-responsive epilepsy and drug-resistant epilepsy were 3.38 ± 0.49 mm3 and 3.05 ± 0.42 mm3, respectively (p>0.05). The perilesional anomalies of FA and MD did not correlate with lesion volumes (p>0.05). Besides, there was no relationship between epilepsy duration and the extent of perilesional WM abnormality (p>0.05).
The normalized volumes of the corpus callosum in patients with drug-responsive epilepsy and drug-resistant epilepsy were 3847.06 ± 675.74 mm3 and 3765.58 ± 542.70 mm3, respectively (p>0.05). Furthermore, in patients with drug-responsive epilepsy, the FA and MD value were 0.52 ± 0.05 and 0.0011 ± 0.00010, respectively. In patients with drug-resistant epilepsy, the FA and MD value were 0.51 ± 0.05 and 0.0012 ± 0.00011, respectively. Thus, there was no significant difference in FA and MD between the two groups (p>0.05).
The normalized volumes of the hemispheric WM were displayed in Fig. 4. There was no statistically significant difference in the volume of hemispheric WM between the two groups. In addition, ANCOVA showed no difference in the integrity of hemispheric WM between the two groups (Fig. 4).
Fig. 4Histograms reflect the volume and diffusional metrics of the hemispheric white matter in patients with drug-resistant epilepsy and drug-responsive epilepsy. (A) The volume of the hemispheric white matter. (B) The fractional anisotropy (FA) of the hemispheric white matter. (C) The mean diffusivity (MD) of the hemispheric white matter.
Our study showed that the WM microstructural abnormalities in epileptic patients with FCD were different based on AED responsiveness. Compared to patients with drug-responsive epilepsy, patients with drug-resistant epilepsy showed a more extensive perilesional WM abnormalities pattern. These findings suggested that the changes in the integrity of perilesional WM are related to AED responsiveness in epileptic patients with FCD.
Recent publications on WM integrity analysis on mesial temporal lobe epilepsy (MTLE) are supportive and congruent with our study findings. Labate et al. revealed that drug-resistant MTLE was associated with more severly WM abnormalities than drug-responsive MTLE [
]. Furthermore, a longitudinal study demonstrated that extensive WM abnormalities had already pre-existed in the patients who had developed drug-resistance overtime compared to those who were drug-responsive epilepsy [
]. Thus, the result showed that WM changes might be a potential prognostic marker for long-term clinical outcomes and were involved in the development of drug resistance in patients with MTLE. However, the characteristics of WM integrity associated with drug-responsive patterns in epileptic patients with FCD have not been deeply probed.
Hong et al. used a multisurface analysis approach, coupled with DTI, showed that decreased FA of perilesional WM could extend outside the FCD type II lesion in patients with drug-resistant epilepsy [
]. However, no differentiation between the AEDs response was provided. Our study is the first to investigate the association between AED response and the WM integrity based on DTI analysis in epileptic patients with FCD. In both drug-resistant and drug-responsive groups, WM abnormalities were observed in the regions adjacent to the structurally visible borders of the lesions. In patients with drug-responsive, the abnormality of WM was more restricted than patients with drug-resistant epilepsy. These findings corroborate the hypothesis that in epileptic patients with FCD, the WM abnormalities extend beyond the MRI visible lesion [
]. Additionally, the corpus callosum is the largest WM structure, which connects the left and right sides of the brain. One study reported the volumes of the corpus callosum were different according to antiepileptic drug responses in patients with newly diagnosed focal epilepsy of unknown etiology. In drug-responsive patients, the volumes of the corpus callosum was lower, compared to those who were drug- nonresponsive patients [
]. However, there was no statistical difference in volumes and integrity of corpus callosum between the two groups of patients with FCD. These discordant findings are likely due to the different etiology and the small smaple in our cohort. Further studies including a larger number of patients with FCD could help to better understand the relationship between the corpus callosum and drug responsive pattern. As epilepsy is considered a network disorder, WM is the substrate connecting its various components as structural connectivity [
]. As structures within an epileptogenic network are involved in the generation and expression of seizures, we speculate that perilesional WM alterations might be one of the factors involved in developing drug resistance in epileptic patients with FCD. Those results may help unravel the underlying neuroanatomical character associated with drug resistance in epileptic patients with FCD, which could improve our ability to match patients to treatments.
In addition, one might raise a reasonable question of whether the perilesional WM abnormalities in patients with FCD are related to epilepsy duration. In our study, epilepsy duration was not correlated with the perilesional WM abnormality. Similarly, Zucca et al. combined 7T MRI and surgical specimens from epileptic patients with FCD and reported no correlation between epilepsy duration and WM abnormalities [
]. However, our study was cross‐sectional, and thus no further comment could be made regarding the causal relationship between the integrity of WM and drug resistance. Additionally, our sample was small, and all patients were reported to have FCD type II. Thus, further prospective and longitudinal studies, including new-onset and treatment naïve patients as well as patients with FCD type I, are warranted to investigate the particular causal association between WM abnormalities and drug resistance. Lastly, as only a few surgical patients underwent post-surgical MRI, the relationship between the volume of resected WM abnormalities and postsurgical seizure freedom was unable to analyze. Thus, further prognostic studies are warranted to investigate whether all the WM abnormalities in proximity of FCD lesions are necessary to be resected in order to achieve seizure freedom.
5. Conclusions
In conclusion, our findings demonstrate a differential pattern of WM microstructural abnormalities according to AED response in epileptic patients with FCD. Reduced microstructural integrity of perilesional WM is more extensive in patients with drug resistance epilepsy when compared o patients with drug-responsive epilepsy. These findings may help to unravel the underlying neuroanatomical character associated with drug resistance in epileptic patients with FCD.
Funding support
This study was supported by the National Natural Science Foundation of China (grant numbers: 81671282; 81801279; 81971207; 82001366), the Natural Science Foundation of Zhejiang Province (grant number: LQ20H090019), China and Zhejiang Medical Science and Technology Project (grant number: 2021KY020), China.
Declaration of Competing interest
None.
References
Barkovich A.J.
Guerrini R.
Kuzniecky R.I.
Jackson G.D.
Dobyns W.B.
A developmental and genetic classification for malformations of cortical development: update 2012.
Alterations in the hippocampal-thalamic pathway underlying secondarily generalized tonic-clonic seizures in mesial temporal lobe epilepsy: A diffusion tensor imaging study.
The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE Diagnostic Methods Commission.
(18)F-FDG PET and high-resolution MRI co-registration for pre-surgical evaluation of patients with conventional MRI-negative refractory extra-temporal lobe epilepsy.
00258-2&id=gr1.jpg){kind=link}
00258-2&id=gr2.jpg){kind=link}
00258-2&id=gr3.jpg){kind=link}
00258-2&id=gr4.jpg){kind=link}