Brain network correlates of epilepsy occurrence in multiple sclerosis and neuroinflammation : Neural Regeneration Research

Secondary Logo

Journal Logo


Brain network correlates of epilepsy occurrence in multiple sclerosis and neuroinflammation

Ciolac, Dumitru; Gonzalez-Escamilla, Gabriel; Winter, Yaroslav; Fleischer, Vinzenz; Grothe, Matthias; Groppa, Sergiu*

Author Information
Neural Regeneration Research 18(8):p 1717-1718, August 2023. | DOI: 10.4103/1673-5374.363188
  • Open

Multiple sclerosis (MS), the most common inflammatory condition of the central nervous system in young adults, is characterized by immune-mediated demyelination and neurodegeneration that translate into heterogeneous clinical phenotypes and disease trajectories. Although focal demyelinating lesions within the white matter are the hallmark of MS pathology, a large amount of lesions has been detected in both cortical and subcortical grey matter tissue. Besides focal pathology, diffuse inflammation and axonal damage are increasingly recognized in normal appearing white matter, as well as grey matter. Among various clinical manifestations, patients with MS may experience epileptic seizures, which can emerge at any time point throughout the disease course. Several clinical factors such as earlier onset of MS, longer disease duration, and higher disability have been related to the increased prevalence of epilepsy in patients with MS (Neuß et al., 2020; Grothe et al., 2021). Acute seizures and epilepsy were also reported in other neuroinflammatory disorders of the central nervous system, e.g., in myelin oligodendrocyte glycoprotein antibody disease, acute disseminated encephalomyelitis or neuromyelitis optica spectrum disorder. A few available neuroimaging and neuropathological studies suggested that the extent of cortical grey matter damage is proportional to the risk of epilepsy occurrence in patients with MS (Calabrese et al., 2012; Nicholas et al., 2016). However, despite the significant progress achieved in elucidating the molecular and cellular basis of MS pathology, mechanisms of increased susceptibility to seizure generation in acute and chronic neuroinflammation are poorly understood, thereby, leaving many questions open. Only a few to list: what are the brain structural fingerprints of epileptic seizures in MS? Is the location of tissue damage within a particular brain region (e.g., hippocampus or thalamus) is critical for initiating epilepsy? Is there a specific “network correlate” of hyperexcitable neuronal circuits and do they render networks vulnerable to ongoing MS-mediated damage? And finally, what are the main “culprit mechanisms” responsible for ictogenesis and epileptogenesis in MS? Based on the aforementioned questions, the overarching aim of this perspective article was to portray the brain network correlates of epilepsy occurrence in MS. First, we describe structural abnormalities of brain tissue associated with a higher susceptibility of epilepsy occurrence. Afterwards, we highlight the network alterations that are linked to epileptogenesis in patients with MS and explain the candidate molecular mechanisms underlying network hyperexcitability. Finally, we provide a conceptual background for future studies.

Several brain structural fingerprints have been found to be linked to epilepsy occurrence in patients with MS: intracortical lesions, regional cortical atrophy, and cortical microstructural alterations (Figure 1A). Previous and recent studies unequivocally showed that MS patients with concomitant epilepsy have a higher lesion burden within the cortical grey matter compartment in comparison to MS patients without epilepsy (Calabrese et al., 2012; Ciolac et al., 2022). Notably, intracortical lesions have a higher rate of accumulation over time in MS patients with epilepsy (Calabrese et al., 2012). The causes behind these observations are unclear but a magnetic resonance imaging-based phenotype of predominant cortical damage can be assumed. It is remarkable that in MS patients with epilepsy the distribution of intracortical lesions across the cortical compartment is not uniform but rather shows a clear predilection for particular anatomical areas. The temporal lobe, particularly the hippocampal formation, is the most heavily affected region by lesions in MS patients with epilepsy (Calabrese et al., 2017; Ciolac et al., 2022). Even within the hippocampal formation, the lesions tend to accumulate within specific hippocampal subfields such as subiculum, hippocampal tail, and hippocampus-amygdala transition area (Ciolac et al., 2022). The role of hippocampal damage in MS-induced epileptogenesis is supported by experimental mouse data. In cuprizone-treated mice, a toxic demyelination model, extensive hippocampal demyelination in association with overt seizure activity has been described (Lapato et al., 2017). Other temporal lobe areas, among which parahippocampal, entorhinal and fusiform cortices and temporal pole were also shown to have high lesion loads (Calabrese et al., 2017; Ciolac et al., 2022).

Figure 1:
Schematic representation of brain pathological alterations and associated molecular mechanisms contributing to epileptogenesis in MS-related neuroinflammation and neurodegeneration.(A) Cortical demyelinating lesions, degeneration of cortical inhibitory neurons, alterations of cortical microstructure, and hippocampal demyelination and neuronal cell loss are the most relevant structural determinants of ictogenesis and epileptogenesis in MS-related neuroinflammation and neurodegeneration. Loss of grey matter integrity leads to network remodeling and increased vulnerability of brain networks that favor and maintain epileptogenesis. (B) Upon immune-mediated demyelination, infiltration of peripheral blood inflammatory cells through the leaky blood-brain barrier as well as activation of resident cells (i.e., microglia and astrocytes) results in proinflammatory cytokine release and upregulation of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) activity. The neurodegenerative component of epileptogenesis is driven by astrogliosis with excitatory neurotransmitter (interleukin-1/6 [IL-1/6] and tumor necrosis factor [TNF]) and ion (Ca2+) dysregulation, and microgliosis with proinflammatory cytokine release and altered neuronal excitability. Loss of cortical gamma-aminobutyric acid (GABA)-ergic and hippocampal cornu ammonis (CA) 1 inhibitory neurons leads to an imbalance between excitatory and inhibitory processes, hence to hyperexcitability and epileptogenesis.

In MS patients with epilepsy, localized cortical atrophy is mostly detected in temporal (parahippocampal, entorhinal, fusiform) and parietal (isthmus and posterior cingulate) lobes of the brain (Calabrese et al., 2017; Ciolac et al., 2022). Spatial colocalization of lesions and atrophy within the temporal lobe areas may suggest that grey matter loss is secondary to a higher amount of grey matter lesions. However, the opposite scenario of epilepsy emergence from areas with a more advanced cortical atrophy should not be underestimated, since it is well established that e.g. temporal lobe seizures in hippocampal sclerosis originate from atrophied hippocampal subfields.

Loss of microstructural grey matter integrity within the cortical compartment might be as well considered one of brain structural correlates of epilepsy. Abnormal diffusion parameters – high volume fraction of the isotropic compartment and low intracellular volume fraction are attested in the hippocampus, parahippocampal, superior/middle/inferior temporal, fusiform, and cingulate cortices of MS patients with epilepsy (Calabrese et al., 2017). The contribution of cortical normal appearing grey matter in epileptogenesis is still controversial as it may just reflect the ongoing neuroaxonal degeneration. Nevertheless, low intracellular volume fraction of the inferior temporal gyrus (along with its cortical thickness) was found to be one of the significant predictors of epilepsy development in this patient population (Calabrese et al., 2017).

Immune-mediated damage and neurodegenerative processes induce early and widespread brain network alterations in patients with MS (Fleischer et al., 2019; Groppa et al., 2021). These network responses are considered to act as compensatory (or adaptive) mechanisms aimed to maintain, yet to a limited extent, an efficient communication within network circuits (Fleischer et al., 2019; Groppa et al., 2021). Both MS patients with and without epilepsy display a more segregated topology (i.e., greater divisibility of a network into smaller subunits) of structural covariance networks compared to healthy subjects, however, with an even more exaggerated network response in MS patients with epilepsy (Ciolac et al., 2022). One may speculate that the identified network pattern in MS patients with epilepsy reflects the shift from an adaptive to a maladaptive response. This maladaptive response may potentiate neuroaxonal damage due to an increased neural activity and thereby render neural networks even more vulnerable to perpetuating MS-driven damage (Groppa et al., 2021). Indeed, MS patients with epilepsy have a higher network vulnerability (i.e., higher sensitivity of the network to lesion effects) compared to MS patients without epilepsy (Ciolac et al., 2022). Increased network vulnerability may also stem from the preferential damage of the hub regions, i.e., of highly interconnected regions that play a central role in mediating information flows (Fleischer et al., 2022), and indicates the depletion of potential backup routes in the parallel organization of neuronal pathways. Hence, in patients with MS network vulnerability may be either the cause or consequence of epilepsy or even a combination of both – network vulnerability facilitates seizures, which in their turn induce excitotoxic neuroaxonal injury with loss of backup routes and hence to vulnerability. This potential scenario is corroborated by experimental models. Cuprizone-induced degeneration of hippocampal CA1 interneurons leads to the generation of seizures, which further aggravate the integrity of intrinsic hippocampal networks and hence increase vulnerability to neuroinflammatory and neurodegenerative processes (Lapato et al., 2017). On this ground, it might be postulated that increased network vulnerability is one of the “network correlates” of MS-induced epileptogenesis.

Overall, inflammatory lesions, neuronal cell loss, and axonal demyelination were proposed as driving mechanisms (or “culprit mechanisms”) of cortical hyperexcitability and epilepsy occurrence in MS (Nicholas et al., 2016). Upon neuroinflammation, a range of pathological phenomena, such as blood-brain barrier disruption, immune cell infiltration, overexpression of pro-inflammatory cytokines and chemokines, and glial cell dysfunction disturb the homeostatic environment of neurons (Figure 1B). Following infiltration into brain tissue, immune cells (T and B lymphocytes, macrophages) produce proinflammatory mediators such as interleukin-1 and interleukin-6 and tumor necrosis factor-alpha that induce phosphorylation of NR2 subunit of N-methyl-D-aspartate receptors with subsequent Ca2+ influx into neurons (Rayatpour et al., 2021; Li et al., 2022). Activated brain tissue-resident cells such as microglia and astrocytes as well secrete proinflammatory cytokines. The released cytokines alter the physiological functioning of voltage-gated channels and enhance the discharge of excitatory neurotransmitters (e.g., glutamate), thereby increasing neuronal excitability that is aggravated by IL-1-mediated inhibition of glutamate uptake by astrocytes. High amounts of glutamate can induce neuronal apoptosis by increasing the intracellular influx of Ca2+ (Rayatpour et al., 2021). Secretion of tumor necrosis factor-alpha leads to an up-regulation of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors and induction of gamma-aminobutyric acid (GABA) receptor endocytosis that cause excitatory overdrive of neurons (Figure 1B). Interferon-beta produced by activated microglia and reactive astrocytes is able to induce a hyper-excitation state with increased firing rates of hippocampal neurons that translates into high seizurogenic and epileptogenic potential (Rayatpour et al., 2021). The created proinflammatory milieu with emerging neuroaxonal damage inevitably leads to an abnormal cortical excitability within local network circuits as well as within distant interconnected networks, thereby serving as premises for an enduring predisposition of seizure generation. But is the immune-mediated inflammation alone able to trigger epileptic seizures? Perhaps not. Inhibitory neuronal cell loss (as an indicator of reduced inhibitory drive) is a critical prerequisite for MS-induced epileptogenesis. Selective loss of GABA-ergic inhibitory interneurons is evidenced in cortical layers IV and VI of the temporal lobe in patients with MS and concomitant epilepsy (Nicholas et al., 2016). Focal loss of inhibitory interneurons was related to type I cortical lesions (also known as leukocortical lesions), which could be the source of inflammatory cells responsible for neuronal injury (Nicholas et al., 2016). A similar pattern of selective loss of GABA-ergic inhibitory neurons is observed in the CA1 hippocampal subfield upon cuprizone-induced demyelination (Lapato et al., 2017). Loss of inhibitory interneurons occurs along with diffuse astrogliosis and alterations in aquaporin-4 expression (Lapato et al., 2017). Interestingly, cuprizone-induced acute generalized demyelination in a genetic mouse model of absence epilepsy results in a lower number of epileptic discharges, increased cortical theta oscillations with concomitant reduction of thalamic rhythmic burst activity (Chaudhary et al., 2022). These apparently contradicting findings are likely to be related to the particularities of the used animal models.

Conclusion and future perspectives: Based on available clinical and experimental evidence, increased network segregation and network vulnerability may be considered network correlates of MS-associated epileptogenesis. Apparently, network vulnerability is at the same time the cause and consequence of epileptogenesis as recurrent seizures result in network vulnerability, which further promotes the generation of seizures. With the aid of advanced neuroimaging and network neuroscience tools, the development of individualized models predicting epilepsy occurrence in MS patients could become possible. Nevertheless, we must gain more robust insights into the underlying pathophysiological basis of the reciprocal interactions among neuroinflammation, neurodegeneration, and the development of epilepsy. Undoubtedly, both inflammatory and neurodegenerative processes contribute to epileptogenesis but the extent and magnitude of their influence varies depending on disease-specific trajectory. In addition to grey matter involvement and network vulnerability, other factors (e.g., the genetic makeup, environmental factors, sex, and comorbidities (Neuß et al., 2020)) are likely to mediate the risk of epilepsy in MS. Maladaptive myelination within the epileptogenic network circuits following recurrent seizures contributes to the progressive aggravation of seizures in animal models of generalized epilepsy (Knowles et al., 2022). Future work should define how MS-induced epileptic seizures impact myelin turnover and how the latter affects the functional properties of involved networks. Also, the influence of epilepsy on disease course and disease-related disability should be addressed in more detail (Grothe et al., 2021). Determining structural, electrophysiological and network correlates of increased seizure risk in chronic neuroinflammatory disorders is important to refine diagnostic algorithms and optimize therapeutic approaches.

This work was supported by the German Research Foundation (DFG) - SFB-TR-128, the Boehringer Ingelheim Fonds BIF-03, and the State University of Medicine and Pharmacy “Nicolae Testemitanu” (project codes 20.80009.8007.40 and 21.80013.8007.2B).

C-Editors: Zhao M, Zhao LJ, Li CH; T-Editor: Jia Y


1. Calabrese M, Grossi P, Favaretto A, Romualdi C, Atzori M, Rinaldi F, Perini P, Saladini M, Gallo P (2012) Cortical pathology in multiple sclerosis patients with epilepsy:a 3 year longitudinal study. J Neurol Neurosurg Psychiatry 83:49–54.
2. Calabrese M, Castellaro M, Bertoldo A, De Luca A, Pizzini F, Ricciardi G, Pitteri M, Zimatore S, Magliozzi R, Benedetti M (2017) Epilepsy in multiple sclerosis:The role of temporal lobe damage. Mult Scler 23:473–482.
3. Chaudhary R, Albrecht S, Datunashvili M, Cerina M, Lüttjohann A, Han Y, Narayanan V, Chetkovich DM, Ruck T, Kuhlmann T (2022) Modulation of pacemaker channel function in a model of thalamocortical hyperexcitability by demyelination and cytokines. Cereb Cortex 32:4397–4421.
4. Ciolac D, Gonzalez-Escamilla G, Winter Y, Melzer N, Luessi F, Radetz A, Fleischer V, Groppa SA, Kirsch M, Bittner S, Zipp F, Muthuraman M, Meuth SG, Grothe M, Groppa S (2022) Altered grey matter integrity and network vulnerability relate to epilepsy occurrence in patients with multiple sclerosis. Eur J Neurol 29:2309–2320.
5. Fleischer V, Radetz A, Ciolac D, Muthuraman M, Gonzalez-Escamilla G, Zipp F, Groppa S (2019) Graph theoretical framework of brain networks in multiple sclerosis:a review of concepts. Neuroscience 403:35–53.
6. Fleischer V, Ciolac D, Gonzalez-Escamilla G, Grothe M, Strauss S, Molina Galindo LS, Radetz A, Salmen A, Lukas C, Klotz L (2022) Subcortical volumes as early predictors of fatigue in multiple sclerosis. Ann Neurol 91:192–202.
7. Groppa S, Gonzalez-Escamilla G, Eshaghi A, Meuth SG, Ciccarelli O (2021) Linking immune-mediated damage to neurodegeneration in multiple sclerosis:could network-based MRI help?. Brain Commun 3:fcab237.
8. Grothe M, Ellenberger D, von Podewils F, Stahmann A, Rommer PS, Zettl UK (2021) Epilepsy as a predictor of disease progression in multiple sclerosis. Mult Scler 28:942–949.
9. Knowles JK, Xu H, Soane C, Batra A, Saucedo T, Frost E, Tam LT, Fraga D, Ni L, Villar K (2022) Maladaptive myelination promotes generalized epilepsy progression. Nat Neurosci 25:596–606.
10. Lapato AS, Szu JI, Hasselmann JP, Khalaj AJ, Binder DK, Tiwari-Woodruff SK (2017) Chronic demyelination-induced seizures. Neuroscience 346:409–422.
11. Li EC, Zheng Y, Cai MT, Lai QL, Fang GL, Du BQ, Shen CH, Zhang YX, Wu LJ, Ding MP (2022) Seizures and epilepsy in multiple sclerosis, aquaporin 4 antibody-positive neuromyelitis optica spectrum disorder, and myelin oligodendrocyte glycoprotein antibody-associated disease. Epilepsia 63:2173–2191.
12. Neuß F, von Podewils F, Wang ZI, Süße M, Zettl UK, Grothe M (2020) Epileptic seizures in multiple sclerosis:prevalence, competing causes and diagnostic accuracy. J Neurol 268:1721–1727.
13. Nicholas R, Magliozzi R, Campbell G, Mahad D, Reynolds R (2016) Temporal lobe cortical pathology and inhibitory GABA interneuron cell loss are associated with seizures in multiple sclerosis. Mult Scler 22:25–35.
14. Rayatpour A, Farhangi S, Verdaguer E, Olloquequi J, Ureña J, Auladell C, Javan M (2021) The cross talk between underlying mechanisms of multiple sclerosis and epilepsy may provide new insights for more efficient therapies. Pharmaceuticals 14:1031.
Copyright: © 2023 Neural Regeneration Research