Seizure: European Journal of Epilepsy 91 (2021) 447–455

Contents lists available at ScienceDirect

Seizure: European Journal of Epilepsy

journal homepage: www.elsevier.com/locate/seizure

Temporal lobe epilepsy: Evaluation of central and systemic

immune-inflammatory features associated with drug resistance.
Andrea Toledo a, b, Sandra Orozco-Suárez c, Marcos Rosetti d, e, Lorenzo Maldonado a,
Sara I. Bautista a, Ximena Flores a, Alfonso Arellano f, Sergio Moreno f, Mario Alonso f,
Iris E. Martínez-Juárez g, Gladis Fragoso h, Edda Sciutto h, Agnès Fleury a, i, *
a
Unidad Mixta de Estudio de la Neuroinflamación, Instituto de Investigaciones Biomédicas (IIBO), Universidad Nacional Autónoma de México (UNAM) / Instituto
Nacional de Neurología y Neurocirugía (INNN), Ciudad de México, México
b
División de Investigación, Facultad de Medicina, Universidad Nacional Autónoma de México (UNAM)
c
Unidad de Investigación Médica en Enfermedades Neurológicas, Hospital de Especialidades, “Bernardo Sepulveda”, CMN Siglo XXI, IMSS, Ciudad de México, México
d
Departamento de Biología Celular y Fisiología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México
e
Unidad Psicobiología del Desarrollo, Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz, Ciudad de México, México
f
Departamento de Neurocirugía, Instituto Nacional de Neurología y Neurocirugía (INNN), Ciudad de México, México
g
Clínica de Epilepsia, Instituto Nacional de Neurología y Neurocirugía (INNN), Ciudad de México, México
h
Departamento de Inmunología, Instituto de Investigación Biomédicas, Universidad Nacional Autónoma de México (UNAM), Ciudad de México, México
i
Departamento de Medicina Genómica y Toxicología Ambiental, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México Ciudad
Universitaria. Apartado Postal 70228. México DF 04510. México

A R T I C L E I N F O A B S T R A C T

Keywords: Neuroinflammation is probably one of the factors involved in drug resistance in people with epilepsy. Finding
Temporal lobe epilepsy peripheral markers reflecting the intensity of neuroinflammation could be of great help to decide for which
Drug-resistant epilepsy patients anti-inflammatory treatment might be an option. In this context, peripheral cytokines levels and
Neuroinflammation
lymphocyte phenotypes were assessed by ELISA and flow cytometry in 3 groups of subjects: drug resistant pa­
Peripheral inflammation
tients with temporal lobe epilepsy (DR-TLE), non DR-TLE patients and healthy controls. The same parameters
were assessed in brain tissue in the DR-TLE group. Differences in the peripheral immune-inflammatory status
between the 3 groups of subjects, and correlations between the central and peripheral immune-inflammatory
status in DR-TLE patients were evaluated.
Forty-one patients with DR-TLE, ten with non-DR-TLE and twenty controls were included. In the periphery,
decrease in regulatory cells were observed in DR-TLE patients compared to controls. In addition, significant
increase of IL-6 and IL-5 was observed in patients with epilepsy (particularly DR-TLE patients). Two groups of
DR-TLE patients with significant differences in several central inflammatory parameters were identified in a
cluster analysis. The inflammatory cluster was associated with a peripheral increase of CD4+CD38+ cells and
different significant correlations between central and systemic inflammatory parameters were observed.
Although their interpretation is not immediate, they demonstrate a clear dialogue between central and pe­
ripheral inflammatory reactions. In conclusion, our results add new elements to better understand the in­
teractions between the central and peripheral compartments in patients with DR-TLE, and to help better define
treatment options in this group of patients.

1. Introduction characterized by an “enduring predisposition to generate epileptic sei­

zures that are the transient occurrence of signs and/or symptoms due to
Epilepsy is a brain disorder that affects people of all ages. It is abnormal excessive or synchronous neuronal activity in the brain” [1,

Abbreviations: PWE, Patients with epilepsy; DR, Drug resistant epilepsy; TLE, Temporal lobe epilepsy; DR-TLE, Drug resistant temporal lobe epilepsy; non-DR-TLE,
Non drug resistant temporal lobe epilepsy; HC, Healthy controls; HS, Hippocampal sclerosis.
* Corresponding author.
E-mail address: afleury@iibiomedicas.unam.mx (A. Fleury).

https://doi.org/10.1016/j.seizure.2021.07.028
Received 25 March 2021; Received in revised form 13 July 2021; Accepted 21 July 2021
Available online 26 July 2021
1059-1311/© 2021 British Epilepsy Association. Published by Elsevier Ltd. This article is made available under the Elsevier license (http://www.elsevier.com/open-
access/userlicense/1.0/).
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

2]. Around 50 million people worldwide have epilepsy, making it one of Table 1
the most common neurological diseases. Nearly 80% of people with Antibody combinations used for flow cytometry in PMBC cells and cerebral
epilepsy live in low and middle-income countries [3]. cortex cells
Drug resistant epilepsy (DR) is defined as the “failure of adequate Antibodies Characteristic
trials of two tolerated, appropriately chosen and used antiepileptic drug CD4+ (APC) Total T helper cells
schedules (whether as monotherapies or in combination) to achieve CD8+ (FITC) Total cytotoxic T cells
sustained seizure freedom” [4]. Its prevalence among epilepsy patients CD4+ APC/CD8+ FITC/CD69+ PE Early activated
is considered to be 30% (95% [CI] 19-42%) [5]. In these cases, the CD4+APC/CD8+ FITC/CD38+ PerCP Late activated
CD4+ FITC/CD8+ PE /CD95+ APC Apoptosis
therapeutic methods currently available to reduce or stop seizures
CD4+/CD25+/FoxP3+ (FITC, APC, PE, Regulatory T cells
include surgery, neurostimulation, ketogenic diet and immunotherapy respectively)
in autoimmune epilepsy [6, 7]. CD4+/CD25+/FoxP3+/CD45RO+ (FITC, APC, PE, Activated regulatory T cells
In a large study that evaluate the types of epilepsy mostly associated PerCP, respectively)
with DR in patients that underwent epilepsy surgery, hippocampal CD3− /+ FITC/CD16+ PE/CD56+ PerCP NK/NKT cells
CD19+ PE/ IL10+ PerCP B cells expressing IL10
sclerosis (HS) was the first pathological diagnosis (36.4% of the pa­ CD11b+ FITC/CD45 PE Activated microglia
tients), followed by tumors (mainly ganglioglioma, 23.6% of the pa­ CD68+ Macrophages and many
tients), and malformations of cortical development (mainly focal microglial cell
cortical dysplasia, 19.8% of the patients) [8]. Temporal lobe (TLE) was CD20+ B lymphocytes
involved in 71.9% of epilepsy surgeries [8]. FITC: fluorescein isothiocyanate; PerCP: peridinin chlorophyll; APC: allophy­
On the other hand, neuroinflammatory pathways are known to cocyanin; PE: phycoerythrin.
contribute to the development and progression of seizures and epilepsy
[9]. Inflammation is the normal response of an organism to damage, by direct interview and revision of medical charts.
helping in its reparation, and that is also the case in the Central Nervous
System, by balancing the enhanced metabolic demand during increased
neuronal activity. However, if neuroinflammatory response becomes 2.2. Systemic inflammation evaluation
chronic or very intense, it can be associated to cellular dysfunction, as in
the case of epilepsy and other neurological diseases [9]. Before obtaining blood and surgery samples all patients signed a
It is known that links between systemic and neuroinflammatory consent form.
status exist. Systemic inflammation can clearly modulate neuro­ A 20 ml venous blood sample drawn from the subjects was obtained
inflammatory condition in different diseases, particularly epilepsy [10, in tubes coated with ethylenediaminetetra-acetic acid (EDTA) (BD
11], and on the other hand neuroinflammation inducing blood-brain Vacutainer). The samples were kept at room temperature and processed
barrier leakage, the peripheral inflammatory status should reflect within 2 hours. The PMBCs were isolated in a Ficoll / Hypaque gradient.
some characteristics of the central inflammatory status. Peripheral Additionally, a serum sample was obtained and immediately frozen until
biomarkers of neuroinflammation have been proposed, [12] although a use (cytokines titration).
clear comparison between central and systemic inflammation is still For DR-TLE patients, extraction of blood was made 24 hours before
lacking. neurosurgery.
Considering the background, this exploratory study aims to compare Due to logistical constraints, the time of sampling and the fasting
the systemic inflammatory state of patients with DR-TLE, non-DR-TLE conditions were not homogeneous between the subjects included.
and healthy controls, and to evaluate the eventual associations between
neuroinflammatory features and systemic inflammatory features in DR- 2.2.1. Peripheral cells phenotype (Flow cytometry)
TLE patients. The immunophenotype of peripheral blood mononuclear cells
(PBMCs) was studied following the procedure previously published
[13].
2. Methods
The human monoclonal antibodies used for leukocyte phenotyping
by flow cytometry are shown in Table 1.
2.1. Patients
All antibodies were titrated for optimal detection of positive pop­
ulations prior to use, following the manufacturer’s (BioLegend) recom­
Three groups of subjects were included: patients with drug-resistant
mended concentrations. Antibody combinations were used to define
temporal lobe epilepsy (DR-TLE), patients with temporal lobe epilepsy
different lymphocyte populations (Table 1). Appropriate isotype con­
responders to treatment (non-DR-TLE), and healthy controls (HC).
trols were included. A fluorescence-activated cell sorter (FACS) Calibur
Inclusion criteria common for these three groups were the absence of
cytometer was used for data acquisition and analyzed with the CellQuest
inflammatory, tumoral and metabolic uncontrolled diseases, the
software.
absence of other neurological conditions, and general laboratories
within normal limits. Patients with epilepsy (PWE) included had all
2.2.2. Cytokines (ELISA)
confirmed diagnosis of temporal lobe epilepsy based on clinical, radio­
Commercial sandwich enzyme linked immunosorbent assay (ELISA)
logical and electrophysiological criteria. Criteria for DR were adopted
kits were employed to quantify in serum and in tissue extracts the pro
from the 2010 International League Against Epilepsy (ILAE) consensus
and anti-inflammatory human cytokines IL-1β, IL-5, IL-6, IL-17A, CCL5,
statement, [4] and non DR patients were considered if they had less than
INFγ and TNF (all from BioLegend, San Diego, CA, USA), following the
one seizure per year (either focal or focal to bilateral tonic-clonic) or had
supplier’s instructions. Sandwich ELISAs were performed in 96-well,
been seizure free for at least six months.
flat-bottom microtiter plates (Nunc-Immuno Plate Maxisorp). The
All the patients included attended the epileptic clinic at the National
detection limits were 3.9 pg/ml for IL-1β, 7.8 pg/ml for IL-5 and IL17-A,
Institute of Neurology and Neurosurgery (INNN) in Mexico City, and in
15.6 pg/ml for IL-6, CCL5, IFN- γ and TNF-α. All analyses were run in
cases of DR-TLE patients, decision of surgical treatment was taken at a
duplicate.
Surgical Epilepsy Session. All of them have more than 15 years as only
patients of this age or above are attended at INNN. This project was
approved by the Research and Bioethics institutional boards, and all the 2.3. Central inflammation evaluation
subjects included gave their informed consent before inclusion.
Demographic, clinical and radiological information were recovered A sample of brain tissue (cerebral cortex and hippocampus) from DR-

448
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

TLE patients who underwent surgery was recovered. The surgeries Table 2
consisted of an anterior temporal lobectomy or a selective amygdalo­ Main characteristics of subjects included
hippocampectomy. On the hippocampus, we only performed patholog­ Temporal lobe epilepsy Healthy P
ical analyzes because the small sample size did not allow us to do all the (n= 51) controls (n=
planned evaluations. Three types of analysis were made: 20)
DR-TLE Non DR-
(n=41) TLE (n=
2.3.1. Pathological analysis 10)
A fragment from each region of the temporal lobe (cerebral cortex
Gender (M/F) 18/ 23 4/6 8 / 12 0.95
and hippocampus) was fixed in 4% paraformaldehyde diluted in phos­ Age at inclusion (mean years 39.1 ± 40.0 ± 37.1 ± 10.4 0.75
phate buffer, samples were included in paraffin and an adjacent block ± SD) 10.6 12.2
was frozen after protection with isopentane solution (CAS: 78-78-4, Education less or equal to 9 24 (58.5) 3 (30) 8* (47.1) 0.25
Sigma Aldrich) and cut to -20 ◦ C using a cryostat (CM-1510,Leica Bio­ years (n, %)
Age at the beginning of 13.9 ± 15.3 ± 7.5 0.68
systems) to be processed by immunofluorescence.
seizures (mean years ± SD) 9.7
Type of seizures (n, %)
2.3.2. Histological analyses • Focal onset without 4 (9.8) 1 (10) 0.98
Paraffin-embedded brain tissues were sectioned coronally using a impaired awareness
microtome at a thickness of 6 μm (RM2125RT, Leica Biosystems) and • Focal onset with impaired 26 (63.4) 2 (20) 0.01
awareness
stained with hematoxylin and eosin (H & E), for an initial evaluation to • Generalized onset 12 (29.3) 2 (20) 0.56
define cytoarchitectural and pathologic features. Immunoperoxidase Antiseizure drugs (n, %)
technique was carried in these same samples for identifying neurons, • Monotherapy 4 (9.8) 1 (10) 0.98
activated microglia and gliosis. The sections were deparaffinized and • Levetiracetam 23 (56.1) 3 (30) 0.14
• Carbamazepine 24 (58.5) 2 (20) 0.03
hydrated to be processed according to previous studies by immunoper­
• Sodium valproate 16 (39.0) 6 (60) 0.23
oxidase for paraffin sections [14]. • Lamotrigine 12 (29.3) 3 (30) 0.96
• Oxcarbazepine 7 (17.1) 1 (10) 0.58
2.3.3. Immunofluorescence techniques • Phenytoin 5 (12.2) 2 (20) 0.52
12 um of cryostat sections were hydrated in PBS and then incubated Body Mass Index (Mean ± 26.5 ± 27.2 ± 4.2 0.68
SD) 4.9**
with 1% normal horse serum in PBS 0.12 M. Later, the sections were
• Underweight (n, %) 1 (2.5) 0 1
incubated with primary antibodies diluted in PBS as follows: CD68+, • Normal weight (n, %) 17 (42.5) 3 (30) 0.72
CD20+, CD3+, CD4+, CD8+, for 24 h at 4◦ C (Santa Cruz Biotech­ • Overweight (n, %) 13 (32.5) 4 (40) 0.72
nology). Then, the sections were washed with PBS 0.12 M and incubated • Obesity (n, %) 9 (22.5) 3 (30) 0.68
Medical comorbidities (n, %)
2 h at 4◦ C with the appropriate secondary antibodies diluted in PBS as
• Depression / anxiety 14 (34.1) 3 (30) 1
follows: a) goat anti- mouse IgG Alexa-488 or goat-anti-rabbit IgG Alexa • Arterial hypertension 3 (7.3) 0 1
488 (BioLegend) according to the primary antibody, following this, the • Diabetes Mellitus 2 (4.9) 0 1
sections were washed and stained with propidium iodide (P-3566, • Calcified Neurocysticercosis 2 (4.9) 1 (10) 0.49
Invitrogen, Thermo Fisher Scientific) and finally, washed and cover- • Dyslipidemia 1 (2.4) 1 (10) 0.36
• Anemia 1 (2.4) 0 1
slipped with Vectashield (H-1000, Vector Laboratories, Inc.) mounting
Associated treatment (n, %)
medium. • Antidepressant /anxiolytic 12 (29.3) 3 (30) 1
The images from fluorescence sections were observed on a Nikon Ti • ACE inhibitors 3 (7.3) 0 1
Eclipse inverted confocal microscope equipped with an A1 imaging • Metformin 2 (4.9) 0 1
• Statins 1 (2.4) 1 (10) 0.36
system controlled with the proprietary software (NIS Elements v.4.50).
• Erythropoietin 1 (2.4) 0 1
Imaging was performed using either a 20x (dry, NA 0.8) or a 60x (oil
immersion, NA 1.4) as specified in the text. Dyes were excited in a DRE: Drug resistant epilepsy / * the information was not available in 3 subjects.
sequential mode using the built-in laser lines. Corresponding fluores­ / ** the information was not available in 1 patient.
cence was read in the following ranges: 425-475 nm for Alexa 486, 535- Student’s t-test was used to evaluate differences between means. Fisher’s exact
test was used to evaluate differences in categorical variables.
617 nm for propidium iodide. Images were acquired and analyzed using
NIS Elements v.4.50 and ImageJ v.1.50i software, respectively.
The number of cells immunoreactive to each antibody were counted 2.3.5. Cytokines
in six serial sections every 100 microns in 330 square micron fields and During surgery, a small piece of cerebral cortex tissue (0.7 g) was
adjusted to a millimeter close to the blood vessels and the meningeal placed in a sterile tube and immediately put in liquid nitrogen. Brain
surface in the case of the cerebral cortex. proteins were extracted as Schieb et al., 2011 [15]. Briefly, brain sam­
ples were weighed and homogenized by sonication in 1 ml per 100 mg
2.3.4. Brain cells phenotypes evaluated by flow cytometry wet tissue of a triple detergent buffer (50 mM HEPES, pH 7.4, 150 mM
Approximately 0.5 to 2 g of cerebral cortex tissue were removed by NaCl, 1.0% Nonidet P-40 (Igepal CA 630)) (v/v), 0.5% sodium deoxy­
surgery, kept on ice and homogenized in 3 volumes of GKN buffer (PBS cholate (w/v), 0.1% SDS (w/v), and Complete protease inhibitor
1X, supplemented with Glucose at 0.02%). Using a cell striner (100 μm), mixture (Roche Applied Science), 1 tablet per 10 ml. After centrifuga­
the tissue was macerated obtaining a homogeneous suspension of cells tion (16,000 × g, 15 min, 4◦ C), the supernatant was collected, and the
which was collected and centrifuged for 10 minutes at 400 g at room protein concentration was determined by the BCA assay (Thermo
temperature. The pellet was resuspended in a digestion buffer which Scientific).
contained 500 U of DNAsa I and 15 U of collagenase II and then was The same ELISA kits and methodology than those used for peripheral
incubated at 37◦ C for 1 hour. The reaction was stopped using 45 ml of evaluation were employed. Assay sensitivity was measured in pg/mg of
GKN-BSA buffer at 4◦ C. After 10 minutes of centrifugation at 400 g, the protein.
cells first were resuspended in a 37% Percoll gradient and subsequently
separated using 30% and 70% percoll gradients. The gradient was 2.4. Statistical analysis
centrifuged for 40 minutes at 1200 rpm. After this period, the cells were
collected and marked by flow cytometry as previously described [13]. The data was collected and organized in Excel (Microsoft) software
and analyzed with SPSS15.0 (SSPS Inc.) and GraphPad Prism 6.0

449
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

Table 3
Percentages of lymphocyte phenotypes and cytokine concentrations in peripheral blood of the people with epilepsy and healthy controls
Healthy controls (HC) (n= 20) Non DR-TLE (n= 10) DR-TLE (n= 41) P1 P2 P3

Lymphocyte phenotypes
CD4+ 39.16 (±17.0)* 38.06 (±12.4) 37.36 (±17.2) 0.92 0.70 0.72
CD8+ 14.11 (±8.2) 9.51 (±6.7) 10.17 (±8.3) 0.17 0.06 0.23
CD8+CD95+ 21.99 (±12.5) 20.95 (±11.3) 16.71 (±11.6) 0.23 0.16 0.09
CD4+CD95+ 15.87 (±9.7) 9.01 (±5.90) 14.43 (±12.0) 0.25 0.38 0.74
CD19+IL-10+ 5.15 (±5.9) 5.76 (±4.8) 2.57 (±4.2) 0.07 0.15 0.03
CD3+CD16+CD56+ 2.70 (±3.5) 2.72 (±4.2) 2.44 (±4.3) 0.96 0.85 0.78
CD3-CD16+CD56+ 15.01 (±8.3) 15.14 (±11.5) 12.94 (±11.5) 0.72 0.57 0.42
CD8+CD69+ 6.45 (±7.1) 4.26 (±7.2) 4.07 (±6.0) 0.40 0.17 0.30
CD8+CD38+ 2.56 (±2.7) 1.82 (±2.1) 1.55 (±2.3) 0.32 0.14 0.20
CD4+CD69+ 2.91 (±3.0) 2.22 (±3.2) 2.26 (±3.4) 0.75 0.44 0.59
CD4+CD38+ 1.42 (±1.2) 0.97 (±0.9) 1.87 (±4.0) 0.69 0.75 0.43
CD4+CD25+FoxP3+ 2.20 (±5.1) 0.28 (±0.5) 0.26 (±0.4) 0.04 0.11 0.11
CD4+CD25+FoxP3+CD45RO+ 2.30 (±4.3) 0.74 (±0.9) 0.47 (±0.6) 0.02 0.08 0.06
Cytokine concentrations (pg/mL)
TNF 2.28 (±8.0) 42.0 (±98.8) 1.93 (±6.9) 0.01 0.46 0.21
CCL5 509.5 (±78.8) 554.2 (±93.2) 498.8 (±74.8) 0.15 0.86 0.28
IL-17A 0.02 (±0.08) 0.81 (±2.4) 0.39 (±1.2) 0.27 0.04 0.75
INF-γ 9.30 (±12.9) 20.12 (±30.3) 19.53 (±36.3) 0.45 0.08 0.41
IL-1β 0.58 (±1.4) 0.98 (±2.1) 0.31 (±0.9) 0.33 0.67 0.23
IL-6 21.11 (±62.4) 43.96 (±136.1) 89.79 (±171.8) 0.21 0.04 0.07
IL-5 0.17 (±0.8) 0.81 (±1.9) 2.97 (±5.7) 0.06 0.004 0.01

*Mean (±Standard deviation)

P1: comparison between the 3 groups, P2: PWE (NoDR-TLE + DR-TLE) vs. healthy controls, P3: DR-TLE vs NoDR-TLE + Healthy controls
Comparisons between groups were made using ANOVA and Tuckey post-hoc test

software (GraphPad Software). The statistical comparisons between decrease of CD19+IL-10+ in DR-TLE patients compared to healthy
variables were performed using parametric or non-parametric tests, controls and non-DR-TLE (p= 0.03) and a significant difference of reg­
depending on the normality of the data. Comparisons of categorical ulatory T cells, both activated and non-activated, between the 3 groups
variables were made using Chi-squared or Fisher’s exact tests. Com­ of patients (p= 0.04 and p= 0.02, respectively). The decrease of both
parisons of means were made using ANOVA and Tukey post-hoc test, types of regulatory T cells was noted in the two groups of PWE, and
Student’s t-test or Mann-Whitney test. Paired correlations between the presented statistical significance between HC and DR-TLE patients (data
numerical values were evaluated using Pearson’s or Spearman’s test. As not shown, p= 0.04 and p= 0.02, respectively). Cluster analysis was not
our study was exploratory, we did not make any corrections for the able to define markers that adequately separate the 3 groups of included
multiple comparisons [16]. The statistical software R (R Core Team) was subjects.
used to perform a clustering analysis on the data. For this, we used the k
means clustering algorithm, an unsupervised machine learning tech­ 3.1.2. Peripheral cytokines in the three groups of subjects included
nique by which are data points can be grouped into several predefined, Results are presented in Table 3. As seen, in all the groups the
non-overlapping subgroups. Data were scaled before any clustering was cytokine with higher concentration was CCL5. IL-5 concentration was
performed. Elbow plots generated by the function fviz_nbclust() from the significantly higher in PWE vs. HC (p= 0.004), in DR-TLE vs. the other 2
package factoextra from the help choose the optimal number of clusters. groups (p= 0.01), and in DR-TLE patients vs. HC (p= 0.06, data not
shown). IL-6 was significantly higher in PWE and in DR-TLE patients
3. Results compared to HC (p= 0.04 and p=0.08 respectively, data not shown).
TNF was significantly different between the 3 groups (p= 0.01), due to
Seventy-one subjects were included: 41 patients with DR-TLE (27 of its increase in non-DR-TLE patients compared to DR-TLE patients and
them underwent epilepsy surgery), 10 with non-DR-TLE, and 20 HC. HC (p= 0.02 and p= 0.01 respectively, data not shown). IL-17A con­
Demographic characteristics were similar between the 3 groups centration was significantly higher in PWE compared to HC (p= 0.04).
(Table 2). By comparing the two groups of PWE, DR-TLE patients had Cluster analysis was not able to define markers that adequately separate
focal onset with more frequent impaired awareness seizures (p= 0.01), the 3 groups of included subjects.
and most were on carbamazepine (p = 0.03). No significant differences
between the two groups of PWE were observed regarding body mass 3.1.3. Correlations between peripheral cellular phenotypes and cytokines in
index, medical comorbidities and their treatments (Table 2). Treatment DR-TLE patients
with LVT was associated with an increase of peripheral The following significant correlations were observed: IL-6 was
CD4+CD25+FoxP3+CD45RO+ lymphocytes (P= 0.04) and treatment negatively associated with percentages of CD4+CD38+ (R= - 0.39; p=
with Oxcarbazepine was associated with an increase of central 0.01), CD4+CD25+FoxP3+CD45RO+ (R= -0.40; p= 0.01), and CD3-
CD4+CD25+FoxP3+ lymphocytes (p=0.02). The time between the last CD16+CD56+ (R= -0.36; p=.0.02).
seizure and sampling / surgery was variable (1 day - 1 month) in DR-TLE IL-1 was positively associated with percentage of
patients. Only 2 patients had presented a seizure within 24 hours before CD4+CD25+FoxP3+ (R= 0.33; p= 0.04) and percentage of
surgery and in both cases, it was a focal onset impaired awareness CD8+CD38+ (R= 0.37; p= 0.02).
seizure. CCL5 was positively correlated with percentage of
CD4+CD25+FoxP3+CD45RO+ (R= 0.40; p= 0.01).
3.1. Peripheral immune-inflammatory status
3.2. Central inflammatory status in DR-TLE patients
3.1.1. Peripheral cellular phenotypes in the three groups of subjects
included 3.2.1. Histological and Immunocytochemical results
Results are shown in Table 3. The main results were a significant We evaluated with these techniques 27 samples of cortex and 24

450
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

Fig. 2. Number of immunoreactive cells per mm2 in the cerebral cortex and in
the hippocampus.

samples of hippocampus.
Of 27 cases in which the temporal cortex was obtained, 22.8% (n=6)
had dual pathology, i.e. focal cortical dysplasia (FCD) identified by
marking the nuclei with NeuN (1 IB, 1 IIA, 1 IIB, and 3 IIIB) associated
with hippocampal sclerosis. Two of the cases classified as FCD IIIB were
related to a low-grade tumor: One neuroglial tumor located in the
frontal region, and one low-grade astrocytoma located in the temporal
cortex, with less than 1% of glial proliferation (Ki -67) and not notice­
able by neuroimaging. These patients were excluded from all analyzes to
avoid bias related to possible immunological alterations associated with
the tumors.
All the 24 revised hippocampus showed signs of hippocampal scle­
rosis; of type I in 19 cases (79.2%) and of type 3 in one case (4.16%),
according to the International consensus classification [17]. In 4 pa­
tients (16.66%) the type of sclerosis was not possible to determine
because the amount of tissue that was provided by the neurosurgeons
Fig. 1. Photomicrographs showing both in the cortex (A) and in hippocampus was not enough to identify the pattern of cell loss. In addition to these
(F) a low transformation of microglia (Iba) in the perivascular area. The pres­ findings, the temporal lobe histopathological changes included moder­
ence of T lymphocytes (CD3 +) is observed on the meningeal surface (B) ate reactive astrogliosis pattern (n=8) in which the bodies of astrocytes
extravasating the cerebral parenchyma in the superficial layers (arrow). In the with their extensions were observed, and severe fibrillar astrogliosis
hippocampus CD3+ cells are observed in the vessels and microvessels (G). In (n=16) with a dense mesh of fine processes marked with GFAP in a
the cerebral cortex CD8 + cells are more abundant in the first layers and on the sclerotic region.
meningeal surface (C). In the hippocampus they are mainly observed in the In the cerebral cortex the presence of activated microglia (IBA-1) was
vessels (H). CD20 + cells are located in the deepest layers of the cerebral cortex
not homogenously distributed and mainly expressed in the perivascular
(D). In the hippocampus (I) these cells are mainly perivascular with few cells
area, in layers I-II (Fig 1A and Fig 2). CD68+ cells and lymphocytes were
extravasated to the parenchyma. CD4 + lymphocytes in the cortex (E) are
located in the perivascular area, while in the hippocampus the extravasation to located mainly in the meningeal surface and in the proximity of vessels,
the parenchyma is higher (J). GL:granular layer with low extravasation to the cortex. The main phenotype was CD3+,
observed in the superficial layers and near vessels (Fig. 1B and Fig. 2).
The same distribution was observed in CD8+ and CD4+, with CD8+
cells being more numerous than CD4+ lymphocytes (Fig 1C and E, Fig
2). CD20+ lymphocytes were particularly located in the nearest of the

451
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

Table 4 was also negatively associated with CD3-CD16+CD56+ in periphery

Percentage of main cellular populations and cytokine concentration in cerebral (R= -0.42; p= 0.049).
cortex from DR-TLE patients When evaluating associations between immunofluorescent results
N Mean ± standard deviation and peripheral ELISA, positive correlations between CD20+ cortex and
Cellular phenotypes
IL-6 in sera (R=0.45, p=0.03), and between CD68+ hippocampus and
CD4+ 13 2.23±1.94 IL-1 sera (R=0.54, P=0.01) were observed.
CD8+ 13 1.52±1.38
CD8+CD95+ 14 0.74±0.95 3.3.2. Central cytokines and its correlations with peripheral cytokine and
CD4+CD95+ 14 0.87±1.43
peripheral cellular phenotypes in DR-TLE patients
CD3+CD16+CD56+ 14 2.99±5.78
CD3-CD16+CD56+ 14 3.08±8.32 IL-1β in cerebral cortex was negatively associated with concentration
CD4+CD25+FoxP3+ 12 0.74±1.03 of IL-6 in periphery (R= - 0.50; p= 0.02).
CD11B / CD45 12 0.52±0.66 Presence of IL-1β in cortex was positively associated with presence of
Cytokine Concentration (pg/mg of proteins) CD3+CD16+CD56+ in periphery (R= 0.44; p= 0.04). On the other
IL-1β 24 77.8±47.62
IL-6 24 48.97±49.65
hand, presence of IL-6 in cortex was negatively correlated with
CCL-5 24 19.92±11.07 CD4+CD25+FoXP3+CD45RO+ in periphery (R=-0.42, p= 0.049).
IL-17A 24 17.79±38.66
INF-γ 24 1.81±2.82 3.3.3. Cluster analysis
IL-5 24 0.32±0.38
We evaluate if clusters could be defined using the 16 central samples
TNF-α 24 0.00
in which complete pathological and cytokines analysis were performed.
Two clusters were made apparent: one with 7 and one with 9 patients,
damaged areas (Fig 1D). (Fig. 3). These two clusters of patients significantly differ regarding the
In the hippocampus the presence of IBA-1 cells was lower and pref­ following variables: CD3+, CD4+, CD8+ in cortex (p=0.004, p=0.04,
erentially located in the proximity of vessels (Fig 1F, Fig 2B). Occurrence p=0.01, respectively), CD3+ in hippocampus (p=0.02), IFNγ in cortex
of the other cellular phenotypes (CD3+, CD4+, CD8+, CD68+, CD20+) (p=0.03), and IL-5 in cortex (p=0.001). All these parameters, apart from
was mainly observed in the vessels with low extravasation (CD3+, Fig IL-5, were increased in cluster 2 while IL-5 was significantly higher in
1G-J, Fig 2B). CD68+ cells were the ones with the highest rate of peri­ cluster 1. We further assessed whether these groups, defined by central
vascular extravasation. results, differed significantly in terms of peripheral markers. Only a
tendency at the limit of statistical significance showed an increase in
3.2.2. Flow cytometry and ELISA CD4 + CD38 + lymphocytes in cluster 2 (p=0.06).
In Table 4, the different cellular populations evaluated in the cere­
bral cortex are shown. As seen, CD4+ and CD3+/-CD16+CD56+ were 4. Discussion
the more abundant whereas macrophages/microglia (CD11b/CD45)
were the lower. Regarding cytokines, IL-1 was the cytokine with higher Drug resistant epilepsy is a worldwide health problem, affecting
concentration in cerebral cortex, followed by IL-6, CCL-5 and IL-17A around 30% of PWE. Although epilepsy surgery is one of the solutions
(Table 4). for some of these patients, a number of them are not candidates to this
A positive correlation between IL-6 and CD4+CD25+FoxP3+ in management, and others do not have access to this type of treatment due
cortex (R= 0.75, P= 0.01) was observed. to the lack of hospital infrastructure, among other barriers [18].
For these patients controlling neuroinflammation could be one of the
3.3. Correlation between central and peripheral inflammatory status in ways to improve their seizure control [19, 20]. In experimental epilepsy
DR-TLE patients (A summary of the significant correlations is presented on it has been demonstrated that targeting specific inflammatory signaling
Table 5 and Graphical abstract) pathways favorably modify the disease course [21]. There are few
human studies to evaluate anti-inflammatory treatments and results
3.3.1. Central cellular phenotypes and its correlations with the peripheral seem to be very heterogeneous between patients, some of them
cellular phenotypes and cytokine profile in DR-TLE patients improving while others don’t.
A significant positive correlation was observed between CD4+ in This is probably due to the variability of neuroinflammation quality
cortex and peripheral CD4+CD25+FoxP3+CD45RO+ (R= 0.56, P= and intensity between patients. In this context, having peripheral
0.04). On the other hand, presence of CD4+CD25+FoxP3 in cortex was markers reflecting the central inflammatory phenomenon would be an
negatively associated (R= -0.65, p=0.02) with peripheral important and useful tool to help individualize the therapeutic decision
CD3+CD16+CD56+. [21]. Our study is framed in this context.
Regarding correlations between central cellular phenotypes and Patients with epilepsy, and most importantly those with DR-TLE,
peripheral cytokines, significant negative correlations were observed present significantly increased peripheral levels of IL-5. Interestingly,
between the presence of CD11B/CD45 in cortex and both peripheral treatment with sodium valproate in children with epilepsy has been
concentration of TNF (R= -0.67, p=0.02) and peripheral concentration associated with an increase of IL-5 [22]. This observation probably
of IFNγ (R=-0.74, p=0.009). Concentration of CD4+ in cortex was contributes to our findings since more than 40% of patients were treated
negatively associated with peripheral concentration of IL-5 (R= - 0.71, with this drug at the time of sampling. In addition, although this is a
p= 0.007), while presence of CD68+ cells in hippocampus was posi­ classical Th2 cytokine, which specially increase in some allergic condi­
tively correlated with peripheral concentration of IL-5 (R= 0.54, tions, it has been found to be elevated in other diseases as well such as in
p=0.01). major depressive disorders, although the exact biological mechanisms
Different negatives correlations were observed when evaluating through which IL-5 is linked to depression are not yet known [23].
central immunofluorescent and peripheral flow cytometry results: Interestingly, IL5 uses the Ras GTPase-extracellular signal-regulated
Presence of IBA in cortex was negatively associated with peripheral kinase (Ras-ERK) pathway, which has been found to control neural stem
CD4+CD95+ (R= -0.43, p= 0.04), CD4+ cells in cortex was negatively cell functions and maintains specific brain structures [24]. Thus, its
associated with both peripheral CD3+CD16+CD56+ (R= -0.47; increase in those patients resistant to treatment may imply the
p=0.02) and peripheral CD8+CD69+ (R= -0.46, p=0.03); and CD20+ involvement of brain tissues to restore the homeostasis condition.
in cortex was negatively associated with peripheral CD8+CD95+ (R= Peripheral levels of inflammatory cytokines (IL-6 and IL-17) were
-0.47; p= 0.02). Additionally, presence of CD68+ cells in hippocampus increased in PWE compared to HC. These results are similar to those

452
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

Fig. 3. The two clusters proposed by the k means algorithm. Patients (n=16) are shown by small, numbered points in the plot. Polygon lines mark the cluster edges
and the unnumbered shape (a circle and a triangle) mark each clusters’ centroid. To represent data in a 2d space, the set of central pathological and cytokines markers
were reduced by means of a principal component analysis.

described in previous studies [25, 26, 27] and likely reflect low-grade expected in a normal brain, lymphoid populations were present. Indeed,
peripheral inflammation in PWE. both CD4 and CD8 T lymphocytes, NK / NKT cells and activated mac­
Peripheral changes in cellular phenotypes were also observed; in rophages were observed, albeit in low percentages. It is interesting to
particular, a significant decrease of CD19/IL10 in DR-TLE patients note that these observations were made in the hippocampus but also in
compared to HC + non DR-TLE patients (p= 0.03), and a decrease in the cortex, confirming the spread of the inflammatory process as re­
activated and non-activated regulatory T cells in DR-TLE patients ported by positron emission tomographic imaging [28]. This lympho­
compared to HC (p= 0.02 and p= 0.04, respectively). The decrease in cytic infiltration, that reflect alterations in the blood-brain barrier and a
the levels of regulatory B and T cells in patients with DR-TLE epilepsy certain degree of inflammation in the cerebral cortex tissues, could be
indicates a deficiency in one of the mechanisms critically involved in the related to drug-resistance. Considering these results, it could be inter­
control of inflammation that could result in an increase of it. These re­ esting to evaluate whether the control of neuroinflammation results in
sults are different from those described in a previous study in which an clinical improvement in these patients.
increase in Tregs cells was observed in TLE patients compared to control The cell populations documented in the temporal lobe were pre­
[25] but are similar to those described in a pediatric study including dominantly lymphocytic with a preponderance of CD3+ positive T cells
patients with intractable epilepsy [26]. Thus, there seems to be a failure and smaller numbers of both cytotoxic and CD4+ T cells as well as
in the complex and variated regulatory mechanisms that may be asso­ CD20+ B cells. This observation reflects the chronic inflammation, as
ciated with resistance to treatment rather than epilepsy per se. observed previously [29,30,31,32] especially in patients with classical
In agreement with the latter finding, central inflammation was HS (ILAE Type I) [33]. The extent to which leukocytes extravasate into
documented in DR-TLE patients. Particularly, and opposite to what it is the brain parenchyma and play a role in seizures remains unclear. Initial

453
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

Table 5
Main significant correlations between central and peripheral immunological features in DR-TLE patients
Central Peripheral Direction of correlation R, p

Flow cytometry
1. CD4+ CD4+CD25+FoxP3+CD45RO+ Positive 0.56, 0.04
2. CD4+ IL-5 Negative -0.71, 0.007
3. CD11B/CD45 TNF Negative -0.67, 0.02
4. CD11B/CD45 IFNγ Negative -0.74, 0.009
5. CD4+CD25+FoxP3+ CD3+CD16+CD56+ Negative -0.65, 0.02
ELISA
6. IL-1β IL-6 Negative -0.5, 0.02
7. IL-1β CD3+CD16+CD56+ Positive 0.44, 0.04
8. IL-6 CD4+CD25+FoxP3+CD45RO+ Negative -0.42, 0.049
Immunocytochemical
9. IBA CD4+CD95+ Negative -0.43, 0.04
10. CD4+ CD3+CD16+CD56+ Negative -0.47, 0.02
11. CD4+ CD8+CD69+ Negative -0.46, 0.03
12. CD20+ CD8+CD95+ Negative -0.47, 0.02
13. CD68+ (Hippocampus) CD3-CD16+CD56+ Negative -0.42, 0.049
14. CD68+ (Hippocampus) IL-5 Positive 0.54, 0.01
15. CD68+ (Hippocampus) IL-1β positive 0.54, 0.01
16. CD20+ IL-6 Positive 0.45, 0.03

Paired correlations were evaluated using Pearson t test or Spearman’s test depending on the normality of the data.

descriptions suggested only limited leukocytic infiltration of the peri­ neuroinflammatory reaction. Several drugs are evaluated in clinical
vascular space, resulting from seizure-induced recruitment of the studies [9], and the obtention of peripheral markers of neuro­
adaptive immune system and simple leakage through a damaged inflammation should be a great contribution to better orient adminis­
blood-brain barrier [34]. tration of such treatment.
Regarding the detection of inflammatory cytokines in brain tissues, We are aware that our study has certain limitations. In particular, the
the presence of IFNγ, IL-17 and CCL-5 (which are not expected to be lack of a control group for the brain samples and the partial assessment
present under homeostatic conditions) support the notion of inflam­ of the hippocampus must be corrected in further studies. The small
matory process appearing in the cerebral cortex; also IL-17 and CCL-5 number of non-DR-TLE patients included is also a problem. In addition,
cytokines could be involved in the cellular recruitment of lymphocytes the heterogeneity of the samples on the time of sampling and on the
in the cerebral cortex of these patients [35]. As expected, IL-6 and IL-1β, fasting conditions of the subjects included can surely contribute to the
pleotropic cytokines that participate in the regulation of multiple central dispersion of our results.
biological phenomena in normal conditions [36, 37] were also detected. In the future, the implementation of longitudinal studies is necessary
However, their levels were significantly increased in most patients, as to integrate all these results. Indeed, we must consider that interactions
expected during an inflammatory process. between the two compartments are dynamic, and therefore, transversal
The lack of detection of TNF in brain tissue was not anticipated, studies showing the situation in a precise moment are not able to eval­
although its expression is known to be more transient than IL-1β [21]. uate the global situation. Such longitudinal studies focusing on the dy­
Interestingly, it has been reported that TNF may exhibit an anticon­ namic of central inflammation have been made in experimental models
vulsant effect [38]; it is therefore possible that its undetectability in the of epilepsy [39]. The evaluation of the systemic inflammation in these
severe patients included could be linked to the lack of response to models should be very informative to better understand the interrelation
treatment. between central and systemic conditions. Clinical longitudinal studies in
Significant correlations were found among central and peripheral PWE, using positron emission tomography (PET) with ligands specific to
inflammatory markers (Table 5 and graphical abstract). This result is inflammatory markers [40], could be of great interest to completely
interesting, and in agreement with previous findings [20]; it suggests understand the complexity of the interactions.
that it will eventually be possible to find peripheral markers reflecting
the intensity of central inflammatory reaction. 5. Conclusions
Particularly, it is interesting to note the positive correlations between
central CD4 T cells with the peripheral regulatory T cell populations, Our study showed that a relationship between central and systemic
and the negative correlations between the peripheral regulatory T cells inflammatory conditions exists in DR-TLE patients. This is an important
and the central pleiotropic IL-6 cytokine. Thus, it is possible that the observation that confirms the feasibility of finding peripheral markers
control of the inflammatory response in patients with epilepsy may be that may be useful in deciding both the possibility of prescription and
orchestrated by the regulatory CD4 T cells and the potential deviation of the chances of success of an anti-inflammatory treatment before
the Th1 response towards a Th2 response. considering the surgical option. Further studies confirming these
The results of the cluster analysis are also relevant. Two groups of exploratory results are necessary.
DRE patients were defined, one with a significant increase in central
inflammatory parameters (higher cell infiltration and IFNγ concentra­ Funding
tion), and the other with a higher concentration of the anti-
inflammatory cytokine IL-5 in the cerebral cortex. These results This work was funded by the Programa de Apoyo a Proyectos de
confirm the heterogeneity of the central inflammatory response in DRE Investigación e Innovación Tecnológica (PAPIIT, DGAPA, UNAM),
patients, and could explain why anti-inflammatory treatments may be of Project number: IN202217.
interest only in some of these patients. The marginal association of the
group of "inflammatory" patients with the peripheral increase in CD4 + Acknowledgements
CD38 +, a marker of activation of T lymphocytes, (p = 0.06) is of great
interest in the context of what has been said previously, that is to find The authors thank Mario Contreras Fleury for English copyediting.
peripheral markers associated with the intensity of the

454
A. Toledo et al. Seizure: European Journal of Epilepsy 91 (2021) 447–455

References [22] Makis AC, Tzoufi M, Kateri MD, Bourantas KL, Papadopoulou ZL. Valproate-
induced eosinophilia in children with epilepsy: role of interleukin-5. J Child Neurol
2005;20:150–2. https://doi.org/10.1177/08830738050200022201.
[1] Fisher RS, Van Emde Boas W, Blume W, Elger C, Genton P, Lee P, et al. Epileptic
[23] Elomaa AP, Niskanen L, Herzig KH, Viinamäki H, Hintikka J, Koivumaa-
seizures and epilepsy: definitions proposed by the International League Against
Honkanen H, et al. Elevated levels of serum IL-5 are associated with an increased
Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia 2005;45:
likelihood of major depressive disorder. BMC Psychiatry 2012;12:1–8. https://doi.
470–2. https://doi.org/10.1111/j.0013-9580.2005.66104.x.
org/10.1186/1471-244X-12-2.
[2] Fisher RS, Acevedo C, Arzimanoglou A, Bogacz A, Cross JH, Elger CE, et al. ILAE
[24] Phoenix TN, Temple S. Spred1, a negative regulator of Ras-MAPK-ERK, is enriched
official report: a practical clinical definition of epilepsy. Epilepsia 2014;55:475–82.
in CNS germinal zones, dampens NSC proliferation, and maintains ventricular zone
https://doi.org/10.1111/epi.12550.
structure. Genes Dev 2010;24:45–56. https://doi.org/10.1101/gad.1839510.
[3] WHO, 2019. Epilepsy: A public health imperative, Geneva: World Health
[25] Vieira ELM, De Oliveira GNM, Lessa JMK, Gonçalves AP, Oliveira ACP, Bauer ME,
Organization, Licence: CC BY-NC-SA 3.0 IGO, ISBN 978-92-4-151593-1.
et al. Peripheral leukocyte profile in people with temporal lobe epilepsy reflects the
[4] Kwan P, Arzimanoglou A, Berg AT, Brodie MJ, Allen Hauser W, Mathern G, et al.
associated proinflammatory state. Brain Behav Immun 2016;53:123–30. https://
Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force
doi.org/10.1016/j.bbi.2015.11.016.
of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51:1069–77.
[26] Ni FF, Li CR, Liao JX, Wang GB, Lin SF, Xia Y, et al. The effects of ketogenic diet on
https://doi.org/10.1111/j.1528-1167.2009.02397.x.
the Th17/Treg cells imbalance in patients with intractable childhood epilepsy.
[5] Kalilani L, Sun X, Pelgrims B, Noack-Rink M, Villanueva V. The epidemiology of
Seizure 2016;38:17–22. https://doi.org/10.1016/j.seizure.2016.03.006.
drug-resistant epilepsy: A systematic review and meta-analysis. Epilepsia 2018;59:
[27] Ouédraogo O, Rébillard RM, Jamann H, Mamane VH, Clénet ML, Daigneault A,
2179–93. https://doi.org/10.1111/epi.14596.
et al. Increased frequency of proinflammatory CD4 T cells and pathological levels
[6] Mehdizadeh A, Barzegar M, Negargar S, Yahyavi A, Raeisi S. The current and
of serum neurofilament light chain in adult drug-resistant epilepsy. Epilepsia 2020;
emerging therapeutic approaches in drug-resistant epilepsy management. Acta
00:1–14. https://doi.org/10.1111/epi.16742.
Neurol Belg 2019;119:155–62. https://doi.org/10.1007/s13760-019-01120-8.
[28] Gershen LD, Zanotti-Fregonara. P, Dustin IH, Liow JS, Hirvonen J, Kreisl WC,
[7] Bakpa OD, Reuber M, Irani SR. Antibody-associated epilepsies: Clinical features,
Jenko KJ, et al. Neuroinflammation in Temporal Lobe Epilepsy Measured Using
evidence for immunotherapies and future research questions. Seizure 2016;41:
Positron Emission Tomographic Imaging of Translocator Protein. JAMA Neurol
26–41. https://doi.org/10.1016/j.seizure.2016.07.002.
2015;72:882–8. https://doi.org/10.1001/jamaneurol.2015.0941.
[8] Blumcke I, Spreafico R, Haaker G, Coras R, Kobow K, Bien CG, et al.
[29] Ravizza T, Gagliardi B, Noé F, Aronica E, Vezzani A. Innate and adaptive immunity
Histopathological Findings in Brain Tissue Obtained during Epilepsy Surgery.
during epileptogenesis and spontaneous seizures: evidence from experimental
N Engl J Med 2017;377:648–56. https://doi.org/10.1056/NEJMoa1703784.
models and human temporal lobe epilepsy. Neurobiol Dis 2008;29:142–60.
[9] Vezzani A, Balosso S, Ravizza T. Neuroinflammatory pathways as treatment targets
https://doi.org/10.1016/j.nbd.2007.08.012.
and biomarkers in epilepsy. Nat Rev Neurol 2019;15:459–72. https://doi.org/
[30] Nakahara H, Konishi Y, Beach TG, Yamada N, Makino S, Tooyama I, et al.
10.1038/s41582-019-0217-x.
Infiltration of T lymphocytes and expression of ICAM-1 in the hippocampus of
[10] Riazi K, Galic MA, Pittman QJ. Contributions of peripheral inflammation to seizure
patients with hippocampal sclerosis. Acta Histochem Cytochem 2010;43:157–62.
susceptibility: cytokines and brain excitability. Epilepsy Res 2010;89:34–42.
https://doi.org/10.1267/ahc.10022.
https://doi.org/10.1016/j.eplepsyres.2009.09.004.
[31] Zattoni M, Mura ML, Deprez F, Schewndener RA, Engelhardt B, Frei K, et al. Brain
[11] Murta V, Ferrari CC. Influence of Peripheral inflammation on the progression of
infiltration of leukocytes contributes to the pathophysiology of temporal lobe
multiple sclerosis: evidence from the clinic and experimental animal models. Mol
epilepsy. J Neurosci 2011;31:4037–50. https://doi.org/10.1523/
Cell Neurosci 2013;53:6–13. https://doi.org/10.1016/j.mcn.2012.06.004.
JNEUROSCI.6210-10.2011.
[12] Terrone G, Balosso S, Pauletti A, Ravizza T, Vezzani A. Inflammation and reactive
[32] Gales J, Payson R. Chronic inflammation in refractory hippocampal sclerosis-
oxygen species as disease modifiers in epilepsy. Neuropharmacology 2019;14:
related temporal lobe epilepsy. Ann Diagn Pathol 2017;30:12–6. https://doi.org/
107742. https://doi.org/10.1016/j.neuropharm.2019.107742.
10.1016/j.anndiagpath.2017.05.009.
[13] Garfias S, Tamaya Domínguez B, Toledo Rojas A, Arroyo M, Rodríguez U, Boll C,
[33] Tezer FI, Firat A, Tuzun E, Unal I, Soylemezoglu F, Bilginer B. Immunopathology in
et al. Peripheral blood lymphocyte phenotypes in Alzheimer and Parkinson’s
drug resistant mesial temporal lobe epilepsy with different types of hippocampal
diseases. Neurologia 2019. https://doi.org/10.1016/j.nrl.2018.10.004. 19 pii:
sclerosis. Int J Neurosci 2018;128:421–8. https://doi.org/10.1080/
S0213-4853 30013-1.
00207454.2017.1389928.
[14] Vega-García A, Rocha L, Guevara-Guzmán R, Guerra-Araiza C, Feria-Romero I,
[34] Marchi N, Teng Q, Ghosh C, Fan Q, Nguyen MT, Desai NK, et al. Blood-brain
Gallardo JM, et al. Magnolia officinalis reduces inflammation and damage induced
barrier damage, but not parenchymal white blood cells, is a hallmark of seizure
by recurrent Status epilepticus in immature rats. Curr Pharm Des 2020;26. https://
activity. Brain Res 2010;1353:176–86. https://doi.org/10.1016/j.
doi.org/10.2174/1381612826666200320121813. 1388-01.
brainres.2010.06.051.
[15] Schieb H, Kratzin H, Jahn O, Möbius W, Rabe S, Staufenbiel M, et al. Beta-amyloid
[35] Wojkowska DW, Szpakowski P, Glabinski A. Interleukin 17A Promotes
peptide variants in brains and cerebrospinal fluid from amyloid precursor protein
Lymphocytes Adhesion and Induces CCL2 and CXCL1 Release from Brain
(APP) transgenic mice: comparison with human Alzheimer amyloid. J Biol Chem
Endothelial Cells. Int J Mol Sci 2017;18:1000. https://doi.org/10.3390/
2011;286:33747–58. https://doi.org/10.1074/jbc.M111.246561.
ijms18051000.
[16] Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology
[36] Erta M, Quintana A, Hidalgo J. Interleukin-6, a major cytokine in the central
1990;1:43–6. PMID: 2081237.
nervous system. Int J Biol Sci 2012;8:1254–66. https://doi.org/10.7150/ijbs.4679.
[17] Blümcke I, Thom M, Aronica E, Armstrong DD, Bartolomei F, Bernasconi A, et al.
Epub 2012 Oct 25.
International consensus classification of hippocampal sclerosis in temporal lobe
[37] Hewett SJ, Jackman NA, Claycomb RJ. Interleukin-1β in Central Nervous System
epilepsy: a Task Force report from the ILAE Commission on Diagnostic Methods.
Injury and Repair. Eur J Neurodegener Dis 2012;1:195–211. PMID: 26082912;
Epilepsia 2013;54:1315–29. https://doi.org/10.1111/epi.12220.
PMCID: PMC4465544.
[18] Ladino LD, Benjumea-Cuartas V, Vargas-Osorio J, Villamil Osorio L, Hernandez
[38] Balosso S, Ravizza T, Perego C, Peschon J, Campbell IL, De Simoni MG, et al.
Vanegas L, Matinez Juarez IE, et al. Barreras de acceso a la cirugia de la epilepsia:
Tumor necrosis factor-alpha inhibits seizures in mice via p75 receptors. Ann Neurol
revision de la bibliografia [Barriers impeding access to epilepsy surgery: a review
2005;57:804–12. https://doi.org/10.1002/ana.20480.
of the literature]. Rev Neurol 2017;65:268–79.
[39] Nguyen DL, Wimberley C, Truillet C, Jego B, Caillé F, Pottier G, et al. Longitudinal
[19] Özkara Ç, Vigevano F. Immuno and antiinflammatory therapies in epileptic
positron emission tomography imaging of glial cell activation in a mouse model of
disorders. Epilepsia 2011;52:45–51. https://doi.org/10.1111/j.1528-
mesial temporal lobe epilepsy: Toward identification of optimal treatment
1167.2011.03036.x.
windows. Epilepsia 2018;59:1234–44. https://doi.org/10.1111/epi.14083.
[20] Rawat C, Kukal S, Dahiya UR, Kukreti R. Cyclooxygenase-2 (COX-2) inhibitors:
[40] Dickstein LP, Liow JS, Austermuehle A, Zoghbi S, Inati SK, Zaghloul K, et al.
future therapeutic strategies for epilepsy management. J Neuroinflammation 2019;
Neuroinflammation in neocortical epilepsy measured by PET imaging of
16:197. https://doi.org/10.1186/s12974-019-1592-3.
translocator protein. Epilepsia 2019;60:1248–54. https://doi.org/10.1111/
[21] Ravizza T, Vezzani A. Pharmacological targeting of brain inflammation in epilepsy:
epi.15967.
Therapeutic perspectives from experimental and clinical studies. Epilepsia Open
2017;16:133–42. doi:0.1002/epi4.12242.

455