REVIEW

published: 06 April 2018

doi: 10.3389/fnana.2018.00025

Anatomy Based Networks and

Topology Alteration in
Seizure-Related Cognitive Outcomes
Qian Wu 1 , Charlie W. Zhao 2 , Zhe Long 3 , Bo Xiao 4 and Li Feng 4 *
1
Department of Neurology, First Affiliated Hospital, Kunming Medical University, Kunming, China, 2 Department of
Neuroscience, Yale School of Medicine, Yale University, New Haven, CT, United States, 3 Sydney Medical School and the
Brain & Mind Institute, The University of Sydney, Camperdown, NSW, Australia, 4 Department of Neurology, Xiangya Hospital,
Central South University, Changsha, China

Epilepsy is a paroxysmal neurological disorder characterized by recurrent and

unprovoked seizures affecting approximately 50 million people worldwide. Cognitive
dysfunction induced by seizures is a severe comorbidity of epilepsy and epilepsy
syndromes and reduces patients’ quality of life. Seizures, along with accompanying
histopathological and pathophysiological changes, are associated with cognitive
comorbidities. Advances in imaging technology and computing allow anatomical and
topological changes in neural networks to be visualized. Anatomical components
including the hippocampus, amygdala, cortex, corpus callosum (CC), cerebellum and
white matter (WM) are the fundamental components of seizure- and cognition-related
topological networks. Damage to these structures and their substructures results in
worsening of epilepsy symptoms and cognitive dysfunction. In this review article, we
survey structural, network changes and topological alteration in different regions of the
Edited by: brain and in different epilepsy and epileptic syndromes, and discuss what these changes
Alberto Munoz,
Complutense University of Madrid,
may mean for cognitive outcomes related to these disease states.
Spain
Keywords: epilepsy, cognition, networks, topology, seizure
Reviewed by:
Ursula Susan Sandau,
Oregon Health & Science University, INTRODUCTION
United States
Jon I. Arellano, Epilepsy is a prevalent neurological disorder characterized by recurrent seizures. An important
Yale School of Medicine,
Yale University, United States
complication of epilepsy and epilepsy syndromes is disruption in cognitive ability. Understanding
how cognitive dysfunction occurs will not only improve the quality of life for patients, but will also
*Correspondence:
help elucidate neural mechanisms underlying cognition.
Li Feng
[email protected] There are many forms of epilepsy, broadly characterized as involving partial or generalized
seizures. Epilepsy syndromes are defined by associated clinical manifestations. Childhood absence
Received: 07 October 2017 epilepsy (CAE), for instance, is characterized by brief, consciousness-impairing staring spells and
Accepted: 20 March 2018 3-Hz discharges on EEG.
Published: 06 April 2018 Some forms of epilepsy and epilepsy syndromes may lead to cognitive impairment, further
Citation: compounding disease burden. Yet, how cognitive impairment relates to seizures is unclear. How
Wu Q, Zhao CW, Long Z, Xiao B and do seizures impact cognition? What are the neural mechanisms mediating this relationship? These
Feng L (2018) Anatomy Based
are important questions being addressed with investigation on multiple levels.
Networks and Topology Alteration in
Seizure-Related Cognitive Outcomes.
In this review article, we focus on changes to neuroanatomy and neural networks that alter
Front. Neuroanat. 12:25. the topological relationship between brain regions. These alterations result in greater seizure
doi: 10.3389/fnana.2018.00025 susceptibility and cognitive dysfunction. Furthermore, increased seizure severity causes further

Frontiers in Neuroanatomy | www.frontiersin.org 1 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

TABLE 1 | A list of abbreviations used in this review article. recurrent seizures and improves epilepsy-associated memory
Abbreviation Meaning
deficit (Cho et al., 2015). In addition, brief recurrent seizures
induced by flurothyl and the tetanus toxin result in dramatic
AHC Amygdalo-hippocampal complex
impairment of hippocampus-based spatial learning and memory
BECTS Benign epilepsy with centro-temporal spikes
CA Cornu ammonis (region of the hippocampus)
(Nishimura et al., 2011). Place cells in cornu ammonis 1 (CA1)
CAE Childhood absence epilepsy are unable to form a stable spatial map when animals experience
CC corpus callosum early life seizures (ELS; Karnam et al., 2009). In the long-term
CPS Complex partial seizure kindling rat model, seizures during cell maturation prevent
DMN Default mode network
newborn neurons from integrating properly into hippocampal
DTI Diffuse tensor imaging
ELS Early life seizures circuits that are important for memory formation. Reducing
FDG-PET Fluorodeoxyglucose positron emission tomography activation of adult-born neurons may be a therapeutic strategy to
FLE Frontal lobe epilepsy reverse cognitive deficits in epileptic syndromes (Fournier et al.,
FS Febrile seizure 2013).
HS Hippocampal sclerosis
JME Juvenile myoclonus epilepsy
In temporal lobe epilepsy (TLE), extra-hippocampal
LSTGp Left posterior superior temporal gyrus volume abnormalities are observed in association with
mTLE Medial temporal lobe epilepsy cognitive dysfunction. Chronic TLE is characterized by extra-
PFC Prefrontal cortex hippocampal brain abnormality and cognitive impairment
rTMS Repetitive transcranial magnetic stimulation in both memory and non-memory domains (Tuchscherer
SMA Supplemental motor area
STP Short term potentiation
et al., 2010). Baseline thalamic volumes of TLE patients are
TLE Temporal lobe epilepsy lower than that of controls and executive functioning is poorer
WM White matter in the TLE group even after correcting for hippocampal
and frontal lobe volumes (Tuchscherer et al., 2010). The
topological changes, thus establishing a vicious cycle. In this thalamus is involved in the evolution and propagation of
survey, we focus on specific brain areas involved in seizure partial seizures and plays a role in cognition. The anterior
initiation and cognition and epilepsy syndromes associated thalamus interacts with the hippocampus and cortex and
with cognitive impairment. Abbreviations used throughout this functions in memory processing and spatial navigation, the
review article are in Table 1. A summary of structural and intralaminar thalamic nucleus and the parafascicular thalamus
functional anatomic findings associated with major epilepsy are involved in behavioral flexibility, and the mediodorsal
syndromes covered in this review article can be found in Table 2. thalamic nucleus in goal-directed behavior (Saniya et al., 2017).
The amygdala also affects cognitive tasks such as emotional
appraisal. The lateral amygdala, which when stimulated
ANATOMICAL BASIS OF causes experiential symptoms in TLE patients, projects to
SEIZURE-RELATED COGNITIVE the hippocampus and temporal neocortex, and selective
OUTCOMES amygdalotomy can be used to effectively treat TLE (Saniya et al.,
2017).
Hippocampal Dysfunction Plasticity and metaplasticity are thought to be fundamental
Animal and clinical studies show that seizures can result to learning and memory and may be involved in epilepsy.
in hippocampal dysfunction associated with cognitive and Synaptic plasticity is modulated by prior synaptic activity,
behavioral disturbances in both the developing and mature a phenomenon termed metaplasticity (Zhang and Luo,
brain (Alvarez et al., 2014). The hippocampus plays an 2011). Dysfunction of plasticity and metaplasticity in the
important role in memory formation. Signals from the occipital, hippocampus is implicated in febrile seizure (FS), a common
temporal and parietal lobes, posterior cingulated cortex and the childhood episode that can impair cognitive function
contralateral hippocampus converge on hippocampal neurons (Zhang and Luo, 2011). FS impairs metaplasticity of the
through the medial and lateral perforant pathways and through lateral perforant path of the rat hippocampus without
the anterior commissure (Aggleton, 2012). Additional synaptic affecting long-term potentiation, suggesting that FS may
integration arises from local recurrent excitatory networks, cause dysfunction in the excitatory status of other pathways
where feedforward and inhibitory connections contribute to the that ultimately lead to brain damage (Zhang and Luo,
acquisition of new episodic memories (Casanova et al., 2014). 2011).
Seizures have been shown to cause aberrant neurogenesis In the following sections, we further explore the relationship
in the hippocampus and form faulty circuits that disrupt between seizures and hippocampal dysfunction, specifically
hippocampal function (Fournier et al., 2013). Aberrant focusing on Hippocampal sclerosis (HS), dendritic pathology and
neurogenesis, including neural progenitor proliferation, cortical alterations.
ectopic granule cell production and mossy fiber sprouting
after seizures, are harmful to new memory formation (Arnold Hippocampal Sclerosis
et al., 2016). In a genetic mouse model, blocking ectopic granule HS is a pathological development associated with benign
cell production using Nestin-δ-HSV-thymidine kinase-EGFP and drug-resistant mTLE and is involved in epilepsy-related
before acute seizures reduces the frequency of spontaneous cognitive dysfunction (Blumcke et al., 2012). Unilateral HS

Frontiers in Neuroanatomy | www.frontiersin.org 2 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

TABLE 2 | Structural and functional anatomic changes involved in the epilepsy syndromes covered in this review article.

Condition Structural Functional

Neuronal injury in l. septal nuclei, amygdala, v. CA1 unable to form spatial mapPrefrontal cortex short term
Early life seizures subiculum/CA1 potentiation alterations (LII/LIII-to-LV and LV-to-LV)
Temporal lobe epilepsy Volume loss in thalamus, hippocampus, cerebellum Lateral amygdala stimulation causes experiential symptoms
Temporal and frontal cortical thinning (only frontal in White matter abnormalities in a. and m. temporal, ips. cerebellum,
some studies) p. callosum, con. frontoparietal
With postictal psychosis: prefrontal and temporal Interictal hypometabolism in epileptic region
thickening Poorly segregated cognitive modules
Temporal lobe epilepsy with Neuronal loss and gliosis centered around CA1 Decrease connectivity between amygdalo-hippocampal complex
hippocampal sclerosis Granule cell dispersion and loss and sprouting of and bil. v. prefrontal, temporal pole, con. p. cingulate; Increase
interneurons in dentate gyrus connectivity with a. cingulate, d. m. prefrontal, bil. temporo-parietal
Mesiotemporal cortex sclerosis, neuronal loss, gliosis junction
Increase in complexity of temporal and frontal cortical Increased synchronization in ips. parahippocampus, midbrain,
folding insula, callosum, bil. sensorimotor cortex, frontoparietal subcortical
Cortical thinning in regions connected with structures; Decreased synchronization in cerebellum, precuneus, p.
hippocampus cingulate, bil. i. l. parietal, m. PFC
Frontal lobe epilepsy Frontal cortical volume loss in children for Frontal, temporal, parietal hypometabolism
left FLE: (ips) s. frontal, paracentral, precuneus,
cingulate, i. parietal, supramarginal, postcentral, s.
temporal (con) s. and m. frontal, m. orbitofrontal,
supramarginal, postcentral, s. temporal banks,
parahippocampus;
right FLE: (ips) precentral, postcentral, transverse
temporal, parahippocampus, lingual, l. occipital (con)
s. front, i. parietal, postcentral, s. temporal, p.
cingulate, lingual
Juvenile myoclonic epilepsy Gray matter changes in s. m. frontal, p. cingulate and White matter abnormalities in bil. a. and s. corona radiata, callosum
a. callosum genu and body, cingulum-temporal connections, p. parietal, and
Gray matter reduction in supplementary motor area frontal
and p. cingulate Reduced connectivity between prefrontal and frontopolar regions;
Increased connectivity between occipital cortex and supplementary
motor area
Hyperconnectivity in subnetwork involving primary motor cortex,
precuneus, cerebellum lobules IV and V, basal ganglia, bil.
parietal/postcentral, subcortical regions and right hippocampus
Childhood absence epilepsy White matter abnormalities in bil. thalamus, a. callosum, upper
brainstem, prefrontal, a. cingulate, parietal, p. cerebellum, bil.
putamen, bil. p. internal capsule
Altered whole-brain topology
Impaired subcortical and orbitofrontal subnetworks
Microstructural changes in callosum and bil. precuneus

Some changes span both categories, in which case we classified it according to its dominant feature. Abbreviations used are positional: a-anterior, p-posterior, v-ventral,
d-dorsal, s-superior, i-inferior, l-lateral, m-medial, ips-ipsilateral, con-contralateral, bil-bilateral. Italicized text are animal studies.

is the most frequent mTLE pathological change, observed in In unilateral mTLE with HS, seizures can induce effective
60%–80% of mTLE patients (Berkovic et al., 1995). HS is connectivity alterations between the non-epileptic amygdalo-
also observed in patients who suffer from neurodegenerative hippocampal complex (AHC) and the rest of human brain
diseases including Alzheimer’s disease and fronto-temporal lobe (Trotta et al., 2013). These changes include a significant
dementia (Bandopadhyay et al., 2014). decrease in connectivity between the non-epileptic AHC and
Histologically, epileptic HS exhibits a consistent pattern the bilateral ventral prefrontal cortical areas, the temporal
of neuronal loss and gliosis centered on the CA1 subfield pole and the posterior cingulated cortex contralateral to
and a more variable loss in CA4 and other subfields HS, and a significant increase in connectivity between the
(Thom, 2009). Other features seen in epileptic HS include non-epileptic AHC and midline structures such as the
the dispersion of the granule cell layer of the dentate anterior cingulate and dorsal medial prefrontal cortices,
gyrus and sprouting of mossy fiber axons (Schmeiser et al., as well as the bilateral temporo-parietal junction (Trotta
2017). Loss and sprouting of dentate gyrus interneurons in et al., 2013). Connectivity alterations also exist between
either a unilateral or bilateral pattern are also observed in the non-epileptic AHC and some limbic and default mode
post mortem (Martinian et al., 2012; Thom et al., 2012). network (DMN) areas. These changes may account for
Interneuronal axon sprouting is a functional process related to the emotional, cognitive and decision-making impairment
network changes through synaptic re-organization (Maglóczky, that occur frequently in mTLE patients (Trotta et al.,
2010). 2013).

Frontiers in Neuroanatomy | www.frontiersin.org 3 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

Dendritic Pathology affected, and cognitive impairment in TLE is associated with the
Abnormalities in dendritic spines are often observed in severity of cortical abnormalities (Gutierrez-Galve et al., 2012).
brain specimens from epilepsy patients and animal models Many studies report cortical abnormalities in the temporal and
of epilepsy (Bartsch et al., 2010). Dendritic spine density frontal regions (Raj et al., 2010; Voets et al., 2011) while some
progressively decreases with increasing duration of infancy- or found abnormalities only in the frontal lobe (Gutierrez-Galve
childhood-onset epilepsy (Casanova et al., 2014). However, the et al., 2012). This discrepancy may arise from differences in
relationship between dendritic pathology and epileptogenesis inclusion criteria and methodology of the studies, thus more
and cognitive deficits is complex (Wong and Guo, 2013). evidence is needed for clarification. In general, in TLE with or
Dendritic spines are small actin rich dendritic protrusions without HS, cortical thinning is associated with loss of volume in
that receive excitatory input from axons and presynaptic signals. the hippocampus and anterior thalamus (Mueller et al., 2010).
They are the major sites of contact for synapsing neurons and TLE patients with HS is associated with neuropathology
provide a metric for the number and strength of signaling primarily in the frontotemporal cortex, which includes
connections between the elements of a functional neuronal mesiotemporal sclerosis, neuronal loss and gliosis (Kaaden et al.,
circuit (Mancuso et al., 2014). Epilepsy or seizures damage 2011). These extratemporal cortical abnormalities contribute
these structures, thus contributing to progressive epileptogenesis, to patterns of cognitive impairment in TLE, including IQ
decreased seizure thresholds and learning and memory problems decline and deficits in language, executive and motor functions
(Wong and Guo, 2013). In turn, dendritic spine abnormalities (Hermann et al., 2009; Keller et al., 2009; Tuchscherer et al.,
can promote hyperexcitability in circuits, further worsening 2010). Surface-based morphometry reveals an increase in the
the epilepsy and associated cognitive dysfunction. Using a complexity of temporal and frontal cortical folding distant from
morphologically and biophysically realistic model of a bursting the epileptic focus (Voets et al., 2011) and widespread cortical
layer 5 pyramidal cell from the cat visual cortex, van Elburg and thinning in regions connected with the hippocampus (Raj et al.,
van Ooyen (2010) showed that alterations in size or topology of 2010).
pyramidal cell morphology, such as shortening or lengthening In TLE with postictal psychosis, cortical thickness is increased
the dendritic tree, or even just modifying the pattern in which the in prefrontal and temporal regions (DuBois et al., 2011). Cortical
branches in the tree are connected, may change neuronal burst abnormalities in extratemporal areas in TLE patients may not
firing and affect information processing and cognition. This only explain the observed cognitive impairment, but also reflect
resembles the neuronal mechanism seen in Alzheimer’s disease, the spread of seizure activity through the thalamus in addition to
mental retardation, epilepsy and chronic stress (van Elburg and limbic pathways (Bernhardt et al., 2008).
van Ooyen, 2010). FLE and its corresponding cognitive outcomes are
Recurrent ELS may suppress dendrite growth by impairing associated with changes in cortical parameters. A study
the addition of new branches and/or impairing the growth of using 18 F-Fluorodeoxyglucose positron emission tomography
existing branches (Casanova et al., 2012). Apical and basilar (FDG-PET) in children with FLE shows widespread
dendrites appear to respond differently to seizures. Recurrent hypometabolism not only in the frontal lobe but also in the
ELS do not affect the length or branch number of apical dendrites extrafrontal cortex, including the temporal and parietal cortices
but reduce that of basilar dendrites in mouse models (Nishimura (da Silva et al., 1997). This suggests that, similar to TLE, FLE
et al., 2011; Yang et al., 2015). The latter may reduce anatomical seizures also spread from a single epileptic focus to brain regions
and molecular substrates for learning and memory. Even though via pathways including cortico-cortical connections.
apical dendrite morphology appear unaffected, seizures may Frontal cortical volumes are smaller in children with
still produce activity-dependent alterations in synaptic efficacy intractable FLE than in healthy children (Lawson et al.,
and affect aspects of hippocampal synaptic plasticity, such as 2002) with similar changes seen in extrafrontal areas (Widjaja
long-term potentiation and depression, which are thought to et al., 2011), which may explain the complicated cognitive
form the cellular basis for hippocampal learning and memory dysfunction in intractable FLE. Surface-based morphometry
(Alarcon et al., 2006). shows that in children with left FLE, cortical thinning is
Interestingly, the effect of seizures on dendrite anatomy present in the left superior frontal, paracentral, precuneus,
and spatial learning appears to be age-dependent. The effects cingulate, inferior parietal, supramarginal, postcentral and
mentioned above are absent when seizures are induced in superior temporal gyri, as well as in the right superior and
older rats. Thus, there may exist a neurodevelopmental time middle frontal, medial orbitofrontal, supramarginal, postcentral,
window during which dendritic growth and maintenance are banks of superior temporal sulcus and parahippocampal gyri.
particularly vulnerable to seizures. This would explain the Conversely, in children with right FLE, cortical thinning is
cognitive impairment seen in children with epilepsy and in adult present in the right precentral, postcentral, transverse temporal,
patients who have had early-onset epilepsy (Hermann et al., parahippocampal, lingual and lateral occipital gyri, as well as in
2008). the left superior frontal, inferior parietal, postcentral, superior
temporal, posterior cingulate and lingual gyri (Widjaja et al.,
Cortical Changes 2011). Widespread cortical thinning and gray matter volume
We next discuss cortical changes in TLE, frontal lobe epilepsy reduction appear to be more prominent in left FLE than in
(FLE), ELS and juvenile myoclonus epilepsy (JME). In TLE, right FLE (Widjaja et al., 2011). We note that although cortical
cortical parameters including area, thickness and volume are abnormalities are observed in FLE, there is currently no direct

Frontiers in Neuroanatomy | www.frontiersin.org 4 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

evidence linking these changes and cognitive dysfunction in FLE cortex. Cognitively, they exhibit impaired verbal fluency,
patients. comprehension and expression, as well as impaired nonverbal
The prefrontal cortex (PFC) is an important brain region memory and mental flexibility (O’Muircheartaigh et al., 2011).
for neonatal seizure-related cognitive impairment, as in the White matter (WM) abnormalities in these regions and networks
case of ELS. Anatomical and electrophysiological studies show are also observed, which we further discuss in the next
that direct connections between the hippocampus and medial section.
PFC as part of the hippocampo-PFC circuit play an important Besides the cortex, seizure-induced abnormalities in the
role in aspects of learning and memory processing including callosal and cerebellar networks affect cognitive function. In
information consolidation and working memory (Preston and TLE, Schneider et al. (2014) found that anterior and mid-callosal
Eichenbaum, 2013). Even a single episode of neonatal seizure corpus callosum (CC) changes affect cognitive performance.
can permanently alter synaptic organization and transmission In chronic intractable TLE patients with intermittent explosive
(Isaeva et al., 2006) in the CA1 region of the hippocampus disorders, total brain and cerebellar volumes, particularly in the
and impair spatial learning and working memory (Kleen et al., left cerebellum, are influenced by patient age and duration of
2011b). Animals that experience ELS display a deficit in epilepsy. These alterations in total brain and cerebellar volumes
behavioral flexibility associated with PFC architectural changes are strongly associated with cognitive impairment, whereas
(Kleen et al., 2011a). ELS may also result in anxiety-like alterations in hippocampal volume have a minor influence on
behavior and impaired spatial learning and memory during the cognitive parameters (Hellwig et al., 2013).
developmental stage (Mlsna and Koh, 2013). These behavioral
deficits are temporally correlated with the presence of neuronal White Matter Changes
injury in the following regions involved in modulation of WM is composed of glial cells and myelinated axons and serves
the hypothalamic-pituitary-adrenal stress response (Mlsna and as signal transmitter from one cerebral region to another or
Koh, 2013): (1) the lateral septal nuclei which is involved in to lower brain centers. Diffusion tensor imaging (DTI) reveals
motivational and affective function; (2) the amygdala which the following associations between WM in different regions of
is involved in anxiety, threat-induced behavioral arousal and the brain and cognitive function: (1) right temporal WM with
emotion; and (3) the ventral subiculum/CA1 which is involved language and executive function; (2) the CC with intelligence
in spatial learning (Mlsna and Koh, 2013). and language; and (3) left parietal WM with language (Kim
The functional deficits with ELS may result from short et al., 2012; Widjaja et al., 2013a). WM impairment may reflect
term potentiation (STP) alterations in the PFC. STP of the connectivity disruptions in cortical processing networks that
PFC is important to its functions including short-term working are necessary for cognitive development (Widjaja et al., 2013a).
memory, information processing and decision-making processes Here we discuss specific WM changes seen in JME, chronic
(Hernan et al., 2013). Using a flurothyl mouse model of ELS, TLE and CAE.
it was observed that recurrent seizures early in development In patients with JME, WM abnormalities are seen in the
affect STP at Layer II/III (LII/III)-to-LV and LV-to-LV networks bilateral anterior and superior corona radiata, genu and body of
in the PFC (Hernan et al., 2013). Both networks are involved CC, cingulum connections to the temporal cortex (Seltzer and
in working memory tasks: the apical dendrites of LV neurons Pandya, 2009), posterior parietal regions (part of the splenium)
serve as the receiver for feedback information from thalamic and multiple frontal regions which result in widespread
inputs as well as from cortico-cortical connections in LII/III interconnection disturbances in the frontal lobe (Kim et al.,
(Kuroda et al., 1998), and the basal dendrites are integral to 2012). These structural abnormalities in the thalamofrontal
the interconnected network of deep pyramidal neurons whose network may account for JME patients’ poor performance in
continued firing during the delayed phase of a working memory frontal functions.
task is thought to underlie short-term working memory (Hernan Patients with chronic epilepsy may have cognitive
et al., 2013). Therefore, alterations in STP in the PFC induced by comorbidities and widespread network abnormalities outside the
ELS may lead to cognitive deficits (Deng et al., 2011), particularly epileptic zone, such as in WM areas that affect cognitive function
in spatial learning and working memory. and global intelligence (Vaessen et al., 2012). In chronic TLE
JME-related cognitive impairment is associated with cortical patients, WM abnormalities are seen in the anterior temporal
abnormalities. fMRI and EEG show gray matter volume lobe, mesial temporal lobe, ipsilateral cerebellum, posterior
alterations in the superior midline frontal regions in JME, regions of the CC, and the frontoparietal lobe contralateral to
with reduced volume seen in some studies (O’Muircheartaigh the side of seizure onset (Riley et al., 2010; Rodríguez-Cruces
et al., 2011) and increased volume in others (Koepp et al., et al., 2018). Abnormalities in the anterior temporal lobe are
2013). Gray matter of the posterior cingulate and the anterior correlated with delayed memory, the mesial temporal lobe with
callosum regions are affected in JME and may contribute immediate memory, and the CC with earlier age of seizure
to reduction in cognitive performance demonstrated in letter onset (Riley et al., 2010). Although chronic epileptic patients
fluency and similarity tasks, concordant with JME frontal with regional or multi-regional WM impairment usually exhibit
lobe dysfunctions (Sonmez et al., 2004). Generally speaking, severe and complicated cognitive impairment, their whole brain
JME patients have subtle focal cortical abnormalities and WM volume does not differ from that of healthy controls or of
gray matter reduction in the mesial frontal cortex, especially patients with little to no cognitive impairment (Vaessen et al.,
the supplementary motor area (SMA) and posterior cingulate 2012). This suggests that impaired WM connectivity rather

Frontiers in Neuroanatomy | www.frontiersin.org 5 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

than WM volumetric change is involved in cognitive decline in synchronization is significantly increased in the ipsilateral
patients with chronic epilepsy. parahippocampal gyrus, midbrain, insula, CC, bilateral
WM changes are observed in patients with CAE, a syndrome sensorimotor cortex and frontoparietal subcortical structures,
of idiopathic generalized epilepsy. In untreated CAE patients, but is decreased in the cerebellum and the DMN, specifically in
there are significant WM abnormalities in the bilateral thalamus, the precuneus and posterior cingulate gyrus, bilateral inferior
anterior CC, upper brainstem, prefrontal areas, anterior lateral parietal, and mesial PFC (Zeng et al., 2013). In absence
cingulate, parietal areas, posterior cerebellar hemispheres, and patients, functional connectivity is increased between the frontal,
in subcortical structures including the bilateral putamen and parietal and temporal lobes but is decreased in the DMN,
bilateral posterior limbs of the internal capsule. These changes which may result in cognitive mental impairment and loss of
point to impairment of WM integrity in the basal ganglia- consciousness during an absence seizure (Luo et al., 2011; Li
thalamocortical circuit of drug-naïve CAE patients, which may et al., 2015).
result in increased cortical excitability and cognitive, linguistic DMN abnormalities are seen during resting interictal periods
and behavioral/emotional deficits during and between seizures without interictal epileptiform discharges (Luo et al., 2011). In
(Yang et al., 2012). patients with complex partial seizures (CPS), the average volume
of activated brain regions is 98% higher than that of controls,
and 81% of activated areas are in cognitive regions of the frontal
NETWORK AND TOPOLOGICAL BASIS OF and temporal lobes, anterior cingulate cortex, precuneus and
SEIZURE-RELATED COGNITIVE cuneus, while the remaining 19% are in the precentral gyrus, the
OUTCOMES superior and medial occipital gyrus, the parahippocampal gyrus,
the inferior parietal lobule and the angular gyrus (Karmonik
Networks et al., 2010). In CPS patients, large areas of activation occur in the
Epileptogenesis engages multiple brain networks. The frontal and temporal lobes, as well as the cuneus and precuneus,
functionality of these networks is based on the level of as opposed to the control group where activation is found mostly
synchronous neural activity or structural correlates between in the parietal lobe (Karmonik et al., 2010). These results suggest
different areas through a complex pattern of increased or that switching from goal-directed behavior to the default mode
decreased connectivity (Stam and van Straaten, 2012). The in CPS patients is impaired.
exact number of networks involved and how they interact in In summary, complex partial, generalized tonic-clonic and
epileptogenesis is unclear. Analysis of known networks provides absence seizures can decrease the activity of the DMN (Danielson
information on: (1) the onset, propagation and termination of et al., 2011). While disorders in other regions may contribute
seizures (Kramer et al., 2008); (2) the preictal, ictal, interictal and to epileptogenesis and propagation, dysfunction in the DMN
postictal state of functional networks in epilepsy (Horstmann appears to be responsible for widespread functional cognitive
et al., 2010; van Dellen et al., 2012); (3) alterations in structural impairment (Zeng et al., 2013).
networks in epilepsy (Bernhardt et al., 2011); and (4) mechanisms
of seizure comorbidities such as cognitive decline and behavioral Brain Adaptive Networks and Connectivities
deficits (Vlooswijk et al., 2011; Vaessen et al., 2012). For a There is evidence that the brain undergoes structural and
detailed discussion on the functional and structural networks connectivity adaptations in response to seizures to preserve
involved and topological changes in epilepsy, we recommend cognitive function. We present some of the evidence in this
reviews by Halász (2010a,b), van Diessen et al. (2013) and Wang section, with a focus on language preservation in seizures of
et al. (2015). Here we discuss some of the networks associated the left posterior superior temporal gyrus (LSTGp), striatal
with seizure-related cognitive dysfunction focusing on the DMN, changes in benign epilepsy with centro-temporal spikes (BECTS)
adaptive networks and the thalamo-frontocortical network. and response to hypometabolism in temporal and frontal lobe
seizures.
The Default Mode Network: A Resting-State Network When the LSTGp is temporarily impaired by a seizure,
The DMN consists of brain regions that are activated during the brain trauma or stroke, the brain adapts to maintain language
resting state and during internally directed cognition (Karmonik comprehension ability. Data from repetitive transcranial
et al., 2010) and deactivated during task engagement (Danielson magnetic stimulation (rTMS) with fMRI studies suggest that this
et al., 2011). It has primary nodes (sites or brain areas) in the adaptation consists of two parts: (1) increased synchronization
precuneus/posterior cingulate and the medial frontal and lateral between compensating regions coupled with decreased
parietal cortices involved in introspective and social cognitive synchronization within the primary language network; and
functions (Danielson et al., 2011). The DMN is abnormal in (2) decreased activation at the rTMS site as well as in distal
several types of epilepsy, which can be either the cause or the regions followed by a recovery process. Adaptation involves
result of seizure-related cognitive comorbidities. three synchronization centers: the contralateral homolog of the
In children with refractory epilepsy, decreased DMN area receiving rTMS (i. e. the right STGp), areas adjacent to
connectivity is seen in the posterior cingulate cortex/precuneus, the rTMS site and the medial frontal gyrus, a region involved
bilateral lateral parietal cortex, and the anterior and in discourse monitoring (Mason et al., 2014). Because of such
mid-cingulate cortex (Widjaja et al., 2013b). In patients insights on the role of the language network in epilepsy-related
with medial temporal lobe epilepsy (mTLE) and HS, regional cognitive outcomes, particularly its adaptive response to injury,

Frontiers in Neuroanatomy | www.frontiersin.org 6 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

this network may be a therapeutic target to prevent cognitive prefrontal and frontopolar regions and increased between the
impairment in epilepsy. occipital cortex and the SMA (Vollmar et al., 2012). This
BECTS is a benign epilepsy syndrome involving the Rolandic may form the anatomical basis of cognitive triggering of
area. It is the most common childhood epilepsy syndrome motor seizures in JME, as well as the link between seizure
and is considered a neurodevelopmental disorder with an semiology, neurophysiology, neuropsychology and imaging
underlying genetic and anatomical basis (Lin et al., 2012). findings (Vollmar et al., 2012). A subnetwork comprising the
Children with BECTS show aberrant volume and morphology in primary motor cortex, precuneus, cerebellum lobules IV and
subcortical regions involved in motor processing and executive V, basal ganglia, bilateral parietal/postcentral gyrus, subcortical
functions with reduced functional connectivity in the Rolandic regions and right hippocampus is hyperconnected in JME
region. A meta-analysis concluded that BECTS has poor patients, which is correlated with decreased auditory memory,
cognition outcomes for language ability, which involves visual verbal fluency and executive function (Caeyenberghs et al., 2015).
processing, auditory processing, single-word reading, expressive Of note, the highly heritable nature of JME, along with the
and receptive language, verbal fluency, processing speed, fluid cognitive dysfunction observed in otherwise healthy siblings of
reasoning and verbal knowledge (Wickens et al., 2017). Structural JME patients, supports the concept of a genetically determined
changes include putamen hypertrophy, dorsoventral elongation thalamo-frontocortical network dysfunction (Wandschneider
of the left caudate and bilateral putamen, and subnuclei et al., 2012).
expansion in the ventral and dorsal striatum (Lin et al., 2012).
But these alterations may not cause cognitive deterioration Topology
but instead be cognitively adaptive, as larger putamen volumes Modern brain mapping techniques such as fMRI and DTI suggest
have been linked to better cognitive performance (Lin et al., that brain function depends on large-scale networks rather
2012). than isolated brain areas. Disorders in components of these
FDG-PET show that 60%–95% of unilateral mTLE patients networks, even a small change in neuronal network topology,
have significant hypometabolism in the epileptic temporal can break the network balance, induce explosive synchronization
regions during the interictal period (Trotta et al., 2011). transition and activity propagation, and lead to epileptic seizures
Significant metabolic changes have concomitantly been observed (Wang et al., 2017). A topological view of brain functioning
in the surrounding and remote brain regions, including the borrows concepts from its mathematical counterpart, concepts
lateral temporal areas, PFC, frontal lobe, thalamus and even such as strength, path length, clustering coefficient and
some areas within the DMN. These structures are involved efficiency, global efficiency, local efficiency, modularity, hub
in epilepsy-induced reorganization of neuronal networks and distribution and small-world. Study methodologies such as
effective connectivity within the mesiotemporal regions or other region of interest approach, unbiased whole brain approach and
distant brain areas (Zhang et al., 2010), the latter accounting for graph theoretical analysis are used to study the correlations
some of the cognitive impairment observed in mTLE (Takaya between network metrics and clinical characteristics. Within this
et al., 2009). Meanwhile, functional reorganization/plasticity in framework, brain images are treated as a close topological space
the non-epileptic temporal lobe may represent a compensatory consisting of functional and structural networks or sub-networks
mechanism sustaining key cognitive functions such as memory associated with cognitive and behavioral functions. Seizures, even
or speech (Bettus et al., 2009; Trotta et al., 2011). psychogenic non-epileptic seizures associated with attention,
Unlike mTLE, there is no consistent pattern of cognitive emotion and sensorimotor systems (Ding et al., 2014), induce
impairment seen in FLE patients, although some evidence an elastic-like deformation in the brain, resulting in alterations
support the notion that cognitive function is impaired of the connectedness, continuity and boundary of areas and
(O’Muircheartaigh and Richardson, 2012). Given that the networks. In this section, we briefly discuss topological changes
frontal lobes consist a large proportion of the cerebral cortex in the brain on the network level and the neuronal level.
and contains rich connections with other brain regions, it Recent studies examined topological changes in functional
is not surprising to either find or not find structure-related and structural networks and sub-networks in epilepsy and their
cognitive deficits in FLE (O’Muircheartaigh and Richardson, relation to seizures and seizure-related cognitive outcomes.
2012). Bonilha et al. (2014) found that children with new-onset epilepsy
have a suboptimal topological structural organization with
Thalamo-Frontocortical Network enhanced network segregation and reduced global integration.
Thalamo-frontocortical network disorders are seen in JME, This results in: (1) structural reorganization that involves the
which is the most common idiopathic epilepsy syndrome redistribution of nodes from the posterior to the anterior head
and is considered a benign seizure disorder (Wandschneider regions; and (2) lower IQ and poorer executive function in these
et al., 2012). The JME phenotype includes a frontal lobe children (Bonilha et al., 2014).
type neuropsychological profile, photosensitivity, hyperexcitable The most common epilepsy syndrome in children, CAE,
motor cortex, and abnormal functional connectivity of the is shown to have altered whole-brain structural network
motor cortex and SMA (Vollmar et al., 2012). Advanced topology, impaired subnetworks in subcortical and orbitofrontal
imaging studies have identified functional and structural structures, microstructural changes in the genu of the CC and
abnormalities in the frontal cortex and thalamus in JME patients bilateral precuneus. The bilateral precuneus is a core area in
(Wandschneider et al., 2012). Connectivity is reduced between the DMN and involved in processing of visuospatial imagery,

Frontiers in Neuroanatomy | www.frontiersin.org 7 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

episodic memory retrieval and consciousness (Liang et al., 2016; CONCLUSION

Qiu et al., 2017).
Others have found topological alterations in cognition- There are still many unknowns in the relationship between
related brain domains in TLE. Using graph analysis, Kellermann epileptogenesis and cognitive and topological alteration.
et al. (2016) divided functional brain domains into several Mechanistic studies thus far have been performed using
modules. They found that TLE patients have poorly segregated animal models while observational studies are performed
cognitive modules compared to non-TLE controls. In the on animals and human patients. Animal models have less
control group, there are numerous segregated, small modules heterogenicity than human patients, which is advantageous in
related to cognitive functioning that include: (1) verbal memory; mechanistic studies where only one or two factors are examined.
(2) language, perception and intelligence; (3) speed and fluency; There are, however, the following limitations with animal
and (4) visual memory and executive function, whereas in models: (1) human epilepsy includes acquired, idiopathic and
TLE patients there are fewer, larger and more mixed modules cryptogenic causes, whereas animal epilepsy models are mostly
including: (1) verbal memory, visual memory and executive induced by chemical or physical methods, which means they
function; (2) speed and fluency; and (3) speed, executive cannot completely imitate the pathophysiological course of
function, perception, language, intelligence and nonverbal human epilepsy; (2) chemical or physical ignition affects the
memory (Kellermann et al., 2016). This phenomenon may whole brain and disrupts the neurogenic niche, which makes
suggest a compensatory mechanism to protect brain networks it difficult to parse out the contribution of specific regions or
and neurocognitive function from epileptic impairment, which structures to cognition; and (3) human beings have higher
may be helpful clinically for cognition training and rehabilitation intelligence and more advanced cognitive functions such as
attempts. computation, language and integrated analysis, which cannot
Topological changes are also seen in the elements of the be studied in animal models, thus animal models are used to
network: neurons. Neurons display a wide range of intrinsic study memory, social behavior and recognition. For readers
firing patterns (van Elburg and van Ooyen, 2010). Burst interested in chemosonbulsant-induced epilepsy and cognitive
firing, the generation of clusters of action potentials with testing in animal models, we recommend a review by Minjarez
short inter-spike intervals, is a relevant pattern for neuronal et al. (2017).
signaling and synaptic plasticity and is modulated by dendritic On the other hand, complicating factors in human studies
morphology and ion channel composition. Computational include the following: (1) most epilepsy patients studied
modeling of neocortical pyramidal cells shows that it is not are receiving antiepileptic drugs that can result in cognitive
only the total length of the apical dendrite but also the impairment. This acts as a confounding factor when we aim
topological structure of its branching pattern that influence to study the direct relationship between epileptogenesis and
inter- and intra-burst spike intervals and predict whether epilepsy-associated cognitive deficit; and (2) patients who suffer
the cell exhibits burst firing (van Elburg and van Ooyen, from chronic epilepsy often have comorbid depression and
2010). Varying the topology of apical dendritic trees by anxiety, which can result in cognitive decline.
swapping branches within the tree, all the trees will have Alterations in anatomic structure, networks and topology are
the same total length and surface area and differ only in associated with seizure-related cognitive impairment. Integrating
the way their branches are connected. This results in wider multiscale imaging approaches such as CT, PET and MRI
connections between pyramidal cells and produce firing patterns for macroscopic observation, super-resolution and electron
ranging from tonic firing to strongly bursting. Only a range microscopy for microscopic research, and computational tools
of dendritic sizes supports burst firing, and this range is for analysis, studying structural or sub-structural alterations in
modulated by dendritic topology. Alterations in the size or the brain is a promising and rewarding approach to understand
topology of pyramidal cells, as seen in epilepsy, may change brain function and dysfunction.
neuronal burst firing patterns and thus affect higher-level
information processing and cognition (van Elburg and van AUTHOR CONTRIBUTIONS
Ooyen, 2010).
QW: conception and design, manuscript writing, final approval
Because changes in topology may influence seizure spread and
of manuscript. CWZ and ZL: conception and design, final
seizure-related cognitive impairment, the following questions
approval of manuscript. BX: final approval of manuscript.
may serve as future research directions: are there rules
LF: conception and design, financial support, final approval of
or principles that can explain and predict seizure spread
manuscript.
and seizure-related cognitive outcomes? To what extent are
topological changes in multiple regions involved? ACKNOWLEDGMENTS
The topographical framework of viewing brain function is
still in early development. There is a need for standard analysis This work was supported by the National Natural Science
methods, as different methods lead to different and sometimes Foundation of China (81601134, 81771407, 81260199,
contradictory findings. For example, the definition of network 81660228), Yunnan Applied Basic Research Projects (2017
weights, such as the definition of edge weights in DTI network FE468(-144)), and the Doctoral Scientific Research Foundation
analysis, can change the calculated network properties. Thus of First Affiliated Hospital of Kunming Medical University
more accurate and unified analysis should be a research focus. (2015BS007).

Frontiers in Neuroanatomy | www.frontiersin.org 8 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

REFERENCES syndrome. J. Neurosci. 31, 10971–10982. doi: 10.1523/JNEUROSCI.2021-

11.2011
Aggleton, J. P. (2012). Multiple anatomical systems embedded within the primate Ding, J., An, D., Liao, W., Wu, G., Xu, Q., Zhou, D., et al. (2014). Abnormal
medial temporal lobe: implications for hippocampal function. Neurosci. functional connectivity density in psychogenic non-epileptic seizures. Epilepsy
Biobehav. Rev. 36, 1579–1596. doi: 10.1016/j.neubiorev.2011.09.005 Res. 108, 1184–1194. doi: 10.1016/j.eplepsyres.2014.05.006
Alarcon, J. M., Barco, A., and Kandel, E. R. (2006). Capture of the late phase DuBois, J. M., Devinsky, O., Carlson, C., Kuzniecky, R., Quinn, B. T., Alper, K.,
of long-term potentiation within and across the apical and basilar dendritic et al. (2011). Abnormalities of cortical thickness in postictal psychosis. Epilepsy
compartments of CA1 pyramidal neurons: synaptic tagging is compartment Behav. 21, 132–136. doi: 10.1016/j.yebeh.2011.03.024
restricted. J. Neurosci. 26, 256–264. doi: 10.1523/JNEUROSCI.3196-05.2006 Fournier, N. M., Botterill, J. J., Marks, W. N., Guskjolen, A. J., and Kalynchuk, L. E.
Alvarez, P. S., Simao, F., Hemb, M., Xavier, L. L., and Nunes, M. L. (2014). (2013). Impaired recruitment of seizure-generated neurons into functional
Effects of undernourishment, recurrent seizures and enriched environment memory networks of the adult dentate gyrus following long-term amygdala
during early life in hippocampal morphology. Int. J. Dev. Neurosci. 33, 81–87. kindling. Exp. Neurol. 244, 96–104. doi: 10.1016/j.expneurol.2012.11.031
doi: 10.1016/j.ijdevneu.2013.12.004 Gutierrez-Galve, L., Flugel, D., Thompson, P. J., Koepp, M. J., Symms, M. R.,
Arnold, C. L., Rademaker, A., Wolf, M. S., Liu, D., Lucas, G., Hancock, J., et al. Ron, M. A., et al. (2012). Cortical abnormalities and their cognitive correlates
(2016). Final results of a 3-year literacy-informed intervention to promote in patients with temporal lobe epilepsy and interictal psychosis. Epilepsia 53,
annual fecal occult blood test screening. J. Community Health 41, 724–731. 1077–1087. doi: 10.1111/j.1528-1167.2012.03504.x
doi: 10.1007/s10900-015-0146-6 Halász, P. (2010a). The concept of epileptic networks. Part 1. Ideggyogy. Sz. 63,
Bandopadhyay, R., Liu, J. Y., Sisodiya, S. M., and Thom, M. (2014). A comparative 293–303.
study of the dentate gyrus in hippocampal sclerosis in epilepsy and dementia. Halász, P. (2010b). The concept of epileptic networks. Part 2. Ideggyogy. Sz. 63,
Neuropathol. Appl. Neurobiol. 40, 177–190. doi: 10.1111/nan.12087 377–384.
Bartsch, T., Schönfeld, R., Müller, F. J., Alfke, K., Leplow, B., Aldenhoff, J., et al. Hellwig, S., Gutmann, V., Trimble, M. R., and van Elst, L. T. (2013). Cerebellar
(2010). Focal lesions of human hippocampal CA1 neurons in transient global volume is linked to cognitive function in temporal lobe epilepsy: a quantitative
amnesia impair place memory. Science 328, 1412–1415. doi: 10.1126/science. MRI study. Epilepsy Behav. 28, 156–162. doi: 10.1016/j.yebeh.2013.04.020
1188160 Hermann, B. P., Lin, J. J., Jones, J. E., and Seidenberg, M. (2009). The emerging
Berkovic, S. F., McIntosh, A. M., Kalnins, R. M., Jackson, G. D., Fabinyi, G. C., architecture of neuropsychological impairment in epilepsy. Neurol. Clin. 27,
Brazenor, G. A., et al. (1995). Preoperative MRI predicts outcome of temporal 881–907. doi: 10.1016/j.ncl.2009.08.001
lobectomy: an actuarial analysis. Neurology 45, 1358–1363. doi: 10.1212/WNL. Hermann, B., Seidenberg, M., and Jones, J. (2008). The neurobehavioural
45.7.1358 comorbidities of epilepsy: can a natural history be developed? Lancet Neurol.
Bernhardt, B. C., Chen, Z., He, Y., Evans, A. C., and Bernasconi, N. (2011). Graph- 7, 151–160. doi: 10.1016/s1474-4422(08)70018-8
theoretical analysis reveals disrupted small-world organization of cortical Hernan, A. E., Holmes, G. L., Isaev, D., Scott, R. C., and Isaeva, E. (2013). Altered
thickness correlation networks in temporal lobe epilepsy. Cereb. Cortex 21, short-term plasticity in the prefrontal cortex after early life seizures. Neurobiol.
2147–2157. doi: 10.1093/cercor/bhq291 Dis. 50, 120–126. doi: 10.1016/j.nbd.2012.10.007
Bernhardt, B. C., Worsley, K. J., Besson, P., Concha, L., Lerch, J. P., Evans, A. C., Horstmann, M. T., Bialonski, S., Noennig, N., Mai, H., Prusseit, J., Wellmer, J.,
et al. (2008). Mapping limbic network organization in temporal lobe et al. (2010). State dependent properties of epileptic brain networks:
epilepsy using morphometric correlations: insights on the relation between comparative graph-theoretical analyses of simultaneously recorded EEG and
mesiotemporal connectivity and cortical atrophy. Neuroimage 42, 515–524. MEG. Clin. Neurophysiol. 121, 172–185. doi: 10.1016/j.clinph.2009.10.013
doi: 10.1016/j.neuroimage.2008.04.261 Isaeva, E., Isaev, D., Khazipov, R., and Holmes, G. L. (2006). Selective impairment
Bettus, G., Guedj, E., Joyeux, F., Confort-Gouny, S., Soulier, E., Laguitton, V., of GABAergic synaptic transmission in the flurothyl model of neonatal
et al. (2009). Decreased basal fMRI functional connectivity in epileptogenic seizures. Eur. J. Neurosci. 23, 1559–1566. doi: 10.1111/j.1460-9568.2006.
networks and contralateral compensatory mechanisms. Hum. Brain Mapp. 30, 04693.x
1580–1591. doi: 10.1002/hbm.20625 Kaaden, S., Quesada, C. M., Urbach, H., Koenig, R., Weber, B., Schramm, J., et al.
Blumcke, I., Coras, R., Miyata, H., and Ozkara, C. (2012). Defining clinico- (2011). Neurodevelopmental disruption in early-onset temporal lobe epilepsy:
neuropathological subtypes of mesial temporal lobe epilepsy with hippocampal evidence from a voxel-based morphometry study. Epilepsy Behav. 20, 694–699.
sclerosis. Brain Pathol. 22, 402–411. doi: 10.1111/j.1750-3639.2012.00583.x doi: 10.1016/j.yebeh.2011.02.020
Bonilha, L., Tabesh, A., Dabbs, K., Hsu, D. A., Stafstrom, C. E., Hermann, B. P., Karmonik, C., Dulay, M., Verma, A., Yen, C., and Grossman, R. G. (2010). Brain
et al. (2014). Neurodevelopmental alterations of large-scale structural networks activation in complex partial seizures during switching from a the goal-directed
in children with new-onset epilepsy. Hum. Brain Mapp. 35, 3661–3672. task to a resting state: comparison of fMRI maps to the default mode network.
doi: 10.1002/hbm.22428 Conf. Proc. IEEE Eng. Med. Biol. Soc. 2010, 5685–5688. doi: 10.1109/IEMBS.
Caeyenberghs, K., Powell, H. W., Thomas, R. H., Brindley, L., Church, C., Evans, J., 2010.5627883
et al. (2015). Hyperconnectivity in juvenile myoclonic epilepsy: a network Karnam, H. B., Zhou, J. L., Huang, L. T., Zhao, Q., Shatskikh, T., and
analysis. Neuroimage Clin. 7, 98–104. doi: 10.1016/j.nicl.2014.11.018 Holmes, G. L. (2009). Early life seizures cause long-standing impairment of the
Casanova, J. R., Nishimura, M., Owens, J. W., and Swann, J. W. (2012). Impact of hippocampal map. Exp. Neurol. 217, 378–387. doi: 10.1016/j.expneurol.2009.
seizures on developing dendrites: implications for intellectual developmental 03.028
disabilities. Epilepsia 53, 116–124. doi: 10.1111/j.1528-1167.2012.03482.x Keller, S. S., Baker, G., Downes, J. J., and Roberts, N. (2009). Quantitative MRI
Casanova, J. R., Nishimura, M., and Swann, J. W. (2014). The effects of early-life of the prefrontal cortex and executive function in patients with temporal lobe
seizures on hippocampal dendrite development and later-life learning and epilepsy. Epilepsy Behav. 15, 186–195. doi: 10.1016/j.yebeh.2009.03.005
memory. Brain Res. Bull. 103, 39–48. doi: 10.1016/j.brainresbull.2013.10.004 Kellermann, T. S., Bonilha, L., Eskandari, R., Garcia-Ramos, C., Lin, J. J., and
Cho, K. O., Lybrand, Z. R., Ito, N., Brulet, R., Tafacory, F., Zhang, L., et al. (2015). Hermann, B. P. (2016). Mapping the neuropsychological profile of temporal
Aberrant hippocampal neurogenesis contributes to epilepsy and associated lobe epilepsy using cognitive network topology and graph theory. Epilepsy
cognitive decline. Nat. Commun. 6:6606. doi: 10.1038/ncomms7606 Behav. 63, 9–16. doi: 10.1016/j.yebeh.2016.07.030
da Silva, E. A., Chugani, D. C., Muzik, O., and Chugani, H. T. (1997). Identification Kim, J. H., Suh, S. I., Park, S. Y., Seo, W. K., Koh, I., Koh, S. B., et al. (2012).
of frontal lobe epileptic foci in children using positron emission tomography. Microstructural white matter abnormality and frontal cognitive dysfunctions
Epilepsia 38, 1198–1208. doi: 10.1111/j.1528-1157.1997.tb01217.x in juvenile myoclonic epilepsy. Epilepsia 53, 1371–1378. doi: 10.1111/j.1528-
Danielson, N. B., Guo, J. N., and Blumenfeld, H. (2011). The default mode 1167.2012.03544.x
network and altered consciousness in epilepsy. Behav. Neurol. 24, 55–65. Kleen, J. K., Sesque, A., Wu, E. X., Miller, F. A., Hernan, A. E., Holmes, G. L.,
doi: 10.3233/BEN-2011-0310 et al. (2011a). Early-life seizures produce lasting alterations in the structure and
Deng, P. Y., Sojka, D., and Klyachko, V. A. (2011). Abnormal presynaptic function of the prefrontal cortex. Epilepsy Behav. 22, 214–219. doi: 10.1016/j.
short-term plasticity and information processing in a mouse model of fragile X yebeh.2011.07.022

Frontiers in Neuroanatomy | www.frontiersin.org 9 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

Kleen, J. K., Wu, E. X., Holmes, G. L., Scott, R. C., and Lenck-Santini, P. P. Preston, A. R., and Eichenbaum, H. (2013). Interplay of hippocampus and
(2011b). Enhanced oscillatory activity in the hippocampal-prefrontal prefrontal cortex in memory. Curr. Biol. 23, R764–R773. doi: 10.1016/j.cub.
network is related to short-term memory function after early-life 2013.05.041
seizures. J. Neurosci. 31, 15397–15406. doi: 10.1523/JNEUROSCI. Qiu, W., Yu, C., Gao, Y., Miao, A., Tang, L., Huang, S., et al. (2017). Disrupted
2196-11.2011 topological organization of structural brain networks in childhood absence
Koepp, M. J., Woermann, F., Savic, I., and Wandschneider, B. (2013). Juvenile epilepsy. Sci. Rep. 7:11973. doi: 10.1038/s41598-017-10778-0
myoclonic epilepsy—neuroimaging findings. Epilepsy Behav. 28, S40–S44. Raj, A., Mueller, S. G., Young, K., Laxer, K. D., and Weiner, M. (2010). Network-
doi: 10.1016/j.yebeh.2012.06.035 level analysis of cortical thickness of the epileptic brain. Neuroimage 52,
Kramer, M. A., Kolaczyk, E. D., and Kirsch, H. E. (2008). Emergent network 1302–1313. doi: 10.1016/j.neuroimage.2010.05.045
topology at seizure onset in humans. Epilepsy Res. 79, 173–186. doi: 10.1016/j. Riley, J. D., Franklin, D. L., Choi, V., Kim, R. C., Binder, D. K., Cramer, S. C., et al.
eplepsyres.2008.02.002 (2010). Altered white matter integrity in temporal lobe epilepsy: association
Kuroda, M., Yokofujita, J., and Murakami, K. (1998). An ultrastructural study of with cognitive and clinical profiles. Epilepsia 51, 536–545. doi: 10.1111/j.1528-
the neural circuit between the prefrontal cortex and the mediodorsal nucleus 1167.2009.02508.x
of the thalamus. Prog. Neurobiol. 54, 417–458. doi: 10.1016/s0301-0082(97) Rodríguez-Cruces, R., Velázquez-Pérez, L., Rodríguez-Leyva, I., Velasco, A. L.,
00070-1 Trejo-Martínez, D., Barrágan-Campos, H. M., et al. (2018). Association of
Lawson, J. A., Cook, M. J., Vogrin, S., Litewka, L., Strong, D., Bleasel, A. F., white matter diffusion characteristics and cognitive deficits in temporal lobe
et al. (2002). Clinical, EEG, and quantitative MRI differences in pediatric epilepsy. Epilepsy Behav. 79, 138–145. doi: 10.1016/j.yebeh.2017.11.040
frontal and temporal lobe epilepsy. Neurology 58, 723–729. doi: 10.1212/WNL. Saniya, K., Patil, B. G., Chavan, M. D., Prakash, K. G., Sailesh, K. S., Archana, R.,
58.5.723 et al. (2017). Neuroanatomical changes in brain structures related to cognition
Li, Q., Cao, W., Liao, X., Chen, Z., Yang, T., Gong, Q., et al. (2015). Altered resting in epilepsy: an update. J. Nat. Sci. Biol. Med. 8, 139–143. doi: 10.4103/0976-
state functional network connectivity in children absence epilepsy. J. Neurol. 9668.210016
Sci. 354, 79–85. doi: 10.1016/j.jns.2015.04.054 Schmeiser, B., Zentner, J., Prinz, M., Brandt, A., and Freiman, T. M. (2017).
Liang, J. S., Lee, S. P., Pulli, B., Chen, J. W., Kao, S. C., Tsang, Y. M., Extent of mossy fiber sprouting in patients with mesiotemporal lobe epilepsy
et al. (2016). Microstructural changes in absence seizure children: a diffusion correlates with neuronal cell loss and granule cell dispersion. Epilepsy Res. 129,
tensor magnetic resonance imaging study. Pediatr. Neonatol. 57, 318–325. 51–58. doi: 10.1016/j.eplepsyres.2016.11.011
doi: 10.1016/j.pedneo.2015.10.003 Schneider, C., Helmstaedter, C., Luders, E., Thompson, P. M., Toga, A. W.,
Lin, J. J., Riley, J. D., Hsu, D. A., Stafstrom, C. E., Dabbs, K., Becker, T., et al. (2012). Elger, C., et al. (2014). Relation of callosal structure to cognitive abilities
Striatal hypertrophy and its cognitive effects in new-onset benign epilepsy with in temporal lobe epilepsy. Front. Neurol. 5:16. doi: 10.3389/fneur.2014.
centrotemporal spikes. Epilepsia 53, 677–685. doi: 10.1111/j.1528-1167.2012. 00016
03422.x Seltzer, B., and Pandya, D. N. (2009). Posterior cingulate and retrosplenial cortex
Luo, C., Li, Q., Lai, Y., Xia, Y., Qin, Y., Liao, W., et al. (2011). Altered functional connections of the caudal superior temporal region in the rhesus monkey. Exp.
connectivity in default mode network in absence epilepsy: a resting-state fMRI Brain Res. 195, 325–334. doi: 10.1007/s00221-009-1795-4
study. Hum. Brain Mapp. 32, 438–449. doi: 10.1002/hbm.21034 Sonmez, F., Atakli, D., Sari, H., Atay, T., and Arpaci, B. (2004). Cognitive function
Maglóczky, Z. (2010). Sprouting in human temporal lobe epilepsy: excitatory in juvenile myoclonic epilepsy. Epilepsy Behav. 5, 329–336. doi: 10.1016/j.
pathways and axons of interneurons. Epilepsy Res. 89, 52–59. doi: 10.1016/j. yebeh.2004.01.007
eplepsyres.2010.01.002 Stam, C. J., and van Straaten, E. C. (2012). The organization of physiological
Mancuso, J. J., Cheng, J., Yin, Z., Gilliam, J. C., Xia, X., Li, X., et al. (2014). brain networks. Clin. Neurophysiol. 123, 1067–1087. doi: 10.1016/j.clinph.2012.
Integration of multiscale dendritic spine structure and function data into 01.011
systems biology models. Front. Neuroanat. 8:130. doi: 10.3389/fnana.2014. Takaya, S., Mikuni, N., Mitsueda, T., Satow, T., Taki, J., Kinoshita, M., et al. (2009).
00130 Improved cerebral function in mesial temporal lobe epilepsy after subtemporal
Martinian, L., Catarino, C. B., Thompson, P., Sisodiya, S. M., and Thom, M. (2012). amygdalohippocampectomy. Brain 132, 185–194. doi: 10.1093/brain/
Calbindin D28K expression in relation to granule cell dispersion, mossy fibre awn218
sprouting and memory impairment in hippocampal sclerosis: a surgical and Thom, M. (2009). Hippocampal sclerosis: progress since Sommer. Brain Pathol.
post mortem series. Epilepsy Res. 98, 14–24. doi: 10.1016/j.eplepsyres.2011. 19, 565–572. doi: 10.1111/j.1750-3639.2008.00201.x
08.011 Thom, M., Liagkouras, I., Martinian, L., Liu, J., Catarino, C. B., and Sisodiya, S. M.
Mason, R. A., Prat, C. S., and Just, M. A. (2014). Neurocognitive brain response (2012). Variability of sclerosis along the longitudinal hippocampal axis in
to transient impairment of Wernicke’s area. Cereb. Cortex 24, 1474–1484. epilepsy: a post mortem study. Epilepsy Res. 102, 45–59. doi: 10.1016/j.
doi: 10.1093/cercor/bhs423 eplepsyres.2012.04.015
Minjarez, B., Camarena, H. O., Haramati, J., Rodríguez-Yañez, Y., Mena- Trotta, N., Goldman, S., Legros, B., Baete, K., Van Laere, K., Van Bogaert, P.,
Munguía, S., Buriticá, J., et al. (2017). Behavioral changes in models of et al. (2013). Changes in functional integration with the non-epileptic temporal
chemoconvulsant-induced epilepsy: a review. Neurosci. Biobehav. Rev. 83, lobe of patients with unilateral mesiotemporal epilepsy. PLoS One 8:e67053.
373–380. doi: 10.1016/j.neubiorev.2017.10.016 doi: 10.1371/journal.pone.0067053
Mlsna, L. M., and Koh, S. (2013). Maturation-dependent behavioral deficits and Trotta, N., Goldman, S., Legros, B., Ligot, N., Guerry, N., Baete, K., et al. (2011).
cell injury in developing animals during the subacute postictal period. Epilepsy Metabolic evidence for episodic memory plasticity in the nonepileptic temporal
Behav. 29, 190–197. doi: 10.1016/j.yebeh.2013.07.018 lobe of patients with mesial temporal epilepsy. Epilepsia 52, 2003–2012.
Mueller, S. G., Laxer, K. D., Barakos, J., Cheong, I., Finlay, D., Garcia, P., et al. doi: 10.1111/j.1528-1167.2011.03271.x
(2010). Involvement of the thalamocortical network in TLE with and without Tuchscherer, V., Seidenberg, M., Pulsipher, D., Lancaster, M., Guidotti, L., and
mesiotemporal sclerosis. Epilepsia 51, 1436–1445. doi: 10.1111/j.1528-1167. Hermann, B. (2010). Extrahippocampal integrity in temporal lobe epilepsy and
2009.02413.x cognition: thalamus and executive functioning. Epilepsy Behav. 17, 478–482.
Nishimura, M., Gu, X., and Swann, J. W. (2011). Seizures in early life suppress doi: 10.1016/j.yebeh.2010.01.019
hippocampal dendrite growth while impairing spatial learning. Neurobiol. Dis. Vaessen, M. J., Jansen, J. F., Vlooswijk, M. C., Hofman, P. A., Majoie, H. J.,
44, 205–214. doi: 10.1016/j.nbd.2011.07.002 Aldenkamp, A. P., et al. (2012). White matter network abnormalities are
O’Muircheartaigh, J., and Richardson, M. P. (2012). Epilepsy and the frontal lobes. associated with cognitive decline in chronic epilepsy. Cereb. Cortex 22,
Cortex 48, 144–155. doi: 10.1016/j.cortex.2011.11.012 2139–2147. doi: 10.1093/cercor/bhr298
O’Muircheartaigh, J., Vollmar, C., Barker, G. J., Kumari, V., Symms, M. R., van Dellen, E., Douw, L., Hillebrand, A., Ris-Hilgersom, I. H., Schoonheim, M. M.,
Thompson, P., et al. (2011). Focal structural changes and cognitive dysfunction Baayen, J. C., et al. (2012). MEG network differences between low- and
in juvenile myoclonic epilepsy. Neurology 76, 34–40. doi: 10.1212/WNL. high-grade glioma related to epilepsy and cognition. PLoS One 7:e50122.
0b013e318203e93d doi: 10.1371/journal.pone.0050122

Frontiers in Neuroanatomy | www.frontiersin.org 10 April 2018 | Volume 12 | Article 25

Wu et al. Networks/Topology Alteration and Cognition

van Diessen, E., Diederen, S. J., Braun, K. P., Jansen, F. E., and Stam, C. J. (2013). Widjaja, E., Zamyadi, M., Raybaud, C., Snead, O. C., and Smith, M. L.
Functional and structural brain networks in epilepsy: what have we learned? (2013b). Impaired default mode network on resting-state FMRI in children
Epilepsia 54, 1855–1865. doi: 10.1111/epi.12350 with medically refractory epilepsy. Am. J. Neuroradiol. 34, 552–557.
van Elburg, R. A., and van Ooyen, A. (2010). Impact of dendritic size and dendritic doi: 10.3174/ajnr.A3265
topology on burst firing in pyramidal cells. PLoS Comput. Biol. 6:e1000781. Wong, M., and Guo, D. (2013). Dendritic spine pathology in epilepsy: cause
doi: 10.1371/journal.pcbi.1000781 or consequence? Neuroscience 251, 141–150. doi: 10.1016/j.neuroscience.2012.
Vlooswijk, M. C., Vaessen, M. J., Jansen, J. F., de Krom, M. C., Majoie, H. J., 03.048
Hofman, P. A., et al. (2011). Loss of network efficiency associated with cognitive Yang, T., Guo, Z., Luo, C., Li, Q., Yan, B., Liu, L., et al. (2012). White
decline in chronic epilepsy. Neurology 77, 938–944. doi: 10.1212/WNL. matter impairment in the basal ganglia-thalamocortical circuit of drug-naive
0b013e31822cfc2f childhood absence epilepsy. Epilepsy Res. 99, 267–273. doi: 10.1016/j.
Voets, N. L., Bernhardt, B. C., Kim, H., Yoon, U., and Bernasconi, N. (2011). eplepsyres.2011.12.006
Increased temporolimbic cortical folding complexity in temporal lobe epilepsy. Yang, X. D., Liao, X. M., Uribe-Marino, A., Liu, R., Xie, X. M., Jia, J.,
Neurology 76, 138–144. doi: 10.1212/WNL.0b013e318205d521 et al. (2015). Stress during a critical postnatal period induces region-specific
Vollmar, C., O’Muircheartaigh, J., Symms, M. R., Barker, G. J., Thompson, P., structural abnormalities and dysfunction of the prefrontal cortex via CRF1.
Kumari, V., et al. (2012). Altered microstructural connectivity in Neuropsychopharmacology 40, 1203–1215. doi: 10.1038/npp.2014.304
juvenile myoclonic epilepsy: the missing link. Neurology 78, 1555–1559. Zeng, H., Pizarro, R., Nair, V. A., La, C., and Prabhakaran, V. (2013). Alterations
doi: 10.1212/WNL.0b013e3182563b44 in regional homogeneity of resting-state brain activity in mesial temporal lobe
Wandschneider, B., Thompson, P. J., Vollmar, C., and Koepp, M. J. (2012). Frontal epilepsy. Epilepsia 54, 658–666. doi: 10.1111/epi.12066
lobe function and structure in juvenile myoclonic epilepsy: a comprehensive Zhang, L., and Luo, X. P. (2011). Plasticity and metaplasticity of lateral perforant
review of neuropsychological and imaging data. Epilepsia 53, 2091–2098. path in hippocampal dentate gyrus in a rat model of febrile seizure. Sheng Li
doi: 10.1111/epi.12003 Xue Bao 63, 124–130.
Wang, Z., Dai, Z., Gong, G., Zhou, C., and He, Y. (2015). Understanding Zhang, Z., Lu, G., Zhong, Y., Tan, Q., Liao, W., Wang, Z., et al. (2010). Altered
structural-functional relationships in the human brain: a large-scale network spontaneous neuronal activity of the default-mode network in mesial temporal
perspective. Neuroscientist 21, 290–305. doi: 10.1177/1073858414537560 lobe epilepsy. Brain Res. 1323, 152–160. doi: 10.1016/j.brainres.2010.01.042
Wang, Z., Tian, C., Dhamala, M., and Liu, Z. (2017). A small change in neuronal
network topology can induce explosive synchronization transition and activity Conflict of Interest Statement: The authors declare that the research was
propagation in the entire network. Sci. Rep. 7:561. doi: 10.1038/s41598-017- conducted in the absence of any commercial or financial relationships that could
00697-5 be construed as a potential conflict of interest.
Wickens, S., Bowden, S. C., and D’Souza, W. (2017). Cognitive functioning in
children with self-limited epilepsy with centrotemporal spikes: a systematic The reviewer JIA declared a shared affiliation, though no other collaboration, with
review and meta-analysis. Epilepsia 58, 1673–1685. doi: 10.1111/epi. two of the authors CWZ and LF to the handling Editor.
13865
Widjaja, E., Mahmoodabadi, S. Z., Snead, O. C. III., Almehdar, A., and Smith, M. L. Copyright © 2018 Wu, Zhao, Long, Xiao and Feng. This is an open-access article
(2011). Widespread cortical thinning in children with frontal lobe epilepsy. distributed under the terms of the Creative Commons Attribution License (CC BY).
Epilepsia 52, 1685–1691. doi: 10.1111/j.1528-1167.2011.03085.x The use, distribution or reproduction in other forums is permitted, provided the
Widjaja, E., Skocic, J., Go, C., Snead, O. C., Mabbott, D., and Smith, M. L. (2013a). original author(s) and the copyright owner are credited and that the original
Abnormal white matter correlates with neuropsychological impairment publication in this journal is cited, in accordance with accepted academic practice.
in children with localization-related epilepsy. Epilepsia 54, 1065–1073. No use, distribution or reproduction is permitted which does not comply with these
doi: 10.1111/epi.12208 terms.

Frontiers in Neuroanatomy | www.frontiersin.org 11 April 2018 | Volume 12 | Article 25