Abstract

Prevention of epilepsy in patients at risk is an urgent global unmet need. In theory, the latent period between an epileptogenic brain insult and the onset of epilepsy may offer a therapeutic window to interfere with epileptogenesis. Numerous preclinical and a few clinical studies on antiepileptogenesis have been performed in the past 20-plus years. The vast majority of these studies used treatments with single, often highly selective drugs shortly after epileptogenic brain injuries, mostly without any success. The negative results may be due to the complex mechanisms of epileptogenesis, which complicate any strategy to interfere with this process. It was therefore proposed to apply principles of network pharmacology to the search for antiepileptogenic treatments. Here the outcome of preclinical studies using rationally chosen drug combinations for antiepileptogenesis is reviewed. Of 24 drug combinations that are discussed here, only four exerted persistent antiepileptogenic efficacy in rodent models of acquired epilepsy. For three of these effective combinations, clinically approved drugs were used, which would facilitate translation into clinical trials. The chapter also discusses future advancements in the search for antiepileptogenic drugs or drug combinations, including the subsequent use of in silico, in vitro, and in vivo platforms as well as “big data” mining approaches and machine learning.

Introduction

Epilepsy prevention is a crucial public health concern. According to the World Health Organization (WHO, 2019), at least a quarter of all epilepsies are preventable because people at risk can be easily identified. Prevention strategies can be categorized as primary and secondary prevention. Primary prevention avoids the disorder by reducing or eliminating underlying causes and risk factors (Thurman et al., 2018). Secondary prevention involves use of therapies early after an insult to limit the extent of brain injury or to interrupt epileptogenesis. This is an unmet clinical need (Pitkänen and Engel, 2014; WHO, 2019; Löscher, 2020). The present review will deal with novel preclinical strategies to develop antiepileptogenic therapies that may interrupt epileptogenesis following brain insults.

A variety of brain insults, including traumatic brain injury (TBI), stroke, infections, and status epilepticus (SE), may induce development of epilepsy by a process termed epileptogenesis (Pitkänen and Engel, 2014; Klein et al., 2018). This process typically takes place during a latent period of weeks to months to years between the brain insult and onset of epilepsy with spontaneous recurrent seizures (SRS), which may offer a therapeutic window of opportunity to prevent or modify development of epilepsy. However, as yet, therapies to prevent epilepsy do not exist. The latent period and thus the process of epileptogenesis is characterized by a variety of structural and functional brain alterations, including neuroinflammation, impairment of blood–brain barrier (BBB) integrity, neurodegeneration, gliosis, axonal and dendritic plasticity, molecular reorganization, epigenetic alterations, changes in neural circuits, all resulting in an epileptic network and neuronal hyperexcitability leading to unprovoked recurrent seizures (Pitkänen and Lukasiuk, 2009; Klein et al., 2018; Löscher, 2020).

The complexity of the epileptogenic brain alterations that eventually result in epilepsy complicates any strategy to interfere with epileptogenesis. We have therefore proposed to apply principles of network pharmacology to the search for antiepileptogenic treatments (Löscher et al., 2013). Network (or systems) pharmacology attempts to model the effects of drug action by simultaneously modulating multiple proteins in a network, using rationally chosen combinations of drugs (Hopkins, 2008; Ainsworth, 2011; Muhammad et al., 2018). Such a network strategy has become a new paradigm in drug discovery. Clinical translation of such strategy would benefit from repurposing of approved drugs that are currently used for other indications. Indeed, the application of network pharmacology in other therapeutic areas has indicated that drug combination therapy increases successful drug repositioning (Sun et al., 2016). Drug combination therapy using two-to-three compounds with different mechanisms of action can overcome several of the challenges of single-drug therapies, particularly in the case of complex diseases. The use of drugs in combination can produce a synergistic effect if each of the drugs impinges on a different target or signaling pathway that results in (1) reduction of required drug concentrations for each individual drug and (2), more importantly, an increased therapeutic efficacy (Hopkins, 2008; Ainsworth, 2011; Sun et al., 2016). Here I will review the efficacy of synergistic drug combinations in preclinical animal and in vitro models of epileptogenesis.

Single Drug versus Drug Combinations for Antiepileptogenesis

In the past 20-plus years, the vast majority of preclinical and clinical studies on antiepileptogenesis used treatment with single drugs shortly after epileptogenic brain injuries in an attempt to prevent epilepsy or modify its severity, mostly without any success (Pitkänen and Kubova, 2004; Pitkänen et al., 2009; Pitkänen and Wasterlain, 2009; Temkin, 2009; Löscher and Brandt, 2010; Pitkänen, 2010; Klein and Tyrlikova, 2017; Löscher, 2020). Limited clinical evidence suggests that statins (such as atorvastatin), vigabatrin, and levetiracetam may exert antiepileptogenic or disease-modifying effects in animal models and patients, but respective studies need to be confirmed (Klein et al., 2020). The negative outcome of most single-drug epilepsy prevention trials in preclinical models or patients and the complexity of epileptogenesis led to the suggestion that network or rational polypharmacy approaches may be more effective (Löscher et al., 2013; White and Löscher, 2014). In the last ~10 years, several preclinical studies on such approaches have been published (Table 67–1).

Table 67–1

Antiepileptogenic Efficacy of Drug Combinations in Different Rodent Models of TLE.

Efficacy of Drug Combination to Prevent or Modify the Development of Epilepsy

The principle for drug selection used for most of the studies, summarized in Table 67–1, was to combine drugs that act on different targets within the epileptogenic process. Except for two studies, all studies used post-SE models of temporal lobe epilepsy (TLE), the most common type of acquired epilepsy after brain injury. Overall, 24 combinations of 2–4 drugs were examined in such models, of which 6 combinations proved to be effective in preventing epilepsy in 30%–100% of the animals (Table 67–1). However, for interpretation of efficacy, it is important to discriminate between treatments that shortened the duration of the initial insult and treatments that were administered without impacting the duration or severity of the initial insult. It is generally accepted that initial insult modification should be differentiated from “true” antiepileptogenic efficacy (Löscher and Brandt, 2010; Galanopoulou et al., 2012), particularly because in patients it will be almost impossible to start treatment during the insult. Examples for potential initial insult modification are the combination studies of Francois et al. (2006) (diazepam and topiramate), Kwon et al. (2013) (two anti-inflammatory drugs), Brandt et al. (2015) (diazepam, phenobarbital, and scopolamine), and Marrero-Rosado et al. (2020) (midazolam and ketamine) (Table 67–1). The study of Brandt et al. (2015) was the first to demonstrate that pilocarpine-induced SE of 60-min duration is not sufficient to cause epilepsy if the SE is completely terminated, including prevention of SE recurrence. In the same paper, it was shown that increasing SE duration from 60 to 90 min leads to the induction of epilepsy in the majority of rats. Such methodological aspects are important to exclude false-positive findings on potential antiepileptogenic treatments.

Only four of the 24 studies on drug combinations summarized in Table 67–1 had a true and persistent antiepileptogenic effect. These studies include antioxidant therapy in the systemic rat kainate model of TLE (Shekh-Ahmad et al., 2019) and our studies in which we systematically evaluated a large series of rationally chosen multitargeted combinations of repurposed drugs, resulting in the identification of three effective drug combinations (Schidlitzki et al., 2020; Welzel et al., 2021). These three combinations are (1) levetiracetam and topiramate; (2) levetiracetam, topiramate, and gabapentin; and (3) levetiracetam, atorvastatin, and ceftriaxone (Table 67–1).

Systematic Evaluation of Drug Combinations for Antiepileptogenesis

The strategies that we used to discover effective antiepileptogenic drug combinations are shown in Figure 67–1. First, a literature review of hundreds of potentially interesting clinically approved drugs was performed to select drugs with relevant mechanism(s) of action and some evidence for an antiepileptogenic or disease-modifying effect in an epilepsy model. The next, most critical step was to decide which combinations of such drugs should be examined in vivo. For this purpose, we have taken two strategies: (1) combining potentially synergistic drugs based on mechanisms of action and (2) a computational in silico approach for network analysis, using the STITCH (Search Tool for Interacting Chemicals) database (http://stitch.embl.de/). STITCH integrates known and predicted interactions between proteins and small molecules for 430,000 chemicals and more than 9.6 million proteins from different species, including humans (Szklarczyk et al., 2016). STITCH allows both analyzing drug-drug interactions and the interactive effects of single drugs and drug combinations on protein networks. The drug combinations finally chosen were first examined for tolerability in naïve mice and mice after SE, as illustrated in Figure 67–1. Doses and dosing intervals for these studies were chosen from the literature and modified based on the tolerability of the drug combinations (Klee et al., 2015; Welzel et al., 2019). Twelve drug combinations were finally chosen for in vivo studies on antiepileptogenesis, using the intrahippocampal kainate mouse model of TLE (Fig. 67–1).

Figure 67–1.

Flow chart illustrating the drug selection process and subsequent forming of drug combinations that were studied in silico and in vivo. Literature research was performed by PubMed (National Library of Medicine, NLM, USA) and the author’s own (more...)

Three Drug Combinations Stand Out in Their Antiepileptogenic Efficacy

As described above, three effective combinations were identified by the strategies shown in Figure 67–1: (1) levetiracetam and topiramate; (2) levetiracetam, topiramate, and gabapentin; and (3) levetiracetam, atorvastatin, and ceftriaxone (Table 67–1). All drug combinations were administered t.i.d. over 5 days, starting 6 h after kainate injection. One-week periods of continuous (24/7) video-EEG monitoring of SRS were started at 3 weeks and 3 months after drug withdrawal to avoid any carryover effects of the treatment on occurrence of SRS. As shown in Figure 67–2A,B, these combinations reduced the incidence and/or frequency of spontaneous recurrent electroclinical seizures, recorded by continuous video-EEG monitoring at 3 months after kainate. The most effective drug combination was levetiracetam, atorvastatin, and ceftriaxone, which completely prevented electroclinical SRS recorded at 3 months after kainate (Welzel et al., 2021). Interestingly, when we tested two different dose regimens for levetiracetam, atorvastatin, and ceftriaxone, only the low-dose regimen was effective (Fig. 67–2A,B), whereas the high-dose regimen was ineffective and, in fact, exerted proepileptogenic effects. This finding is in line with the principles of network pharmacology, because a synergistic drug combination should lead to a reduction of required drug doses for each individual drug (Sun et al., 2016).

Figure 67–2.

Antiepileptogenic efficacy of the three drug combinations that were discovered by the strategies illustrated in Figure 67–1. Overall, 12 drug combinations were tested in vivo in the intrahippocampal kainate mouse model of TLE (Klee et al., 2015; (more...)

For all of the combinations illustrated in Figure 67–2A,B, STITCH indicated significant drug-drug protein network interactions. To examine whether these drug combinations also exerted synergistic interactions in vivo, we determined the antiepileptogenic efficacy of monotherapy versus double or triple therapy, as shown in Figure 67–2C. Neither levetiracetam nor topiramate alone exerted any antiepileptogenic effect (Schidlitzki et al., 2020). Furthermore, in terms of focal electrographic (nonconvulsive) seizures, the most frequent type of SRS in the intrahippocampal kainate mouse model, only the triple combination of levetiracetam, topiramate, and gabapentin significantly reduced the incidence of these seizures, while a double combination (levetiracetam, topiramate) or monotherapy was ineffective (Fig. 67–2C), indicating a synergistic effect of the triple combination.

Potential Mechanisms of Effective Drug Combinations

Different approaches were chosen to determine the potential mechanisms of the drug combinations shown in Figure 67–2, using the levetiracetam plus topiramate combination for proof-of-concept (Schidlitzki et al., 2020). As shown in Figure 67–3, analysis of this drug combination by STITCH for human brain and mouse (for which tissue-specific networks are not available yet) indicates both a significant drug-drug interaction (shown by the red line between drugs) and interactions between the two drugs via the proteins that they affect, resulting in a complex drug-drug protein network interaction. Interactions occur via FAS (TNF receptor superfamily member 6) and caspase 3 as well as via voltage-gated sodium channels (SCNA1A and others). The STITCH analysis also illustrates the wide variety of protein targets that are affected by both drugs. As known from single drug studies (Rogawski et al., 2016), both levetiracetam and topiramate are not highly selective for a single target but interact, in a complementary fashion, with various receptors and ion channels that are thought to be relevant for epileptogenesis (Pitkänen et al., 2015; Mazarati and Sankar, 2016; Klein et al., 2018), including the kainate and AMPA subtypes of glutamate receptors, the GABA_A receptor, different types of sodium and potassium channels, subtypes of carbonic anhydrase, Trak1, and SV2A. As shown by the STITCH analysis, treatment with LEV and TPM combines these various targets, thus resulting in a network pharmacology approach that is likely to explain the promising disease-modifying effect of the combination.

Figure 67–3.

Known and predicted drug-drug protein network interactions of the combination of levetiracetam and topiramate in humans (A) and mouse (Mus musculus; B), calculated by the STITCH database. Drug-protein and protein-protein networks are shown by the confidence (more...)

In a second approach to determining the potential mechanisms of the combination of levetiracetam and topiramate, high-throughput RNA-sequencing (RNA-seq) of the ipsilateral hippocampus of mice treated with the levetiracetam/topiramate combination was used (Schidlitzki et al., 2020). Genome-wide gene expression analysis of the ipsilateral hippocampus in treated and control mice revealed 19 known genes significantly differentially expressed. Twelve of these 19 genes have previously been implicated in epileptogenesis or epilepsy and could have potential disease-modifying effects in epilepsy including Nr4a1 (Nuclear receptor subfamily 4 group A member 1), a downstream target of CREB, thought to be a key regulator of epileptogenesis (Zhang et al., 2016), and Sstr4 (somatostatin receptor 4), a previously identified candidate for inhibition of epileptogenesis (Vezzani and Hoyer, 1999). These data thus suggest that drug-induced cellular differential expression plays a role in mediating the antiepileptogenic effects of this drug combination. Interestingly, in the contralateral hippocampus, only one gene (Gh; growth hormone) was significantly differentially expressed in drug-treated mice (Schidlitzki et al., 2020).

Unexpectedly, none of the drug combinations found effective in preventing or modifying the development of epilepsy reduced the hippocampal damage that occurs in the intrahippocampal kainate mouse model, although monotherapy with these drugs has been reported to exert neuroprotective effects in different brain injury models (Schidlitzki et al., 2020; Welzel et al., 2021). However, for any neuroprotective effect, the timing of the treatment after brain injury is important. As shown by Schidlitzki et al. (2020), the excitotoxic kainate exposure or the SE or both induced significant neurodegeneration in the hippocampus before drug treatment started at 6 h following kainate. The rapid onset of the kainate and/or SE detrimental effects may also explain that the drug combination did not reduce the impairment of the BBB and the neuroinflammation observed by multimodal brain imaging. Thus, to target these consequences of kainate, treatment should start as early as possible after kainate. This, however, could result in initial insult modification rather than an antiepileptogenic or disease-modifying effect. The finding that epilepsy could be prevented in some mice and SRS frequency could be reduced in most mice by transient (5 days) treatment shortly after SE, whereas neurodegeneration, BBB impairment, and neuroinflammation were not significantly reduced or prevented, indicates that adding drugs that interfere with these processes may further increase the efficacy of the combination treatment.

Indeed, as shown in Figure 67–2, the addition of gabapentin to the levetiracetam and topiramate combination markedly increased the antiepileptogenic efficacy, indicating synergistic interaction, which was also indicated by STITCH analysis (not illustrated). Although gabapentin is often proposed to act primarily via modulation of the α2δ subunit of voltage-gated Ca²⁺ channels and thus the synaptic transmitter release machinery, the role of calcium channels in the antiseizure and antiepileptogenic mechanism of this drug is uncertain (Sills and Rogawski, 2020). Evidence for a role of the α2δ subunit in gabapentin’s antiepileptogenic activity stems from the observation that brief gabapentin treatment after injury limits new excitatory synapse formation by preventing binding of thrombospondins to the α2δ-1 subunit, resulting in long-lasting effects that limit aberrant excitatory connectivity (Prince et al., 2016). In addition to effects mediated by the α2δ-1 subunit, gabapentin has been reported to increase GABA turnover in rodents and patients and to exert antiinflammatory effects that would likely contribute to the antiepileptogenic effect of the levetiracetam and topiramate combination (Sills and Rogawski, 2020). Early evidence that gabapentin might have antiepileptogenic effects was obtained in experiments on juvenile (P35) rats treated with gabapentin beginning 24 hours after kainic acid–induced SE (Cilio et al., 2001). In the “undercut” model of posttraumatic epilepsy (PTE), a brief 3-day treatment with gabapentin after injury prevented progressive increases in excitatory connectivity and epileptogenesis following neocortical trauma (Takahashi et al., 2018). In the lithium-pilocarpine model of TLE, early gabapentin treatment during the latent period reduced microglial activation and macrophage infiltration, which indicates antiinflammatory activity (Rossi et al., 2017). Thus, the marked effect of adding gabapentin to the levetiracetam/topiramate combination found in our experiments was not unexpected.

Effects on Diverse versus Similar Targets for Antiepileptogenesis

Our data may indicate that rationally selected drugs that target multiple proteins within an epileptogenic network are an effective way to go. However, some of the data summarized in Table 67–1 suggest that combining drugs with a similar mechanism of action may also be an effective means of preventing epilepsy, which appears to be a competing hypothesis. In the study of Shekh-Ahmad et al. (2019), two antioxidant drugs in combination were shown to prevent epilepsy in 67% of the mice following kainate-induced SE, while either compound alone was less effective. However, in a similar study by Pauletti et al. (2019), a combination of two antioxidant drugs (N-acetylcysteine and sulforaphane) in an electrically induced SE model of TLE, no prevention of epilepsy was observed. Similarly, a combination of two antiinflammatory drugs with different mechanisms of action did not prevent epilepsy in the rat pilocarpine model, although treatment started immediately before pilocarpine (Kwon et al., 2013). A combination of two drugs (ifenprodil and NBQX) that target different subtypes of the glutamate receptor was not effective to exert persistent antiepileptogenic efficacy in the intrahippocampal kainate mouse model (Schidlitzki et al., 2017). Thus, except for the study of Shekh-Ahmad et al. (2019), combining drugs with a similar mechanism of action was less effective than combination treatments that target multiple proteins within an epileptogenic network.

Top-Down versus Bottom-Up Target-Based Approaches in Identifying New Antiepileptogenic Therapies

There are at least two principal strategies for identifying an efficacious network approach for antiepileptogenesis. One is the top-down approach used in the studies summarized in Table 67–1, in which rationally chosen drug combinations that affect different targets within an epileptogenic network are tested in animal models. This top-down approach is relatively unspecific, thus resembling phenotypic drug screening (Swinney and Anthony, 2011; Eder and Herrling, 2016; Moffat et al., 2017). For this strategy, repurposing (or repositioning) of generic drugs can be used to facilitate translation of promising drug combinations into clinical trials, because such drugs have been clinically used for years and have well-known safety profiles (Ainsworth, 2011). If the novel combinations of generic drugs thus chosen involve new formulations, new routes of administrations, applications to new disorders, or doses that are considerably lower than those used in monotherapy, they can still be intellectually protected (Nosengo, 2016).

The second, bottom-up approach starts by identifying specific targets or critical nodes in the network and then searches for drugs (or develops new compounds) that selectively affect these targets or nodes (Swinney and Anthony, 2011; Eder and Herrling, 2016). This bottom-up target-based approach can very effectively develop novel treatments for a validated target or node, but the process of target validation is complex and associated with a high degree of uncertainty (Sams-Dodd, 2005; Swinney and Anthony, 2011). Indeed, in different fields of drug development, the target-based paradigm has not resulted in increased productivity over the traditional phenotypic approach; instead, the contribution of phenotypic screening to the discovery of first-in-class small-molecule drugs exceeded that of target-based approaches (Sams-Dodd, 2005; Swinney and Anthony, 2011; Eder et al., 2014; Moffat et al., 2017). Driven by advances in biology, engineering, and informatics, new paradigms integrate phenotypic with target-based algorithms into comprehensive, systems-level approaches offering value-added strategies for optimized drug discovery (Waldman and Terzic, 2013).

Future Advancements in the Search for Synergistic Antiepileptogenic Drug Combinations

Evaluating the antiepileptogenic potential of single drugs or drug combinations in in vivo models with SRS is extremely laborious and expensive, which renders dose-effect experiments or drug/drug combination screening difficult if not impossible. For instance, the systematic evaluation of drug combinations for antiepileptogenesis illustrated in Figure 67–1 took us more than 6 years, although a large group of scientists was involved. In theory, there are different approaches to resolve this dilemma. One is the use of high-throughput in vitro models of epileptogenesis, such as organotypic hippocampal or hippocampal-entorhinal cortex slice cultures, where epileptogenesis occurs on a compressed time scale (Dzhala and Staley, 2015; Berdichevsk et al., 2016; Drion et al., 2018). Interesting treatments identified in vitro are then evaluated in in vivo models of chronic epilepsy (Berdichevsk et al., 2016). An interesting example of a drug combination (phenobarbital and bumetanide) tested for antiepileptogenesis in an organotypic hippocampal slice culture is described in Table 67–1. The combination did not affect posttraumatic epileptogenesis (Dzhala and Staley, 2015). However, previous in vivo experiments with the phenobarbital-bumetanide combination in the pilocarpine rat model of TLE showed disease-modifying activity on epilepsy-associated behavioral abnormalities (Brandt et al., 2010), illustrating the limitations of in vitro testing.

Another interesting example of combining in vitro and in vivo methods in the search for novel disease-modifying therapies was recently reported by Lipponen et al. (2019). The authors developed a pipeline for the discovery of transcriptomics-derived disease-modifying therapies and used it to validate treatments in vitro and in vivo that could be repurposed for post-TBI treatment. In silico LINCS analysis identified desmethylclomipramine, ionomycin, rapamycin, and trimipramine as candidate treatments, modulating the TBI-induced transcriptomics networks. These drugs were tested in neuron-BV2 microglial co-cultures, monitoring drug effects on neuroinflammation, neurotoxicity, and neuronal survival. Based on the in vitro findings, BBB permeability, target engagement of Nrf2 target genes, and therapeutic time window in silico, desmethylclomipramine was chosen for in vivo validation in a lateral fluid-percussion injury (FPI) model of TBI in rats. Despite the favorable in silico and in vitro outcomes, in vivo assessment of clomipramine, which metabolizes to desmethylclomipramine, failed to demonstrate favorable effects on motor and memory tests, and, in fact, worsened the composite neuroscore. The development of SRS after TBI was not examined in this study. The authors concluded that their pipeline provides a rational stepwise procedure for evaluating favorable and unfavorable effects of systems biology discovered compounds that modulate post-TBI transcriptomics (Lipponen et al., 2019).

Another approach to screen a large number of drugs or drug combinations is computer-aided drug design, which is useful for the development of multitargeted drugs or combination therapies (Sun et al., 2016). For instance, Reutlinger and coworkers (2014) developed a quantitative polypharmacology model for 640 human drug targets. Casas et al. (2019) explored synergistic network pharmacology in ischemic stroke by an integrative in silico approach based on a primary target, NADPH oxidase type 4 (NOX4), to identify a mechanistically related cotarget, NO synthase (NOS). The in silico prediction was then validated and explored using both cellular in vitro and mouse in vivo stroke models, showing that both NOX4 and NOS inhibition is highly synergistic, leading to a significant reduction of infarct volume, direct neuroprotection, and BBB stabilization (Casas et al., 2019). The authors suggested that this systems medicine approach provides a ground plan to decrease current drug development failure in other complex indications. A similar systems approach was recently used by Anderson et al. (2020) for drug repurposing in chordoma, a devastating rare cancer with no approved medicine. Those authors used two strategies to identify synergistic effective combinations of drugs: validated Bayesian machine learning models of chordoma inhibition, followed by screening compounds of interest and their combinations in vitro.

Concerning the prevention of epilepsy, Kirchner et al. (2020) have described a systems biology approach of the human brain, using a “NeuroRepository” of electrically mapped epileptic tissues and associated data. This unbiased “Big Data” mining approach provides an opportunity to explore the underlying electrical, cellular, and molecular mechanisms of the human epileptic brain with the aim to identify novel druggable targets for antiepileptogenesis.

The public STITCH database that we used in our experiments (Figs. 67–1 and 67–3) integrates data on protein-protein, protein-chemical, and chemical-chemical interaction (Kuhn et al., 2008; Szklarczyk et al., 2016). STITCH was developed by the European Molecular Biology Laboratory (EMBL) and can be used in pharmacology, toxicology, or teratology research. In the current, fifth release of STITCH, which shares protein space with STRING v10, functionality was implemented to filter out the proteins and chemicals not associated with a given tissue (Szklarczyk et al., 2016). An example is shown in Figure 67–3A for the human brain, whereas tissue-specific data cannot be generated yet for the mouse (Fig. 67–3B). In silico evaluation of numerous drug combinations by STITCH correctly predicted two of the three effective drug combinations shown in Figure 67–2. However, the major disadvantage of STITCH for the search of novel drug combinations for antiepileptogenesis is that the database does not include disease-related alterations in protein networks that may be relevant for drug-protein interactions. Here, machine learning approaches may be beneficial, as they have been used in the search for biomarkers of PTE (Pitkänen et al., 2020).

Independent of the approach being used, interesting drug combinations thus identified preclinically need to be translated to the clinical arena. Here, the fact that almost all existing data on synergistic drug combinations stem from in vivo testing in rodent models of SE-induced TLE may limit successful translation, because SE is a rare cause of acquired epilepsy in humans (Klein et al., 2018). Furthermore, in the absence of any positive validation by clinically effective antiepileptogenic treatments, the translational value, if any, of post-SE models of epileptogenesis is not known (Löscher, 2020). Given the fact that most acquired epilepsies in humans arise from TBI, stroke, and brain infections (Klein et al., 2018), promising antiepileptogenic treatments identified in post-SE models must be validated in TBI, stroke, or encephalitis models before translation to the clinic. However, most available TBI, stroke, or encephalitis models have inherent problems, including the long latency to onset of SRS and low incidence of SRS, necessitating drug studies to be long and have large sample sizes (Löscher, 2020). Nevertheless, such models are increasingly being used in antiepileptogenic drug trials, and we have started to evaluate the most effective drug combinations identified in post-SE models of TLE (cf. Table 67–1) in a TBI/PTE model, using FPI in rats, which leads to epilepsy in ~40% of animals within 2–3 months (Pitkänen et al., 2017). Based on the commonalities in epileptogenesis across models and humans (Klein et al., 2018), we expect to establish a antiepileptogenic drug combination in a mouse CCI model of PTE for subsequent translation to first proof-of-concept clinical trial in patients with TBI. As shown in other areas of medicine, the use of repurposable medications will significantly decrease the duration, costs, and risks of such translation (Nosengo, 2016; Sun et al., 2016; Cha et al., 2018; Clout et al., 2019; Klein et al., 2020).

Conclusions and Outlook

Despite advances in understanding mechanisms of epileptogenesis, there is currently no approved treatment that prevents the development or progression of epilepsy in patients at risk. However, over the last ~10 years, several novel promising therapeutic approaches have been identified in animal models. Among those, the rational combination of drugs with different but synergistic mechanisms of action may provide a potentially useful antiepileptogenic strategy. An important benefit for the translation of such a network pharmacology approach to patients is the repurposing of drugs that are already clinically available. Repurposing (or repositioning) of approved drugs has recently gained new momentum for rapid identification and development of new therapeutics for diseases that lack effective drug treatment (Nosengo, 2016; Sun et al., 2016). A recent report lists 118 repurposed drug products for 203 new central nervous system indications prior to January 2016; 102 approved and 101 in development (Clout et al., 2019). Drug combinations of two or more compounds with different mechanisms of action increase successful drug repositioning (Sun et al., 2016). The use of drugs in combination can produce a synergistic effect if each of the drugs impacts a different target or signaling pathway that results in the reduction of required drug doses for each individual drug, as shown for the triple combination of levetiracetam, atorvastatin, and ceftriaxone in the intrahippocampal kainate mouse model of TLE (Welzel et al., 2021). Over the past decades, multitargeted and combinatorial therapies achieved considerable therapeutic efficacy by modulating the activities of the targets in complex diseases such as HIV-1 infection, cancer, asthma, and diabetes mellitus (Muhammad et al., 2018). For brain diseases, clinical trials on combinations of repurposed drugs, including losartan and atorvastatin or memantine and donepezil, are currently being performed in Alzheimer disease (Cha et al., 2018; Ihara and Saito, 2020) and Parkinson disease (Athauda and Foltynie, 2018). In epilepsy, drug repurposing has become an important strategy in the treatment of patients with therapies targeted to their specific pathophysiology (Demarest and Brooks-Kayal, 2018; Sisodiya, 2020). One important example is the use of vigabatrin for the prevention or modification of epilepsy in patients with tuberous sclerosis complex (Jozwiak et al., 2020). We envision that synergistic combinations of repurposed drugs as presented in this review will be demonstrated to prevent epilepsy in patients at risk within the next decade.

Acknowledgments

Disclosure Statement

References

1. Ainsworth, C.  2011, “Networking for new drugs”, Nat.Med., vol. 17, no. 10, pp. 1166–1168. [PubMed: 21988973]
2. Anderson, E., Havener, T. M., Zorn, K. M., Foil, D. H., Lane, T. R., Capuzzi, S. J., Morris, D., Hickey, A. J., Drewry, D. H., & Ekins, S.  2020, “Synergistic drug combinations and machine learning for drug repurposing in chordoma”, Sci.Rep., vol. 10, no. 1, p. 12982. [PMC free article: PMC7395084] [PubMed: 32737414]
3. Athauda, D. & Foltynie, T.  2018, “Drug Repurposing in Parkinson’s Disease”, Cns.Drugs, vol. 32, no. 8, pp. 747–761. [PubMed: 30066310]
4. Berdichevsky, Y., Saponjian, Y., Park, K. I., Roach, B., Pouliot, W., Lu, K., Swiercz, W., Dudek, F. E., & Staley, K. J.  2016, “Staged anticonvulsant screening for chronic epilepsy”, Ann.Clin.Transl.Neurol., vol. 3, no. 12, pp. 908–923. [PMC free article: PMC5224819] [PubMed: 28097203]
5. Brandt, C., Nozadze, M., Heuchert, N., Rattka, M., & Löscher, W.  2010, “Disease-modifying effects of phenobarbital and the NKCC1 inhibitor bumetanide in the pilocarpine model of temporal lobe epilepsy”, Journal of Neuroscience, vol. 30, pp. 8602–8612. [PMC free article: PMC6634618] [PubMed: 20573906]
6. Brandt, C., Töllner, K., Klee, R., Bröer, S., & Löscher, W.  2015, “Effective termination of status epilepticus by rational polypharmacy in the lithium-pilocarpine model in rats: window of opportunity to prevent epilepsy and prediction of epilepsy by biomarkers”, Neurobiol.Dis., vol. 75, pp. 78–90. [PubMed: 25549873]
7. Casas, A. I., Hassan, A. A., Larsen, S. J., Gomez-Rangel, V., Elbatreek, M., Kleikers, P. W. M., Guney, E., Egea, J., López, M. G., Baumbach, J., & Schmidt, H. H. H. W.  2019, “From single drug targets to synergistic network pharmacology in ischemic stroke”, Proc.Natl.Acad.Sci.U.S.A, vol. 116, no. 14, pp. 7129–7136. [PMC free article: PMC6452748] [PubMed: 30894481]
8. Cha, Y., Erez, T., Reynolds, I. J., Kumar, D., Ross, J., Koytiger, G., Kusko, R., Zeskind, B., Risso, S., Kagan, E., Papapetropoulos, S., Grossman, I., & Laifenfeld, D.  2018, “Drug repurposing from the perspective of pharmaceutical companies”, Br.J.Pharmacol., vol. 175, no. 2, pp. 168–180. [PMC free article: PMC5758385] [PubMed: 28369768]
9. Cilio, M. R., Bolanos, A. R., Liu, Z., Schmid, R., Yang, Y., Stafstrom, C. E., Mikati, M. A., & Holmes, G. L.  2001, “Anticonvulsant action and long-term effects of gabapentin in the immature brain”, Neuropharmacology, vol. 40, pp. 139–47. [PubMed: 11077080]
10. Clout, A. E., Della, P. O., Hanna, M. G., Orlu, M., & Pitceathly, R. D. S.  2019, “Drug repurposing in neurological diseases: an integrated approach to reduce trial and error”, J.Neurol.Neurosurg.Psychiatry, vol. 90, no. 11, pp. 1270–1275. [PubMed: 31171583]
11. de Souza, A. G., Chaves Filho, A. J. M., Souza Oliveira, J. V., de Souza, D. A. A., Lopes, I. S., de Carvalho, M. A. J., de Lima, K. A., Florenço Sousa, F. C., Mendes Vasconcelos, S. M., Macedo, D., & França Fonteles, M. M.  2019, “Prevention of pentylenetetrazole-induced kindling and behavioral comorbidities in mice by levetiracetam combined with the GLP-1 agonist liraglutide: Involvement of brain antioxidant and BDNF upregulating properties”, Biomed.Pharmacother., vol. 109, pp. 429–439. [PubMed: 30399578]
12. Demarest, S. T. & Brooks-Kayal, A.  2018, “From molecules to medicines: the dawn of targeted therapies for genetic epilepsies”, Nat.Rev.Neurol., vol. 14, no. 12, pp. 735–745. [PubMed: 30448857]
13. Drion, C. M., Kooijman, L., Aronica, E., van Vliet, E. A., Wadman, W. J., Chameau, P., & Gorter, J. A.  2019, “Curcumin reduces development of seizurelike events in the hippocampal-entorhinal cortex slice culture model for epileptogenesis”, Epilepsia, vol. 60, no. 4, pp. 605–614. [PubMed: 30747999]
14. Dzhala, V. & Staley, K. J.  2015, “Acute and chronic efficacy of bumetanide in an in vitro model of posttraumatic epileptogenesis”, CNS.Neurosci.Ther., vol. 21, no. 2, pp. 173–180. [PMC free article: PMC4391014] [PubMed: 25495911]
15. Eder, J. & Herrling, P. L.  2016, “Trends in Modern Drug Discovery”, Handb.Exp.Pharmacol., vol. 232, pp. 3–22. [PubMed: 26330257]
16. Francois, J., Koning, E., Ferrandon, A., & Nehlig, A.  2006, “The combination of topiramate and diazepam is partially neuroprotective in the hippocampus but not antiepileptogenic in the lithium-pilocarpine model of temporal lobe epilepsy”, Epilepsy Research, vol. 72, no. 2-3, pp. 147–163. [PubMed: 16945504]
17. Galanopoulou, A. S., Buckmaster, P. S., Staley, K. J., Moshe, S. L., Perucca, E., Engel, J., Jr., Löscher, W., Noebels, J. L., Pitkänen, A., Stables, J., White, H. S., O’Brien, T. J., & Simonato For The American Epilepsy Society Basic Science Committee And The International League Against Epilepsy Working Group On Recommendations For Preclinical Epilepsy Drug Discovery 2012, “Identification of new epilepsy treatments: Issues in preclinical methodology”, Epilepsia, vol. 53, no. 3, pp. 571–582. [PMC free article: PMC3551973] [PubMed: 22292566]
18. Hopkins, A. L.  2008, “Network pharmacology: the next paradigm in drug discovery”, Nat.Chem.Biol., vol. 4, no. 11, pp. 682–690. [PubMed: 18936753]
19. Ihara, M. & Saito, S.  2020, “Drug Repositioning for Alzheimer’s Disease: Finding Hidden Clues in Old Drugs”, J.Alzheimers.Dis., vol. 74, no. 4, pp. 1013–1028. [PubMed: 32144994]
20. Jozwiak, S., Kotulska-Jozwiak, K., Bebin, M., & Wong, M.  2020, “Modifying genetic epilepsies—results from studies on tuberous sclerosis complex and their potential mpact.”, Neuropharmacology, vol. 166, p. 107908. [PubMed: 31962286]
21. Kirchner, A., Dachet, F., & Loeb, J. A.  2020, “Identifying targets for preventing epilepsy using systems biology of the human brain”, Neuropharmacology, vol. 168, p. 107757. [PMC free article: PMC7959577] [PubMed: 31493467]
22. Klee, R., Töllner, K., Rankovic, V., Römermann, K., Schidlitzki, A., Bankstahl, M., & Löscher, W.  2015, “Network pharmacology for antiepileptogenesis: tolerability of multitargeted drug combinations in nonepileptic vs. post-status epilepticus mice.”, Epilepsy Research, vol. 118, pp. 34–48. [PubMed: 26600369]
23. Klein, P. & Tyrlikova, I.  2017, “Prevention of epilepsy: Should we be avoiding clinical trials?”, Epilepsy Behav., vol. 72, pp. 188–194. [PubMed: 28647441]
24. Klein, P., Dingledine, R., Aronica, E., Bernard, C., Blümcke, I., Boison, D., Brodie, M. J., Brooks-Kayal, A. R., Engel, J., Jr., Forcelli, P. A., Hirsch, L. J., Kaminski, R. M., Klitgaard, H., Kobow, K., Lowenstein, D. H., Pearl, P. L., Pitkänen, A., Puhakka, N., Rogawski, M. A., Schmidt, D., Sillanpää, M., Sloviter, R. S., Steinhauser, C., Vezzani, A., Walker, M. C., & Löscher, W.  2018, “Commonalities in epileptogenic processes from different acute brain insults: Do they translate?”, Epilepsia, vol. 59, pp. 37–66. [PMC free article: PMC5993212] [PubMed: 29247482]
25. Klein, P., Friedman, A., Hameed, M., Kaminski, R., Bar-Klein, G., Klitgaard, H., Koepp, M., Jozwiak, S., Prince, D., Rotenberg, A., Vezzani, A., Wong, M., & Löscher, W.  2020, “Repurposed molecules for antiepileptogenesis: missing an opportunity to prevent epilepsy?”, Epilepsia, vol. 61, pp. 359–386. [PMC free article: PMC8317585] [PubMed: 32196665]
26. Kuhn, M., von Mering, C., Campillos, M., Jensen, L. J., & Bork, P.  2008, “STITCH: interaction networks of chemicals and proteins”, Nucleic Acids Res., vol. 36, no. Database issue, p. D684–D688. [PMC free article: PMC2238848] [PubMed: 18084021]
27. Kwon, Y. S., Pineda, E., Auvin, S., Shin, D., Mazarati, A., & Sankar, R.  2013, “Neuroprotective and antiepileptogenic effects of combination of anti-inflammatory drugs in the immature brain”, J Neuroinflammation., vol. 10, p. 30. [PMC free article: PMC3599749] [PubMed: 23442201]
28. Lipponen, A., Natunen, T., Hujo, M., Ciszek, R., Hämäläinen, E., Tohka, J., Hiltunen, M., Paananen, J., Poulsen, D., Kansanen, E., Ekolle Ndode-Ekane, X., Levonen, A. L., & Pitkänen, A.  2019, “In Vitro and In Vivo Pipeline for Validation of Disease-Modifying Effects of Systems Biology-Derived Network Treatments for Traumatic Brain Injury-Lessons Learned”, Int.J.Mol.Sci., vol. 20, no. 21. [PMC free article: PMC6861918] [PubMed: 31671916]
29. Löscher, W. & Brandt, C.  2010, “Prevention or modification of epileptogenesis after brain insults: experimental approaches and translational research”, Pharmacol.Rev., vol. 62, pp. 668–700. [PMC free article: PMC3014230] [PubMed: 21079040]
30. Löscher, W., Klitgaard, H., Twyman, R. E., & Schmidt, D.  2013, “New avenues for antiepileptic drug discovery and development”, Nat.Rev.Drug Discov., vol. 12, pp. 757–776. [PubMed: 24052047]
31. Löscher, W.  2020, “The holy grail of epilepsy prevention: Preclinical approaches to antiepileptogenic treatments”, Neuropharmacology, vol. 167, p. 107605. [PubMed: 30980836]
32. Marrero-Rosado, B. M., de Araujo, F. M., Kundrick, E. R., Walker, K. A., Stone, M. F., Schultz, C. R., Nguyen, D. A., & Lumley, L. A.  2020, “Ketamine as adjunct to midazolam treatment following soman-induced status epilepticus reduces seizure severity, epileptogenesis, and brain pathology in plasma carboxylesterase knockout mice”, Epilepsy Behav., vol. 111, p. 107229. [PMC free article: PMC7541728] [PubMed: 32575012]
33. Mazarati, A. & Sankar, R.  2016, “Common Mechanisms Underlying Epileptogenesis and the Comorbidities of Epilepsy”, Cold Spring Harb.Perspect.Med, vol. 6, no. 7, a022798. [PMC free article: PMC4930916] [PubMed: 27371669]
34. Moffat, J. G., Vincent, F., Lee, J. A., Eder, J., & Prunotto, M.  2017, “Opportunities and challenges in phenotypic drug discovery: an industry perspective”, Nat.Rev.Drug Discov., vol. 16, no. 8, pp. 531–543. [PubMed: 28685762]
35. Muhammad, J., Khan, A., Ali, A., Fang, L., Yanjing, W., Xu, Q., & Wei, D. Q.  2018, “Network Pharmacology: Exploring the Resources and Methodologies”, Curr.Top.Med Chem., vol. 18, no. 12, pp. 949–964. [PubMed: 29600765]
36. Noé, F. M., Polascheck, N., Frigerio, F., Bankstahl, M., Ravizza, T., Marchini, S., Beltrame, L., Bandero, C. R., Löscher, W., & Vezzani, A.  2013, “Pharmacological blockade of IL-1beta/IL-1 receptor type 1 axis during epileptogenesis provides neuroprotection in two rat models of temporal lobe epilepsy”, Neurobiol.Dis., vol. 59, pp. 183–193. [PubMed: 23938763]
37. Nosengo, N.  2016, “Can you teach old drugs new tricks?”, Nature, vol. 534, no. 7607, pp. 314–316. [PubMed: 27306171]
38. Pauletti, A., Terrone, G., Shekh-Ahmad, T., Salamone, A., Ravizza, T., Rizzi, M., Pastore, A., Pascente, R., Liang, L. P., Villa, B. R., Balosso, S., Abramov, A. Y., van Vliet, E. A., Del Giudice, E., Aronica, E., Patel, M., Walker, M. C., & Vezzani, A.  2019, “Targeting oxidative stress improves disease outcomes in a rat model of acquired epilepsy”, Brain, vol. 142, no. 7, p. e39. [PMC free article: PMC6598637] [PubMed: 31145451]
39. Pitkänen, A. & Kubova, H.  2004, “Antiepileptic drugs in neuroprotection”, Expert Opin.Pharmacother., vol. 5, no. 4, pp. 777–798. [PubMed: 15102563]
40. Pitkänen, A. & Lukasiuk, K.  2009, “Molecular and cellular basis of epileptogenesis in symptomatic epilepsy”, Epilepsy Behav., vol. 14 Suppl 1, pp. 16–25. [PubMed: 18835369]
41. Pitkänen, A., Immonen, R. J., Grohn, O. H., & Kharatishvili, I.  2009, “From traumatic brain injury to posttraumatic epilepsy: what animal models tell us about the process and treatment options”, Epilepsia, vol. 50 Suppl 2, pp. 21–29. [PubMed: 19187291]
42. Pitkänen, A. & Wasterlain, C. C.  2009, “Commentary: translation of laboratory findings to clinic in the treatment of status epilepticus--a mission impossible?”, Epilepsia, vol. 50, no. 7, pp. 1834–1835. [PubMed: 20831523]
43. Pitkänen, A.  2010, “Therapeutic approaches to epileptogenesis—Hope on the horizon”, Epilepsia, vol. 51 (Suppl. 3), pp. 2–17. [PMC free article: PMC2923397] [PubMed: 20618393]
44. Pitkänen, A. & Engel, J., Jr. 2014, “Past and present definitions of epileptogenesis and its biomarkers”, Neurotherapeutics., vol. 11, no. 2, pp. 231–241. [PMC free article: PMC3996117] [PubMed: 24492975]
45. Pitkänen, A., Lukasiuk, K., Dudek, F. E., & Staley, K. J.  2015, “Epileptogenesis”, Cold Spring Harb.Perspect.Med, vol. 5, no. 10, p. a022822. [PMC free article: PMC4588129] [PubMed: 26385090]
46. Pitkänen, A., Kyyriäinen, J., Andrade, P., Pasanen, L., & Ndode-Ekane, X. E.  2017, “Epilepsy after traumatic brain injury,” in Models of seizures and epilepsy. Second Edition.,  A. Pitkänen et al., eds., Academic Press, London, U.K., pp. 661–682.
47. Pitkänen, A., Paananen, T., Kyyriäinen, J., Das, G. S., Heiskanen, M., Vuokila, N., Banuelos-Cabrera, I., Lapinlampi, N., Kajevu, N., Andrade, P., Ciszek, R., Lara-Valderrábano, L., Ekolle Ndode-Ekane, X., & Puhakka, N.  2020, “Biomarkers for posttraumatic epilepsy”, Epilepsy Behav. vol. 121, p. 107080. [PubMed: 32317161]
48. Prince, D. A., Gu, F., & Parada, I.  2016, “Antiepileptogenic repair of excitatory and inhibitory synaptic connectivity after neocortical trauma”, Prog.Brain Res., vol. 226, pp. 209–227. [PubMed: 27323945]
49. Reutlinger, M., Rodrigues, T., Schneider, P., & Schneider, G.  2014, “Multi-objective molecular de novo design by adaptive fragment prioritization”, Angew.Chem.Int.Ed Engl., vol. 53, no. 16, pp. 4244–4248. [PubMed: 24623390]
50. Rogawski, M. A., Löscher, W., & Rho, J. M.  2016, “Mechanisms of Action of Antiseizure Drugs and the Ketogenic Diet”, Cold Spring Harb.Perspect.Med., vol. 6, no. 5, p. a022780 [PMC free article: PMC4852797] [PubMed: 26801895]
51. Rossi, A., Murta, V., Auzmendi, J., & Ramos, A. J.  2017, “Early Gabapentin Treatment during the Latency Period Increases Convulsive Threshold, Reduces Microglial Activation and Macrophage Infiltration in the Lithium-Pilocarpine Model of Epilepsy”, Pharmaceuticals.(Basel), vol. 10, no. 4, p. 93. [PMC free article: PMC5748648] [PubMed: 29182533]
52. Sams-Dodd, F.  2005, “Target-based drug discovery: is something wrong?”, Drug Discov.Today, vol. 10, no. 2, pp. 139–147. [PubMed: 15718163]
53. Schidlitzki, A., Twele, F., Klee, R., Waltl, I., Römermann, K., Broer, S., Meller, S., Gerhauser, I., Rankovic, V., Li, D., Brandt, C., Bankstahl, M., Töllner, K., & Löscher, W.  2017, “A combination of NMDA and AMPA receptor antagonists retards granule cell dispersion and epileptogenesis in a model of acquired epilepsy”, Sci.Rep., vol. 7, no. 1, p. 12191. [PMC free article: PMC5610327] [PubMed: 28939854]
54. Schidlitzki, A., Bascunana, P., Srivastava, P. K., Welzel, L., Twele, F., Töllner, K., Käufer, C., Gericke, B., Feleke, R., Meier, M., Polyak, A., Ross, T. L., Gerhauser, I., Bankstahl, J. P., Johnson, M. R., Bankstahl, M., & Löscher, W.  2020, “Proof-of-concept that network pharmacology is effective to modify development of acquired temporal lobe epilepsy”, Neurobiol.Dis., vol. 134, p. 104664. [PubMed: 31678583]
55. Shekh-Ahmad, T., Lieb, A., Kovac, S., Gola, L., Christian, W. W., Abramov, A. Y., & Walker, M. C.  2019, “Combination antioxidant therapy prevents epileptogenesis and modifies chronic epilepsy”, Redox.Biol., vol. 26, p. 101278. [PMC free article: PMC6692059] [PubMed: 31382215]
56. Sills, G. J. & Rogawski, M. A.  2020, “Mechanisms of Action of Currently Used Antiseizure Drugs.”, Neuropharmacology, vol. 168, p. 107966. [PubMed: 32120063]
57. Sisodiya, S. M.  2020, “Precision medicine and therapies of the future”, Epilepsia. [PMC free article: PMC8432144] [PubMed: 32776321]
58. Sun, W., Sanderson, P. E., & Zheng, W.  2016, “Drug combination therapy increases successful drug repositioning”, Drug Discov.Today, vol. 21, no. 7, pp. 1189–1195. [PMC free article: PMC4907866] [PubMed: 27240777]
59. Swinney, D. C. & Anthony, J.  2011, “How were new medicines discovered?”, Nat.Rev.Drug Discov., vol. 10, no. 7, pp. 507–519. [PubMed: 21701501]
60. Szklarczyk, D., Santos, A., von Mering, C., Jensen, L. J., Bork, P., & Kuhn, M.  2016, “STITCH 5: augmenting protein-chemical interaction networks with tissue and affinity data”, Nucleic Acids Res., vol. 44, no. D1, p. D380–D384. [PMC free article: PMC4702904] [PubMed: 26590256]
61. Takahashi, D. K., Jin, S., & Prince, D. A.  2018, “Gabapentin Prevents Progressive Increases in Excitatory Connectivity and Epileptogenesis Following Neocortical Trauma”, Cereb.Cortex, vol. 28, no. 8, pp. 2725–2740. [PMC free article: PMC6041890] [PubMed: 28981586]
62. Temkin, N. R.  2009, “Preventing and treating posttraumatic seizures: the human experience”, Epilepsia, vol. 50 Suppl 2, pp. 10–13. [PubMed: 19187289]
63. Thurman, D. J., Begley, C. E., Carpio, A., Helmers, S., Hesdorffer, D. C., Mu, J., Touré, K., Parko, K. L., & Newton, C. R.  2018, “The primary prevention of epilepsy: A report of the Prevention Task Force of the International League Against Epilepsy”, Epilepsia, vol. 59, no. 5, pp. 905–914. [PMC free article: PMC7004820] [PubMed: 29637551]
64. Vezzani, A. & Hoyer, D.  1999, “Brain somatostatin: a candidate inhibitory role in seizures and epileptogenesis”, Eur.J.Neurosci., vol. 11, no. 11, pp. 3767–3776. [PubMed: 10583466]
65. Waldman, S. A. & Terzic, A.  2013, “Systems-based discovery advances drug development”, Clin.Pharmacol.Ther., vol. 93, no. 4, pp. 285–287. [PubMed: 23511779]
66. Welzel, L., Twele, F., Schidlitzki, A., Töllner, K., Klein, P., & Löscher, W.  2019, “Network pharmacology for antiepileptogenesis: tolerability and neuroprotective effects of novel multitargeted combination treatments in nonepileptic vs. post-status epilepticus in mice.”, Epilepsy Research, vol. 151, pp. 48–66. [PubMed: 30831337]
67. Welzel, L., Bergin, D., Schidlitzki, A., Twele, F., Johne, A., Klein, P., & Löscher, W.  2021, “Systematic evaluation of rationally chosen multitargeted drug combinations: a combination of low doses of levetiracetam, atorvastatin and ceftriaxone exerts antiepileptogenic effects in a mouse model of acquired epilepsy.”, Neurobiol.Dis., vol. 149, p. 105227. [PubMed: 33347976]
68. White, H. S. & Löscher, W.  2014, “Searching for the ideal antiepileptogenic agent in experimental models: single treatment versus combinatorial treatment strategies”, Neurotherapeutics., vol. 11, no. 2, pp. 373–384. [PMC free article: PMC3996126] [PubMed: 24425186]
69. WHO 2019, Epilepsy: a public health imperative. World Health Organization, Geneva.
70. Zhang, Y., Chen, G., Gao, B., Li, Y., Liang, S., Wang, X., Wang, X., & Zhu, B.  2016, “NR4A1 Knockdown Suppresses Seizure Activity by Regulating Surface Expression of NR2B”, Sci.Rep., vol. 6, p. 37713. [PMC free article: PMC5120300] [PubMed: 27876882]