Optimizing outcomes in drug-resistant mesial temporal lobe epilepsy patients undergoing stereoelectroencephalography-guided radiofrequency thermocoagulation

Abstract

Background:

Mesial temporal lobe epilepsy (MTLE) epileptiform discharges have been reported to arise from the hippocampus or the extrahippocampal medial temporal cortex, such as the amygdala, and then propagate to the temporal lobe cortex. The surgical ablation of which of these structures would result in a better postoperative outcome is debatable.

Objective:

To assess the possible factors that might have influenced the postoperative outcome of a group of drug-resistant mesial MTLE patients who underwent stereoelectroencephalography (SEEG)-guided radiofrequency thermocoagulation (RFTC).

Design:

Single-center, retrospective.

Methods:

The present study utilized a pre- and postoperative gray matter voxel-by-voxel ablation mapping comparison approach, along with a white matter mapping of longitudinal changes in the native space technique, to evaluate the association between the post-SEEG implantation signal recordings (obtained from clinically relevant electrode contacts used during RFTC) and the post-RFTC ablation volume of the different selected regions of interest (ROIs).

Results:

The study included 22 patients (12 men and 10 women, mean age 28.86 ± 14.04 years). Sixteen patients (72.72%) were seizure-free (SF), and six patients (27.27%) were non-SF. Five patients (22.72%) experienced mild side effects following RFTC. The post-RFTC follow-up period varied from 12 to 48 months, with an average of 24.17 ± 9.86 months. The SF group was associated with a higher number of implanted electrode contacts in the amygdala that were used during RFTC, a larger preoperative volume of the amygdala; a larger ablation volume of both the amygdala and rhinal cortex. The ablation volume of the white matter was statistically similar between both groups.

Conclusion:

This study provides valuable insights into the significance of the amygdala and rhinal cortex as ROIs in the preoperative evaluation of patients with MTLE. Future implantation scheme plans should consider evaluating the preoperative volume of these ROIs. Additionally, increasing the number of electrode contacts implanted within these regions might be beneficial to capture more clinically relevant signals and enhance their ablation volume.

Keywords

amygdala gray matter MTLE RFTC rhinal cortex SEEG white matter

Introduction

Despite receiving the most effective anti-epileptic drugs (AEDs), approximately 25% of patients with epilepsy experience suboptimal clinical seizure control.¹ In the case of temporal lobe epilepsy (TLE), which is the most common form of adult epilepsy,² approximately 22% of patients fail to respond to AED treatment.³ Mesial temporal lobe epilepsy (MTLE), which comprises approximately 80% of TLE, is characterized by atrophy and sclerosis of the hippocampus and is also known as mesial temporal sclerosis.^4,5

Surgery is the main treatment for drug-resistant MTLE, with the goal of removing or disconnecting the epileptogenic zone (EZ).⁶ And in order to assess its possible removal or disconnection, an intracranial stereoelectroencephalography (SEEG) exploration might be necessary.⁷ Anterior temporal lobectomy and selective amygdalohippocampectomy are effective treatments for medically refractory MTLE,^8–11 with postoperative seizure-free (SF) rates ranging from 60% to 80%.^12,13 However, both surgical methods may lead to potential side effects such as cognitive impairment, visual field defects, intracranial hemorrhage, and neurological damage.¹⁴ Minimally invasive techniques such as SEEG-guided radiofrequency thermocoagulation (SEEG-guided RFTC), which has shown favorable outcomes and low risks, provide good alternatives to conventional surgery.^15–17 And in the case of SEEG, the utilization of intracranial electrodes that have already been implanted for clinical investigation, eliminates any further risk associated with invasive procedures. This allows for direct targeting of the EZ identified through SEEG signal analysis. RFTC utilizes radiofrequency current pulses delivered through SEEG electrodes to create targeted lesions in the epileptogenic regions.^18,19

The MTLE epileptiform discharges have been reported to arise from the hippocampus or the extrahippocampal medial temporal cortex, such as the amygdala, and then propagate to the temporal lobe cortex.^20,21 However, the distributions of seizure-onset zone (SOZ) in these regions may be specific to different patients.²² An intraoperative electrocorticography study hasshown that some MTLE patients exhibited the SOZs in the hippocampus–amygdala complex, while others displayed them in the hippocampus or amygdala alone.²³ And both the medial temporal structures and temporal neocortex have also been reported as the SOZs in several MTLE patients using SEEG recording.²⁴

The aim of this study was to assess the possible factors that might have influenced the postoperative outcome of a group of drug-resistant MTLE patients who underwent SEEG-guided RFTC after the refusal of the resective surgery option. Based on the evidence supporting that epileptiform discharges arise from different regions in MTLE patients and that SOZ can originate from different regions, we hypothesized that the partial surgical ablation of these different regions of interest (ROIs) within the temporal lobe, could be associated with positive postoperative outcome. To test this hypothesis, we intend to analyze the association between the post-SEEG implantation signals recordings and the post-RFTC ablation volume of the different temporal lobe structures and compare the findings between the SF and non-seizure-free (NSF) groups.

Methods

The study workflow is summarized in Figures 1 and 2.

Figure 1.

SEEG implantation planning scheme and anatomical position of electrode contacts.

Figure 2.

Gray and white matter RFTC volumetric analysis.

Patients selection

We performed a retrospective analysis on individuals diagnosed with drug-resistant MTLE who met the criteria outlined by the International League Against Epilepsy.²⁵ The study was carried out in accordance with the ethical guidelines outlined in the Declaration of Helsinki and received approval from the Ethics Committee Board of Fujian Medical University Union Hospital (No. 2024KY153), and written informed consent was duly obtained from all participants as well as from the parents or guardians of those who were minors.

We retrospectively reviewed the medical data and video-SEEG recordings of a consecutive cohort of patients who underwent SEEG-guided RFTC from October 2019 to April 2023 based on the STROBE guidelines. The selection of patients was based on the following criteria: (1) suspicion of unilateral MTLE established by the presence of ictal temporal lobe epileptiform discharges observed during video-electroencephalography (VEEG) monitoring; (2) confirmation of unilateral MTLE established by the presence of ictal and interictal discharges captured by SEEG; (3) failure to respond to a minimum of three AEDs; (4) medical history of focal seizures linked with MTLE semiology; (5) refusal to undergo resective surgery; and (6) at least 1 year follow-up. The exclusion criteria encompassed patients with malignancies, cavernous malformations, arteriovenous malformations, infarctions, temporal plus epilepsy, and those with dual pathology. Among 28 unilateral MTLE patients who underwent SEEG-guided RFTC, only 22 (12 males and 10 females) met the above-mentioned criteria and were included in this study.

Every patient underwent noninvasive preoperative assessment to localize the EZ. This assessment included semiology, scalp VEEG, neuropsychology testing, positron emission tomography, and magnetic resonance imaging (MRI) scans. Subsequently, another set of MRI scans was acquired 6 months following surgery.

MRI acquisition

All image acquisitions were conducted on a 3T MR scanner (MAGNETOM Prisma; Siemens Healthcare, Erlangen, Germany) equipped with a 64-channel head coil. The structural MRI protocol included pre-contrast sagittal T1-weighted magnetization-prepared rapid acquisition with gradient echo image with a voxel size of 0.9375 × 0.9375 × 0.9 mm³, axial T2-weighted (T2W) fast spin-echo images, axial fluid-attenuated inversion recovery T2W images.

The diffusion images were acquired using a grid sampling scheme to acquire 128 diffusion q-space samples, including 21 b values of 200, 350, 400, 550, 750, 950, 1100, 1150, 1500, 1650, 1700, 1850, 1900, 2050, 2250, 2400, 2450, 2600, 2650, 2950, and 3000 s/mm² along 3, 2, 4, 4, 3, 12, 7, 5, 6, 1, 14, 8, 4, 12, 4, 4, 8, 20, 4, 1, and 2 directions, respectively. The other scan parameters were repetition time, 2400 ms; echo time, 87 ms; slice acceleration factor, 3; n GRAPPA, 2; field of view, 230 × 230 mm²; voxel size, 2.5 × 2.5 × 2.5 mm³, without gap and acquisition time, 5 min 27 s. As the eddy current artifact produced by the grid sampling scheme cannot be corrected using eddy current correction, the bipolar pulse was used to handle eddy current at the sequence level.

Electrode implantation and SEEG recording

SEEG-guided RFTC was indicated as a surgical option to our selected patients after their refusal to undergo resective surgery. Our working theory regarding the localization of the EZs was the basis for determining the number and location of the implanted electrodes. This hypothesis was based on the above-mentioned noninvasive preoperative assessment. Given that all of our patients were diagnosed with MTLE during preoperative assessment, we conducted implantation in various regions of the temporal lobe, including both the mesial and lateral portions. Specifically, we targeted the hippocampus, amygdala, rhinal cortex (entorhinal and perirhinal cortex), parahippocampal cortex, as well as the superior, middle, and inferior temporal gyrus (Figure 1). During preoperative planning, the three-dimensional T1-weighted brain MRI and magnetic resonance angiography (artery and vein phases) were processed using the stereotactic software SinoPlan (Sinovation Medical, Beijing, China) to define the avascular trajectories of the electrodes. The software guided the robotic arm attached to a Mayfield or Dora 3-pin frame while the electrodes (0.8 mm in diameter, 8–16 contacts, 3.5 mm distance between contacts, 1.5 mm insulator length) were being implanted. A postoperative computed tomography (CT) scan was performed and co-registered with the preoperative MRI scan to check for correct electrode implantation and absence of postoperative hemorrhage.

SEEG recordings using the VEEG equipment Neuvo Amplifier (Neuroscan Compumedics, Abbotsford, Victoria, Australia) were carried out under chronic conditions, with reduced medication, for a duration of 1 week to 1 month, with the goal of capturing spontaneous habitual seizures. With a bandpass filter that ranged from 0 to 2500 Hz and a 10,000 Hz sampling rate, this system allowed for the simultaneous recording of up to 256 contacts. Intracranial EEG activity was recorded between contacts at various positions along each electrode’s axis. Visual analysis was performed on SEEG traces to distinguish interictal and ictal patterns. The SOZ was visually identified as the area that showed the first detectable change in SEEG before the seizure manifested clinically. The SEEG ictal onsets were deemed to be clinically significant when they showed periodic rapid discharge of spikes or low-voltage fast activity in the beta and gamma frequency ranges. Interictal epileptiform discharges (IEDs) were deemed to be clinically significant whether they contained spikes, polyspikes, spike-and-wave complexes, or polyspike-and-wave complexes. Following two or three habitual seizures or auras, the patients underwent electrical stimulations. Electrically elicited seizures (EES) that showed clinical similarities to spontaneous habitual seizures were considered significant.

RFTC procedure

To facilitate clinical monitoring of the patient during the procedure, SEEG-guided RFTC was carried out without anesthesia, at the end of the recording period and prior to removing the electrodes, using a radiofrequency generator system Model No. R2000B-M1 (BNS, Beijing, China). The generator achieved a peak output of 7.5 W, resulting in lesions of around 5–7 mm in diameter.^7,26 These lesions were created within a 40-s interval between two contiguous contacts of the same electrode, or between two adjacent contacts of different electrodes, which exhibited notable patterns of ictal onset, IEDs, and/or EES during SEEG recordings. When patients had a history of recurrent seizures prior to RFTC, SEEG monitoring was occasionally continued for an additional few days after thermocoagulation. Following the initial SEEG-guided RFTC procedure, seizures that persisted would be treated with a second RFTC procedure. Patients were discharged 1–2 days following the conclusion of the final RFTC procedure, with the removal of the electrodes under anesthesia.

RFTC outcome and data analysis

The general characteristics of the patients were evaluated to identify potential predictors of seizure outcomes following RFTC. Post-RFTC outcome was assessed at 1, 3, 6 months, and then on a yearly interval. Potential adverse effects were documented, and the outcomes of the patients were assessed by an epileptologist at least 1 year following the procedure using the Engel outcome scale and categorized into two groups: NSF (Engel ≠ IA), and SF (Engel IA). The precise anatomical location of the thermolesions was assessed with another set of MRI scans acquired 6 months following surgery.

Preoperative image processing and parcellation

The preoperative T1-weighted MRI underwent automatic parcellation with standard anatomic definitions using FreeSurfer version 7.4.0 (https://surfer.nmr.mgh.harvard.edu/).²⁷ Cortical reconstruction (recon-all) on preoperative MRI was performed to remove each patient’s brain mask (skull stripping) and obtain cortical parcellation.²⁷ The DKT (Desikan-Killiany Atlas) labeling protocol following recon-all was used since it subdivides the parahippocampal gyrus into the rhinal cortex (anterior portion) and parahippocampal cortex (posterior portion). Prior to use, the accuracy of each parcellated image was checked, and if necessary, corrective actions were taken in accordance with the FreeSurfer documentation.

Localization and visualization of electrode contacts

To assess the precise location of the postoperative SEEG electrodes’ contacts, we used the Brainstorm toolbox (https://neuroimage.usc.edu/brainstorm/Introduction). We imported each patient’s FreeSurfer data into Brainstorm and co-registered their postoperative CT to the preoperative MRI using SPM12 (https://www.fil.ion.ucl.ac.uk/spm/software/spm12/). We manually identify the position of each electrode’s contact tip and skull entry, and specify the number of electrode’s contacts, electrode length and diameter, contacts length, spacing, and diameter, based on the implantation scheme provided by the neurosurgeons, and the postoperative CT, which showed the actual position of each electrode (Figure 1).

Manual segmentation of post-RFTC cavities

The postoperative MRIs were co-registered to the preoperative MRIs using SPM12. The post recon-all DKT atlas file of each patient was imported to 3D Slicer version 5.2.2 (http://www.slicer.org), and was automatically co-registered with the postoperative co-registered MRI. The post-RFTC cavities from the co-registered postoperative MRI were manually segmented using the “Segment Editor” module in 3D Slicer (Figure 2). Manual segmentation was performed in the conventional manner by delineating the cavity boundaries on each millimeter slice. Since 3D Slicer supports simultaneous observation in three orthogonal planes, segmentations were performed in all three panels to guarantee accurate 3D reconstructions in a fourth panel. The preoperative images were then superimposed upon the segmented cavities. The preexisting cavities filled with cerebrospinal fluid from the preoperative images were manually segmented since they were merged with the post-RFTC cavities. Therefore, the actual post-RFTC segmented cavity volume was the subtraction of the segmentation of the preexisting cavities, from the segmented post-RFTC cavities. The entire process was repeated at least twice for each subject to guarantee accuracy and consistency, with a third repetition in case of significant discrepancies between the initial two.

Gray matter ablation volumetric analysis

The actual segmented cavities were superimposed upon the post-recon-all DKT atlas file. Five gray matter ROIs ipsilateral to the surgical side were left visible: hippocampus, amygdala, rhinal cortex, parahippocampal cortex, and the lateral temporal lobe structures (combination of the superior, middle, and inferior temporal gyrus). Each anatomical structure was studied separately by making the others invisible. Each structure’s actual RFTC volume was calculated by manually segmenting the degree of overlap between their anatomical parcellation from FreeSurfer and the actual segmented cavities (Figure 2). The entire process was repeated at least twice for each subject to guarantee accuracy and consistency, with a third repetition in case of significant discrepancies between the initial two.

White matter injury volumetric analysis

Clinical research on brain injuries generally measures white matter tract damage using two methods. The first method uses diffusion-weighted imaging (DWI) to directly assess white matter connection integrity. The second method overlays a patient’s lesion map on an atlas of white matter tracts derived using DWI datasets from healthy controls to indirectly evaluate white matter damage. In this study, we employed the first method. Differential tractography²⁸ using the February 8, 2023, build of DSI Studio (https://dsi-studio.labsolver.org/doc/gui_t3_dt.html) was used for the volumetric analysis of the different injured fibers (Figure 2). Mapping longitudinal change in the native space technique was used to assess the neuronal change reflected by a decrease in anisotropy in the repeat scans (i.e. pre- and postoperative DWI) of each patient. The generated fibers were converted into ROIs with their labels and volume. We excluded those with a mean volume percentage of less than 1% and those contralateral to the surgical side. The remaining ones were then combined as one ROI, which we labeled as temporal lobe-associated white matter.

Post-SEEG implantation and post-RFTC ablation volume association

We analyzed the potential association between the post-SEEG implantation signals within the different structures (from electrode contacts that exhibited relevant ictal onset patterns, IEDs, and EES, that were used during RFTC), and the post-RFTC ablation volume of these different ROIs relative to seizure outcome. We examined four main parameters that might have influenced the seizure outcome between the two groups: (1) the number of electrode contacts implanted in the different structures; (2) the number of electrode contacts implanted in the different structures used during RFTC; (3) preoperative volume of the different gray matter ROIs ipsilateral to the surgical side; and (4) post-RFTC ablation volume of the different structures.

Statistical analysis

The “Segments Statistic” module from 3D Slicer was used to calculate the ablation volume of the different gray matter ROIs. Other statistical analyses were performed using SPSS version 27 (IBM Corporation, Armonk, NY, USA), two-sided p < 0.05 was considered significant. The potential association and correlation between the post-ablation volume percentage of the different ROIs and the post-RFTC outcome were examined using Mann–Whitney U test, binary logistic regression analysis, and Spearman’s rank correlation analysis. Fisher exact test was used for categorical analysis.

Results

Patients demographics

There were 12 men and 10 women with an average age of 28.86 ± 14.04 years. Seizure frequency varied significantly among individuals, ranging from several per month to multiple per day. MRI findings were seemingly negative in 11 patients, MTLE was caused by hippocampal sclerosis (HS) in 10 patients, infection and HS in 1, and in 11 patients the cause was unknown (Table 1). Fourteen patients (63.63%) had ablation performed on both the mesial and lateral surfaces of the temporal lobe, and eight patients (36.36%) underwent ablation solely on the mesial temporal lobe. Twelve patients had secondary generalization, which was negatively correlated with seizure outcome (Table 2).

Table 1.

Patient demographics.

Patient No.	Etiology	Seizure duration (years)	Seizure frequency	MRI findings	RFTC side	Partial ablated structures	Number of electrodes	Responder to RFTC	Transient complications	Long-term complications	Follow-up (months)	RFTC outcome
1	HS	7	Monthly	HS	Left	Amy, PC, Hippo, LTL, WM	14	NSF	Slight vision and memory decline	Slight vision and memory decline	48	Engel IIA
2	HS	24	Monthly	HS	Right	Amy, RC, PC, Hippo, LTL, WM	10	SF	No	No	28	Engel IA
3	Infection/HS	6	Monthly	HS	Left	RC, PC, Hippo, LTL, WM	10	NSF	No	No	21	Engel IB
4	HS	12	Monthly	HS	Left	Amy, RC, PC, Hippo, LTL, WM	14	SF	No	No	29	Engel IA
5	HS	1	Daily	HS	Left	Amy, RC, PC, Hippo, LTL, WM	19	SF	No	No	25	Engel IA
6	HS	6	Monthly	HS	Left	Amy, RC, PC, Hippo, LTL, WM	12	SF	No	No	20	Engel IA
7	Unknown	18	Monthly	Negative	Right	Amy, RC, PC, Hippo, LTL, WM	11	SF	No	No	18	Engel IA
8	Unknown	13	Monthly	Negative	Left	Amy, RC, PC, Hippo, LTL, WM	12	SF	No	No	16	Engel IA
9	Unknown	8	Monthly	Negative	Left	Amy, RC, PC, Hippo, WM	14	SF	No	No	15	Engel IA
10	Unknown	23	Monthly	Negative	Right	Amy, RC, PC, Hippo, WM	15	SF	Slight memory decline and bad temper	Slight memory decline and bad temper	28	Engel IA
11	Unknown	3	Monthly	Negative	Right	Amy, RC, PC, Hippo, LTL, WM	13	SF	Memory decline	Memory decline	22	Engel IA
12	Unknown	6	Weekly	Negative	Right	Amy, RC, PC, Hippo, LTL, WM	10	SF	No	No	34	Engel IA
13	HS	10	Monthly	HS	Right	Amy, RC, Hippo, WM	14	NSF	No	No	21	Engel IIA
14	HS	5	Daily	HS	Right	Amy, RC, PC, Hippo, WM	13	NSF	No	No	23	Engel IIA
15	Unknown	7	Monthly	Negative	Right	Amy, PC, Hippo, WM	16	SF	No	No	14	Engel IA
16	Unknown	24	Monthly	Negative	Right	Amy, Hippo, LTL, WM	14	SF	Déjà vu	Déjà vu	12	Engel IA
17	HS	36	Weekly	HS	Left	Amy, RC, PC, Hippo, LTL, WM	12	NSF	No	No	14	Engel IIA
18	Unknown	11	Monthly	Negative	Left	Amy, RC, PC, Hippo, LTL, WM	13	SF	No	No	17	Engel IA
19	Unknown	28	Monthly	Negative	Left	Amy, RC, PC, Hippo, WM	14	SF	No	No	38	Engel IA
20	Unknown	1	Monthly	Negative	Right	Amy, RC, PC, Hippo, LTL, WM	10	SF	No	No	31	Engel IA
21	HS	11	Monthly	HS	Right	Amy, PC, Hippo, WM	11	NSF	Slight memory decline	Slight memory decline	26	Engel IIA
22	HS	20	Monthly	HS	Right	Amy, RC, PC, Hippo, WM	12	SF	No	No	12	Engel IA

Amy, amygdala; Hippo, hippocampus; HS, hippocampal sclerosis; LTL, lateral temporal lobe; MRI, magnetic resonance imaging; NSF, non-seizure-free; PC, parahippocampal cortex; RC, rhinal cortex; RFTC, radiofrequency thermocoagulation; SF, seizure-free; WM, white matter.

Table 2.

Correlation analysis of the variables with the outcome.

Parameters/(M ± SD)	r	p
Hippocampus ablation volume (%)	0.193	0.389
Amygdala ablation volume (%)	0.515	0.014
Rhinal cortex ablation volume (%)	0.484	0.022
Parahippocampal cortex ablation volume (%)	0.016	0.943
Temporal lobe-associated white matter injury volume (%)	−0.080	0.722
Epilepsy duration (years)	0.081	0.721
Secondary generalization	−0.559	0.007
Sex	−0.149	0.508
Surgical side	0.056	0.805
Age (years)	−0.008	0.972
Follow-up (months)	−0.089	0.695
Ablation of lateral surface of the temporal lobe	0.174	0.440

Outcome and morbidity following RFTC

The post-RFTC follow-up period varied from 12 to 48 months, with an average of 24.17 ± 9.86 months. A total of 283 electrodes were implanted across 12 patients in the right hemisphere and 10 patients in the left hemisphere, with an average of 12.86 ± 2.21 electrodes per patient. Following thermocoagulation, the MRI scans showed distinct areas of necrosis that were clearly visible among the coagulated contacts along the electrode paths. At the last follow-up visit, 16 patients (72.72%) were SF (Engel IA) while 6 patients (27.27%) were NSF (Engel ≠ IA) (Table 1). A total of five patients (22.72%) had mild side effects after undergoing RFTC. These side effects showed some improvement by their last follow-up.

Association between the different ROIs and the different parameters

A total of 15 fibers were selected following the differential tractography analysis: anterior commissure, arcuate fasciculus, cingulum parahippocampal parietal, corpus callosum body, corpus callosum forceps major, corpus callosum tapetum, fornix, inferior fronto-occipital fasciculus, inferior longitudinal fasciculus, middle cerebellar peduncle, middle longitudinal fasciculus, superior longitudinal fasciculus, thalamic radiation anterior, thalamic radiation superior, and uncinate fasciculus. Their ablation volume was not statistically associated with postoperative outcomes (p > 0.05) (Table 3).

Table 3.

Relationship between electrode contacts and seizure outcome.

ROIs	Number of electrode contacts implanted		Number of implanted electrode contact used during RFTC		Preoperative volume (mm³)		Ablation /injury volume (%)
ROIs	NSF/SF	p	NSF/SF	p	NSF/SF	p	NSF/SF	p
Amygdala	6.83 ± 4.21/12.44 ± 5.13	0.029	5.67 ± 3.61/11.81 ± 4.72	0.012	1603.50 ± 210.27/1907.44 ± 280.41	0.039	11.71 ± 11.29/26.85 ± 10.77	0.018
Hippocampus	30.17 ± 13.65/29.94 ± 9.10	0.458	24.17 ± 12.09/26.19 ± 8.06	0.195	3501.83 ± 640.89/3909.00 ± 807.78	0.185	19.61 ± 8.58/24.88 ± 9.08	0.376
Rhinal cortex	1.00 ± 1.54/1.81 ± 1.83	0.357	0.67 ± 0.81/1.44 ± 1.75	0.412	1390.67 ± 354.19/1545.00 ± 453.40	0.417	1.55 ± 2.21/9.75 ± 9.53	0.027
Parahippocampal cortex	4.00 ± 3.40/4.31 ± 3.92	0.911	3.17 ± 2.78/1.44 ± 1.15	0.185	1710.17 ± 353.75/2047.94 ± 352.56	0.065	5.25 ± 4.62/5.22 ± 5.30	0.941
Lateral temporal lobe	32.00 ± 5.17/33.69 ± 9.06	0.460	5.33 ± 6.50/9.00 ± 11.21	0.561	422,236.83 ± 5090.55/42,515.13 ± 5144.80	0.883	1.45 ± 1.60/1.48 ± 1.80	1
Temporal lobe associated white matter	70.50 ± 17.32/76.75 ± 14.96	0.555	17.33 ± 8.47/19.25 ± 11.79	0.912	/	/	5.20 ± 0.88/4.93 ± 1.28	0.712

NSF, non-seizure-free; RFTC, radiofrequency thermocoagulation; ROIs, regions of interest; SF, seizure-free.

Compared to the NSF group, the SF group had a significantly greater number of electrode contacts implanted in the amygdala (6.83 ± 4.21 vs 12.44 ± 5.13) and used during RFTC (5.67 ± 3.61 vs 11.81 ± 4.72), p = 0.029 and 0.012, respectively. Additionally, this group exhibited a larger ablation volume of the amygdala (26.85 ± 10.77) and the rhinal cortex (9.75 ± 9.53), compared to the NSF group (11.71 ± 11 and 1.55 ± 2.21), p = 0.018 and 0.027. The preoperative volume of the amygdala in the NSF group was significantly lower compared to the SF group (1603.50 ± 210.27 vs 1907.44 ± 280.41), p = 0.039. No significant findings were observed in any of the four main parameters relative to seizure outcome in the other ROIs (Table 3), and in epilepsy duration, sex, the total number of implanted electrodes independent to the ROIs, age, follow-up, surgical side, and ablation of the lateral surface of the temporal lobe (Table 4). The ablation of both the amygdala and rhinal cortex was positively correlated to seizure outcome (Table 2); however, only the ablation of the amygdala was statistically significant in the logistic regression model (Table 5).

Table 4.

Relationship between clinical characteristics and seizure outcome.

Parameters	NSF	SF	p
Epilepsy duration (years)	12.50 ± 11.74	12.81 ± 8.91	0.712
Number of electrodes implanted	12.33 ± 1.63	13.06 ± 2.40	0.573
Age (years)	27.83 ± 18.12	29.25 ± 12.89	0.971
Follow-up (months)	25.50 ± 11.70	22.44 ± 8.17	0.685
Parameters/n (%)	Outcome		p
Parameters/n (%)	NSF	SF	p
Ablation of the lateral surface of the temporal lobe
Yes	3 (50.00)	11 (68.75)
No	3 (50.00)	5 (31.25)	0.624
Sex
Men	4 (66.66)	8 (50.00)
Women	2 (33.33)	8 (50.00)	0.646
Surgical side
Right	3 (50.00)	9 (56.25)
Left	3 (50.00)	7 (43.75)	1
Secondary generalization
With	5 (83.33)	7 (43.75)
Without	1 (16.66)	9 (56.25)	0.161

NSF, non-seizure-free; SF, seizure-free.

Table 5.

Binary logistic regression of the two statistically significant variables and outcome.

Parameters	B	SE	Wals	df	p	OR	95% CI
Amygdala ablation volume (%)	0.142	0.070	4.093	1	0.043	1.153	1.004–1.324
Rhinal cortex ablation volume (%)	0.349	0.301	1.342	1	0.247	1.417	0.786–2.558

Discussion

The present investigation utilized a voxel-by-voxel ablation mapping comparison approach, along with the mapping of longitudinal changes in the native space technique, to evaluate the association between the post-SEEG implantation signal recordings (obtained from clinically relevant electrode contacts used during RFTC) and the post-RFTC ablation volume of the different selected ROIs in MTLE patients. Our research showed that improved seizure outcome was associated with: (1) a higher number of implanted electrode contacts in the amygdala which later were used during RFTC; (2) a larger preoperative amygdala volume; and (3) a larger ablation volume of both the amygdala and rhinal cortex. The subsequent sections will provide a detailed analysis of these factors in light of previous research findings.

Previous research has demonstrated that there is an association between the extent of entorhinal and parahippocampal cortex resection and postoperative outcome.^29,30 In a study of 87 TLE patients (47 SF, 40 NSF), Keller et al. reported that SF patients tended to have a greater entorhinal cortex and amygdala resection volume than NSF patients.³¹ A cohort of refractory MTLE patients who underwent resective surgery evaluated by Sagher et al. reported that 86%–90% of patients were SF 2 years after surgery, with up to 90% of the amygdala, hippocampal, and entorhinal cortex been removed.³² A recent study conducted by Gleichgerrcht et al. found that better seizure outcome in a group of TLE patients was associated with the resection of the anterior hippocampus–amygdala–piriform or entorhinal–perirhinal regions and disconnection of the frontal, limbic, and temporal regions through white matter tract loss.³³ However, none of these studies were able to assess if removing these regions individually or in combination improved surgical outcomes. Our results showed that a higher ablation volume of the amygdala and rhinal cortex contributed to positive postoperative outcomes, but as in the above-mentioned studies, we also were unable to determine whether surgical freedom was achieved by the partial ablation of these two ROIs, or whether their ablation in conjunction with the other ROIs improved outcomes. Postoperatively, none of the white matter tracts analyzed through differential tractography showed any significant association with seizure outcome. This lack of association is likely attributed to the similar and limited extent of injury incurred by these tracts during RFTC in both groups.

Regarding the preoperative volumetric assessment of the amygdala, the majority of studies have primarily concentrated on investigating its enlargement.^34–37 Shakhatreh et al., showed that postoperative seizure outcomes in a group of TLE patients were similar regardless of the presence of amygdala enlargement on preoperative MRI.³⁶ Na et al., revealed that in TLE patients with amygdala enlargement, a favorable prognosis was predicted by a rapid volume reduction rate of the enlarged amygdala and a favorable initial treatment response to AEDs.³⁵ However, these studies were conducted either in patients undergoing AEDs treatment, or in patients who underwent resective surgery with no comparison between SF and NSF groups in regard to the preoperative volume of the amygdala. Our study evaluated the disparity in the preoperative volume of the ROIs (gray matter) between the SF and NSF groups. Our findings showed a larger preoperative amygdala volume in the SF group compared to the NSF group. However, due to the absence of a control group or multiple preoperative MRI for comparison, we were unable to determine whether this observed larger preoperative amygdala volume was a result of enlargement. We conducted a supplementary analysis to investigate the potential volumetric asymmetry between the ipsilateral and contralateral amygdala in each group, and the number of electrodes implanted specifically in this region (Supplemental Tables 1 and 2). These supplementary analyses were not statistically significant. We theorized that the larger preoperative volume of the amygdala seen in the SF group may have led to the greater number of implanted electrodes’ contacts. This, in turn, has resulted in an increased recording of clinically relevant signals such as ictal onset patterns, IEDs, and EES. These relevant signals which were used to guide the selection of electrode contacts during RFTC, have led to a larger ablation volume within the amygdala in contrast to the NSF group. Our study also showed that positive seizure outcome was also associated with a higher ablation volume of the rhinal cortex. However, unlike the amygdala, the rhinal cortex didn’t show any statistical differences between the two groups in terms of the number of electrode contacts implanted, contacts used during RFTC, or preoperative volume. Visual inspection showed that in addition to the contacts implanted in the rhinal cortex (that were used during RFTC), the ablation of the amygdala and hippocampus also resulted in the ablation of some of its regions. The larger ablation volume of the rhinal cortex seen in the SF group was due to the greater number of implanted electrode contacts in the amygdala used during RFTC. Specifically, the contacts implanted nearby contributed to the increase in its ablation area.

The total ablation of the mesial temporal lobe complex structure may not be necessary for the treatment of MTLE in patients who refuse to undergo resective surgery. Postoperatively, 72.72% of our MTLE patients were SF, which is in line with the studies of Malikova et al., reporting nearly 75% of MTLE patients being SF 2 years after stereotactic radiofrequency amygdalohippocampectomy (SAHE),³⁸ and of Liscak et al., also reporting 78% of SF MTLE patients 2 years following SAHE.³⁹ The occurrence or progression of epilepsy may be halted by minimally invasive interventions targeting key therapeutic ROIs via surgical destruction or modulation without the need for craniotomy.⁴⁰

Limitations

This study has several limitations, notably its retrospective design, relatively small sample size, and single-center origin. As such, our findings may not be representative of the broader population of refractory MTLE patients.

Conclusion

This study provides valuable insights into the significance of the amygdala and rhinal cortex as ROIs in the preoperative evaluation of patients with MTLE. The future implantation scheme plans should consider evaluating the preoperative volume of these ROIs. Additionally, it might be beneficial to increase the number of electrode contacts implanted within these regions to capture more clinically relevant signals and increase their ablation volume.

Supplemental Material

sj-docx-1-tan-10.1177_17562864241286867 – Supplemental material for Optimizing outcomes in drug-resistant mesial temporal lobe epilepsy patients undergoing stereoelectroencephalography-guided radiofrequency thermocoagulation

Supplemental material, sj-docx-1-tan-10.1177_17562864241286867 for Optimizing outcomes in drug-resistant mesial temporal lobe epilepsy patients undergoing stereoelectroencephalography-guided radiofrequency thermocoagulation by Stéphane Jean, Rifeng Jiang, Yihai Dai, Weitao Chen, Weihong Liu, Donghuo Deng, Panashe Tevin Tagu, Xiaoqiang Wei, Shan Chen, Xinrong Fang and Shiwei Song in Therapeutic Advances in Neurological Disorders

Footnotes

Acknowledgements

Stéphane Jean: manuscript, figures, neuroimaging analysis, and interpretation. Rifeng Jiang: neuroimaging analysis and interpretation. Shiwei Song: SEEG analysis and study conception. Yihai Dai: data acquisition. Weitao Chen: gathering information on patients. Weihong Liu: statistical analysis. Donghuo Deng: follow-up. Panashe Tevin Tagu: manuscript proofreading. Shan Chen: EEG analysis. Xinrong Fang: SEEG analysis. Xiaoqiang Wei: EEG analysis.

Declarations

ORCID iDs

Stéphane Jean

Shiwei Song

Supplemental material

Supplemental material for this article is available online.

References

Shorvon

Goodridge

DMG

. Longitudinal cohort studies of the prognosis of epilepsy: contribution of the National General Practice Study of Epilepsy and other studies. Brain 2013; 136(11): 3497–3510.

Engel

Mesial temporal lobe epilepsy: what have we learned?

Neuroscientist 2001; 7(4): 340–352.

Laxer

Trinka

Hirsch

, et al. The consequences of refractory epilepsy and its treatment. Epilepsy Behav 2014; 37: 59–70.

Tatum

WO.

Mesial temporal lobe epilepsy. J Clin Neurophysiol 2012; 29(5): 356–365.

Berkovic

Andermann

Olivier

, et al. Hippocampal sclerosis in temporal lobe epilepsy demonstrated by magnetic resonance imaging. Ann Neurol 1991; 29(2): 175–182.

Rosenow

Presurgical evaluation of epilepsy. Brain 2001; 124(9): 1683–1700.

Guenot

Isnard

Ryvlin

, et al. SEEG-guided RF thermocoagulation of epileptic foci: feasibility, safety, and preliminary results. Epilepsia 2004; 45(11): 1368–1374.

Engel

Early surgical therapy for drug-resistant temporal lobe epilepsy. JAMA 2012; 307(9): 922.

Penfield

Temporal lobe epilepsy. Br J Surg 1954; 41(168): 337–343.

10.

Olivier

Transcortical selective amygdalohippocampectomy in temporal lobe epilepsy. Can J Neurol Sci 2000; 27(1): S68–S76.

11.

Olivier

Temporal resections in the surgical treatment of epilepsy. Epilepsy Res Suppl 1992; 5: 175–188.

12.

Engel

Jr Van Ness

Rasmussen

, et al. Outcome with respect to epileptic seizures. In: Engel

Jr (ed.) Surgical treatment of the epilepsies, 2nd ed. New York, NY: Raven Press, 1993, pp. 609–621.

13.

McIntosh

AM.

Temporal lobectomy: long-term seizure outcome, late recurrence and risks for seizure recurrence. Brain 2004; 127(9): 2018–2030.

14.

Brotis Alexandros

Giannis

Kapsalaki

, et al. Complications after anterior temporal lobectomy for medically intractable epilepsy: a systematic review and meta-analysis. Stereotact Funct Neurosurg 2019; 97(2): 69–82.

15.

Guenot

Isnard

Ryvlin

, et al. Neurophysiological monitoring for epilepsy surgery: the Talairach SEEG method. Stereotact Funct Neurosurg 2001; 77(1–4): 29–32.

16.

Quigg

Harden

CL.

Minimally invasive techniques for epilepsy surgery: stereotactic radiosurgery and other technologies. J Neurosurg 2014; 121(Suppl_2): 232–240.

17.

Catenoix

Mauguière

Montavont

, et al. Seizures outcome after stereoelectroencephalography-guided thermocoagulations in malformations of cortical development poorly accessible to surgical resection. Neurosurgery 2015; 77(1): 9–15.

18.

Bourdillon

Isnard

Catenoix

, et al. Stereo electroencephalography-guided radiofrequency thermocoagulation (SEEG-guided RF-TC) in drug-resistant focal epilepsy: results from a 10-year experience. Epilepsia 2016; 58(1): 85–93.

19.

Cossu

Fuschillo

Casaceli

, et al. Stereoelectroencephalography-guided radiofrequency thermocoagulation in the epileptogenic zone: a retrospective study on 89 cases. J Neurosurg 2015; 123(6): 1358–1367.

20.

Bonilha

Martz

Glazier

, et al. Subtypes of medial temporal lobe epilepsy: influence on temporal lobectomy outcomes? Epilepsia 2011; 53(1): 1–6.

21.

Pitskhelauri

Kudieva

Kamenetskaya

, et al. Multiple hippocampal transections for mesial temporal lobe epilepsy. Surg Neurol Int 2021; 12: 372.

22.

Johnson

Cai

Doss

, et al. Localizing seizure onset zones in surgical epilepsy with neurostimulation deep learning. J Neurosurg 2022; 138(4): 1002–1007.

23.

Suzuki

Sugano

Nakajima

, et al. The epileptogenic zone in pharmaco-resistant temporal lobe epilepsy with amygdala enlargement. Epileptic Disord 2019; 21(3): 252–264.

24.

Maillard

Vignal

Gavaret

, et al. Semiologic and electrophysiologic correlations in temporal lobe seizure subtypes. Epilepsia 2004; 45(12): 1590–1599.

25.

Kwan

Arzimanoglou

Berg

, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2009; 51(6): 1069–1077.

26.

Catenoix

Mauguiere

Guenot

, et al. SEEG-guided thermocoagulations: a palliative treatment of nonoperable partial epilepsies. Neurology 2008; 71(21): 1719–1726.

27.

Dale

Fischl

Sereno

MI.

Cortical surface-based analysis. NeuroImage 1999; 9(2): 179–194.

28.

Yeh

Zaydan

Suski

, et al. Differential tractography as a track-based biomarker for neuronal injury. NeuroImage 2019; 202: 116131.

29.

Bonilha

Yasuda

Rorden

, et al. Does resection of the medial temporal lobe improve the outcome of temporal lobe epilepsy surgery? Epilepsia 2007; 48(3): 571–578.

30.

Siegel

Wieser

Wichmann

, et al. Relationships between MR-imaged total amount of tissue removed, resection scores of specific mediobasal limbic subcompartments and clinical outcome following selective amygdalohippocampectomy. Epilepsy Res 1990; 6(1): 56–65.

31.

Keller

Richardson

Schoene-Bake

, et al. Thalamotemporal alteration and postoperative seizures in temporal lobe epilepsy. Ann Neurol 2015; 77(5): 760–774.

32.

Sagher

Thawani

Etame

, et al. Seizure outcomes and mesial resection volumes following selective amygdalohippocampectomy and temporal lobectomy. Neurosurg Focus 2012; 32(3): E8.

33.

Gleichgerrcht

Drane

Keller

, et al. Association between anatomical location of surgically induced lesions and postoperative seizure outcome in temporal lobe epilepsy. Neurology 2021; 98(2): e141–e151.

34.

Sun

Cui

, et al. Temporal lobe epilepsy with amygdala enlargement: a subtype of temporal lobe epilepsy. BMC Neurol 2014; 14(1): 194.

35.

Lee

Hong

, et al. Volume change in amygdala enlargement as a prognostic factor in patients with temporal lobe epilepsy: a longitudinal study. Epilepsia 2019; 61(1): 70–80.

36.

Shakhatreh

Sinclair

McLean

, et al. Amygdala enlargement in temporal lobe epilepsy: histopathology and surgical outcomes. Epilepsia 2024; 65(6): 1709–1719.

37.

Ballerini

Tondelli

Talami

, et al. Amygdala subnuclear volumes in temporal lobe epilepsy with hippocampal sclerosis and in non-lesional patients. Brain Commun 2022; 4(5): fcac225.

38.

Malikova

Kramska

Vojtech

, et al. Stereotactic radiofrequency amygdalohippocampectomy: two years of good neuropsychological outcomes. Epilepsy Res 2013; 106(3): 423–432.

39.

Liscak

Malikova

Kalina

, et al. Stereotactic radiofrequency amygdalohippocampectomy in the treatment of mesial temporal lobe epilepsy. Acta Neurochir (Wien) 2010; 152(8): 1291–1298.

40.

Royer

Bernhardt

Larivière

, et al. Epilepsy and brain network hubs. Epilepsia 2022; 63(3): 537–550.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB