Antiseizure Medications that Did Not Reach the Epilepsy Market: An Assessment of Factors Contributing to Their Failed Clinical Development Over the Last Three Decades
Institute of Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
David R. Bloom Center for Pharmacy, The Hebrew University of Jerusalem, Jerusalem, Israel
Department of Medicine (Austin Health), The University of Melbourne, Melbourne, VIC Australia
Department of Neuroscience, School of Translational Medicine, Monash University, Melbourne, VIC Australia
Abstract
The EILAT Conferences on New Antiepileptic Drugs (AEDs), which have taken place biennially since 1992, have traditionally offered a forum for stakeholders from industry and academia to present updates on potential antiseizure medications (ASMs) in development. We reviewed publicly accessible data on compounds that were presented at EILAT conferences over this period but failed to reach the ASM market. The overarching aim was to determine the most likely reason(s) for terminating development and to provide potentially useful clues to improve the efficiency of ASM development in the future. We restricted our analysis to investigational compounds in clinical development that targeted common epilepsies. Of 56 such compounds, 15 reached the ASM market, 11 are still in development, and 30 had their development for epilepsy indications terminated. Compounds whose development was terminated include atimepazole, beprodone, cannabidivarin, carabersat, conantokin-G, dezinamide, elpetrigine, flunarizine, fluorofelbamate, ICA-105665, isovaleramide, JNJ-40411813, losigamone, naluzotan, padsevonil, pitolisant, ralitoline, remacemide, safinamide, soretolide, talampanel, tonabersat, T2000, T2007, valnoctamide, valrocemide, VX-765, zandatrigine, zuranolone, and 534U87. For most of these compounds, termination of development occurred during phase 1 or phase 2. Unfavorable pharmacokinetic properties, such as short half-life or high drug–drug interaction potential, were a common likely cause of early termination. Overall, the most common reason for terminating clinical development was lack of efficacy, possibly related, at least in some cases, to use of suboptimal trial designs. Review of preclinical data suggested that, for many compounds, suboptimal clinical efficacy was unlikely to be primarily explained by their mechanism of action. In some instances, failure to pursue an epilepsy indication could be explained by prioritization of development for other neurological or psychiatric conditions. This may reflect the perception of the drug market for common epilepsies being relatively crowded, and only attractive for compounds with outstanding safety or efficacy advantages over existing medications.
Article notes
Untitled section
Received 2025 Nov 26; Accepted 2026 Jan 26; Issue date 2026.
Key Points
| We reviewed available data for investigational compounds presented since 1992 at the EILAT Conferences on New Antiepileptic Drugs (AEDs), to determine the most likely reason(s) for terminating their clinical development for an epilepsy indication. |
| For most compounds, termination of development occurred during phase 1 or phase 2; the most common reasons for terminating clinical development were unfavorable pharmacokinetic properties, drug interactions, and lack of efficacy, in some instances possibly related to use of suboptimal trial designs. |
| Experience acquired with investigational epilepsy medications that did not reach market can be valuable in guiding development of future treatments in this area. |
Introduction
Drug development is a costly and risky endeavor. The overall success rate from first-in-man trials (phase 1) to regulatory approval for central nervous system (CNS) drugs is lower than the overall success rate (8% versus 11%) and much lower than the success rate in other therapeutic areas such as antimicrobial therapy and cardiovascular medicine (about 20%) [1]. There is clearly a need to understand reasons(s) leading to termination of clinical development for specific drug classes, because understanding past failures could provide guidance to improve drug development in the future. To address this need, we considered it of interest to evaluate available data for compounds that over the last three decades underwent clinical development as potential antiseizure medications (ASMs) but did not make it into the epilepsy drug market.
We based our analysis on a review of compounds presented at the EILAT Conferences on New Antiepileptic Drugs (AEDs), of which we have been co-organizers since inception. The first of these conferences took place in Eilat, Israel, in 1992. Since then, 16 more conferences have been held biennially, the latest (EILAT XVII) in Madrid, Spain, in May 2024. The core part of the EILAT Conferences consists in presentations on new compounds in various stages of development for epilepsy. Progress reports summarizing these presentations have been published after each conference since 1995, first in Epilepsy Research and since 2018 in Epilepsia.
In the present article, we will focus on investigational compounds that entered clinical development and targeted common epilepsies. Of 56 such compounds presented at the EILAT Conferences, 15 reached the ASM market, 11 are still in development as potential epilepsy treatments, and 30 had their development for epilepsy terminated (Table 1). The latter 30 compounds are discussed in some detail below. A review of their putative mechanisms of action illustrates the diversity of targets pursued over the years in epilepsy drug discovery, including ion channels, glutamate receptor subtypes, targets associated with γ-aminobutyric acid (GABA) transmission, serotonin (5-HT) receptor subtypes, synaptic vesicle glycoprotein type 2 (SV2), melatonin receptors, histamine type 3 receptors, connexin 43 hemichannels, cannabinoid receptors, mediators of brain inflammation, and others. In addition to a brief overview of the pharmacological properties of these compounds, we will discuss results acquired during their clinical evaluation, and likely reasons for terminating their development. To provide an insight on the evolution in approaches used in epilepsy drug discovery and development, compounds are listed in chronological order of their presentation at EILAT Conferences.
| Development for epilepsy halted or terminated (30 compounds) | Reached the epilepsy drug market (15 compounds)a | Still in development or development outcome uncertain (11 compounds) |
|---|---|---|
| Flunarizine (1992, 1996) | Felbamate (1992) | DP-VPA (1998, 2000, 2002) |
| Remacemide (1992, 1994, 1996, 1998, 2000) | Gabapentin (1992) | Carisbamatec (2006, 2008, 2010) |
| Dezinamide (1994,1996) | Tiagabine (1992, 1994, 1996) | Huperzine A (2006, 2008, 2010, 2014, 2016, 2018) |
| Ralitoline (1996) | Topiramate (1992, 1994) | Seletracetamd (2006) |
| Valrocemide (1996, 1998, 2000, 2002, 2004, 2006, 2008, 2010, 2012) | Stiripentolb (1992, 1994, 2004, 2006) | OV329 (2018, 2022, 2024) |
| Losigamone (1998) | Levetiracetam (1994, 1996, 1998) | Azetukalner (2018, 2020, 2022, 2024) |
| Soretolide (1998) | Rufinamideb (1994, 1996, 1998, 2000, 2006) | Darigabat (2020, 2022) |
| 534U87 (1998) | Retigabine (1996, 1998, 2000, 2002, 2004, 2006, 2008, 2010) | AMT-260 (2024) |
| Isovaleramide (2000, 2004) | Ganaxoloneb (1998, 2006, 2008, 2010, 2012, 2014, 2016, 2018, 2020, 2022) | Opakalim (BHV-7000) (2024) |
| Safinamide (2000, 2002, 2004) | Pregabalin (1998, 2000, 2002, 2004) | RAP-219 (2024) |
| Talampanel (2000, 2002, 2004, 2006) | Lacosamide (2000, 2002, 2004, 2006, 2008, 2010) | Vormatrigine (2024) |
| Carabersat (2002) | Brivaracetam (2004, 2006, 2008, 2010, 2012, 2014, 2016) | |
| Conantokin-G (2002) | Eslicarbazepine acetate (2004, 2006, 2008, 2010) | |
| Atipamezole (2004) | Cenobamate (2008, 2010, 2012, 2014, 2020) | |
| Fluorofelbamate (2004, 2006) | Perampanel (2010, 2012, 2014) | |
| JZP-4 (2006, 2008, 2010) | ||
| T2000 (2008, 2010) | ||
| T2007 (2008, 2010) | ||
| Tonabersat (2008, 2012) | ||
| ICA-105665 (2010, 2012) | ||
| Valnoctamide (2010, 2012, 2014, 2016, 2018) | ||
| VX-765 (2012) | ||
| Cannabidivarin (2014, 2016, 2018) | ||
| Beprodone (2014) | ||
| Naluzotan (2014) | ||
| Pitolisant (2014) | ||
| Zuranolone (2014, 2016) | ||
| JJ-40411813 (2018, 2020) | ||
| Padsevonil (2018) | ||
| Zandatrigine (2018, 2020, 2022) |
Search Strategy and Assessment Criteria
Records of the scientific program of successive EILAT Conferences were reviewed, and compounds were selected for review on the basis of the following inclusion criteria: (i) individual chemical entities; (ii) availability of publicly accessible data on safety, pharmacokinetics, or efficacy in human subjects; (iii) evidence of intended use for the treatment or prevention of seizures associated with common focal or generalized epilepsies; and (iv) explicit disclosure by the sponsor that treatment for an epilepsy indication was placed on hold or terminated, or no publicly accessible evidence of further clinical development in epilepsy being considered for at least the last 5 years. We excluded (i) cell therapies and dietary therapies; (ii) compounds for which only preclinical data are publicly accessible; (iii) compounds targeting specifically rare epileptic disorders, such as developmental and epileptic encephalopathies (DEEs), progressive myoclonic epilepsies, and autoimmune epileptic disorders; and (iv) compounds intended for use specifically for the treatment of status epilepticus, or as rescue therapies for seizure emergencies. Compounds whose targeted indication included the treatment of common epilepsies but were subsequently developed for other indications were not excluded.
Information on selected compounds was retrieved by conducting a literature search on PubMed by using as sole search term(s) the name/code name(s) of the compound. When the search yielded more than 200 articles, its scope was restricted by using as additional terms “seizures AND/OR epilepsy.” If < 20 articles were retrieved in PubMed, an additional search was made on Google using again as search terms the name/code name(s) of the compound. Clinical trial history for each compound was also reviewed by a search on ClinicalTrials.gov using the name/code name(s) of each compound as search term. References cited in relevant retrieved articles, personal files, and websites of sponsoring companies were also reviewed when available. Our search covered the period up to 31 August 2025.
For some compounds, the primary reason for terminating clinical development was disclosed directly by the sponsor in press releases or relevant publications. This usually occurred after an adequately powered trial failed to meet the primary efficacy endpoint (as, for example, in the case of JNJ-40411813, padsevonil, and zandatrigine) or when unexpected clinical toxicity findings emerged (as for ICA-105665 and 534U87). For most compounds, the reason(s) for terminating clinical development had to be inferred from available data. We considered “unfavorable pharmacokinetics” as a factor likely to have caused or contributed to the decision to discontinue development when a compound had a short half-life (≤ 7 h), typically requiring three times daily (t.i.d.) or four times daily (q.i.d.) dosing in clinical studies. Flunarizine was also considered to have unfavorable pharmacokinetics owing to its large interindividual pharmacokinetic variability and a long half-life (2–7 weeks), which complicates trial design and dose adjustments. As discussed in Sect. 3.4, valrocemide’s unfavorable pharmacokinetics relate instead to its being partially metabolized in humans to valproic acid, thereby weakening the rationale for its development. Many compounds categorized as having “unfavorable pharmacokinetics” had additional shortcomings such as Michaelis–Menten pharmacokinetics (e.g., dezinamide) or the potential to be a victim or perpetrator of clinically significant drug interactions (e.g., remacemide, talampanel). “’Lack of efficacy” was inferred to be a factor causing or contributing to termination of development when randomized controlled trials (RCTs) failed to differentiate the compound from placebo in relevant seizure endpoints, or when no evidence suggestive of efficacy emerged from initial exploratory uncontrolled studies if development was terminated prior to conduction of RCTs. Within this category, we also included compounds that showed “modest efficacy,” arbitrarily defined as inferiority to an established active comparator or as a < 15% difference in responder rate (percentage of patients with ≥ 50% reduction in seizure frequency from baseline) between active treatment and placebo in the presence of appreciable adverse effects. Of note, for many compounds, more than one factor was inferred to have contributed to the decision to terminate development. Further details are provided in the subsections where data for individual compounds are discussed.
Compounds for which publicly accessible information include seizure outcome data after multiple dosing in at least ten patients with epilepsy are discussed in some detail in the sections below. Information for compounds for which seizure outcome data were not obtained or are not publicly accessible, or are limited to single-dose studies or studies in less than ten individuals, is discussed separately in Sect. 3.18.
Overview of Assessed Compounds
Flunarizine
Flunarizine (Fig. 1) is a selective voltage-dependent calcium channel blocker [2, 3], with a chemical structure related to the older antihistamines hydroxyzine and cinnarizine [4]. Flunarizine was initially marketed in Europe in the early 1980s and is still available in many countries (but not the USA and Japan), mainly as a prophylactic treatment for migraine [5].
Flunarizine shows broad spectrum activity in animal models, including maximal electroshock (MES)-induced seizures in mice, audiogenic seizures in DBA/2 mice, seizures induced by subcutaneous (s.c.) pentylenetetrazole (PTZ) in rats, and amygdala-kindled seizures in rats [4]. In humans, it has > 80% oral bioavailability and a very long half-life (2–7 weeks) related to its large volume of distribution (78 L/kg), which in turn is due to its extensive tissue penetration [6, 7]. In patients taking enzyme-inducing comedication, the half-life of flunarizine is shorter but still in the order of about 2 weeks [8].
Flunarizine has been extensively investigated as add-on therapy in patients with uncontrolled, mostly focal seizures. A 2013 Cochrane Review identified eight double-blind placebo-controlled trials, one of which was excluded from further assessment because its duration was < 8 weeks [9]. Of the remaining trials, seven used a two-period crossover design, with a flunarizine (or placebo) period of 12 or 16 weeks, and evaluated flunarizine doses of 15 mg/day or less [10–16]. Of note, only one trial used a wash-out period between treatments [10]. In these studies, flunarizine was generally well tolerated, but efficacy results were disappointing, with most trials failing to identify a significant difference in responder rates between the two treatments [17]. The poor efficacy results could be explained by methodological flaws, including a small sample size (20–30 participants, except for one study with 90 patients enrolled, but only 62 evaluable), use of a fixed and possibly suboptimal dose, and, most importantly, failure to recognize that the crossover design can be problematic when assessing a drug with the pharmacokinetic characteristics of flunarizine. Because of flunarizine’s long half-life, steady-state is unlikely to have been reached in the first half of the treatment period in all studies. Moreover, for patients who received flunarizine in the first treatment period, any effect is likely to have been carried over during the second (placebo) period. This is well illustrated by a report from one of these trials indicating that plasma levels of flunarizine in patients who had received the drug first were still detectable at the end of the 4-month placebo phase [12]. Both the delay in reaching steady-state and the presence of carry-over effects would be expected to attenuate flunarizine’s effect size (if any) in those studies.
The methodological shortcomings discussed above were recognized by the investigators of a subsequent trial sponsored by the Antiepileptic Drug Development Program of the National Institute of Neurological Disorders and Stroke (NINDS) of the US National Institutes of Health (NIH) [7]. First, they conducted a pilot study to determine the tolerability of different target plasma concentrations of flunarizine, and the feasibility of predicting steady-state plasma flunarizine concentrations from single-dose pharmacokinetic data [8]. On the basis of the results of this study, they designed a parallel-group, double-blind trial where patients with focal seizures were randomized to receive either placebo or an individualized flunarizine dose predicted to yield the desired target plasma concentration of the drug (60 ng/mL) [18]. Dose individualization was based on flunarizine’s single-dose pharmacokinetic parameters determined in each patient prior to randomization and allowed to minimize the impact of pharmacokinetic variability on drug response. To accelerate achievement of steady state-conditions, loading doses were administered during the first 6 days of treatment, the duration of which was 25 weeks. The mean individualized dose was 40 mg/day (range: 7–138 mg/day), which is almost three times the highest dose (15 mg/day) investigated in the nine placebo-controlled trials completed before then. A total of 93 patients were randomized, and 92 provided seizure data for the entire 25-week period. The median plasma flunarizine concentration was 71 ng/mL (range 45–186 ng/mL), close to the 60 ng/mL target. The mean reduction in seizure frequency from baseline was 24.4% in the flunarizine group compared with 5.7% in the placebo group, a difference that was statistically significant (P < 0.002) but regarded as “modest.” Responder rates were even lower—19.6% for the flunarizine group versus 12.8% for the placebo group. Adverse events were more common in the flunarizine group, including weight gain. Eight patients discontinued blinded treatment, all in the flunarizine group and all because of CNS adverse events, including depression in three cases. In an open-label extension trial limited to 16 individuals who completed the double-blind trial, further dose increments were found to increase the frequency of adverse events, but seizure control did not improve in most patients [18]. No subsequent controlled trials assessing flunarizine as an ASM appear to have been performed.
In conclusion, development of flunarizine in the treatment of epilepsy most likely did not progress owing to modest efficacy at doses that were associated with severe adverse effects in a small subset of patients. Unfavorable pharmacokinetic characteristics also played a role. Overall, the initial development of flunarizine for epilepsy provides instructive lessons, such as the repeated application of an inappropriate trial design, and failure to adequately investigate dose–response relationships. However, the approach used by the NIH–NINDS investigators to identify target plasma concentrations and address the challenges posed by flunarizine’s long half-life and variable clearance is scientifically admirable, even though justification for a 25-week placebo exposure could be questioned by modern standards.
Remacemide
Remacemide (Fig. 2) emerged from a drug discovery program aimed at designing a molecule with a three-dimensional structure similar to that of phenytoin [3]. It is a diphenyl-ethyl-acetamide derivative with one chiral center. It acts as a low-affinity N-methyl-d-aspartate (NMDA) receptor antagonist, and it also has sodium channels blocking properties. Remacemide shows broad spectrum antiseizure activity in animal models [19, 20]. Its desglycine primary metabolite is also active in these models and may contribute to clinical effects [21]. The enantiomers of remacemide and desglycine-remacemide differ in anticonvulsant potency, but these differences have not been considered biologically significant [19].
In humans, remacemide undergoes rapid absorption and is eliminated primarily by metabolism, with a half-life of 3–4 h [21]. Remacemide metabolism is enhanced by enzyme-inducing ASMs. Conversely, remacemide inhibits carbamazepine metabolism and increases significantly plasma carbamazepine concentrations. In patients receiving remacemide 600 mg/day, the increase in plasma carbamazepine concentration was estimated at 22% [22], but more substantial increases have been reported at higher remacemide doses [23, 24]). Phenytoin plasma concentrations can also be increased by remacemide [25].
Two pivotal double-blind, placebo-controlled, parallel-group add-on trials in patients with mostly focal epilepsy evaluated remacemide doses of 300, 600, and 800 mg/day on a twice daily (b.i.d.) dosing regimen [26] and 300, 600, and 1200 mg/day on a four times daily (q.i.d.) regimen [27]. To maintain the blinding, the dose of concomitantly administered carbamazepine and phenytoin was adjusted when necessary to keep the plasma concentrations of these ASMs unchanged compared with baseline, but despite such adjustment, the concentration of these drugs still increased slightly during remacemide treatment. There was no evidence of improvement in seizure frequency at remacemide doses below 800 mg. At 800 mg/day, the responder rate was 30% in the remacemide group, compared with 15% in the placebo group [26]. At 1200 mg/day, 23% of patients were responders compared with 7% on placebo [27]. Patients on remacemide were more likely to discontinue treatment prematurely, most likely owing to adverse effects, many of which were dose-dependent. Overall, the efficacy of remacemide was rated by Cochrane reviewers to be modest, and inferior to that of other ASMs available at the time [28].
Possibly owing to recognition of the modest efficacy results in the add-on setting and associated drug interactions, AstraZeneca went on to conduct a double-blind, parallel-group monotherapy trial in patients with newly diagnosed epilepsy using a novel double triangular sequential design with preplanned interim analyses. The trial was designed as a superiority trial, with time to first seizure after the 6-week titration as primary endpoint. Patients with two or more focal or generalized tonic–clonic seizures in the previous 12 months were randomized to receive 600 mg daily of remacemide or carbamazepine [29]. The trial was terminated after the second interim analysis and enrollment of 449 patients, when remacemide was found to be inferior to carbamazepine in preventing seizure recurrence (P = 0.003). No further clinical trials of remacemide in epilepsy appear to have been conducted since then. Thus, the unsuccessful development of remacemide appears to have been due to disappointing efficacy results, even at doses associated with appreciable adverse effects. Of note, the design of the monotherapy trial can be questioned, considering that the chosen target dose of remacemide (600 mg/day) had not shown evidence of efficacy in placebo-controlled add-on trials, and it was probably unrealistic to assume that it would have shown superior efficacy to 600 mg/day carbamazepine in a monotherapy setting. Apart from efficacy considerations, the decision not to continue remacemide development was probably reinforced by unfavorable pharmacokinetics requiring multiple daily dosing, and by clinically relevant interactions with concomitant ASMs, particularly carbamazepine.
Dezinamide
Dezinamide (AHR-11748; ADD94057, Fig. 3) is an aryloxy–azetidine–carboxamide derivative. It is also the desmethyl metabolite of fluzinamide, a compound investigated in the 1980s as a potential ASM [30, 31]. In preclinical models, dezinamide shows an activity profile similar to that of valproic acid (VPA) but with higher potency [31]. Dezinamide’s mode of action may be related to blockade of voltage-gated sodium channels, but this has not been confirmed [32].
In healthy volunteers, doses up to 400 mg q.i.d for 27 days were well tolerated [32]. In a subsequent adjunctive-therapy safety study in nine patients with epilepsy, the highest tolerated dose ranged from 300 to 2000 mg/day on a q.i.d dosing regimen [33]. In the latter study, the half-life of dezinamide after a single dose prior to starting multiple dosing was about 7 h on average, with a wide range (2–12 h). There was also considerable variability in tolerated plasma dezinamide levels, with patients on phenytoin tolerating the drug better than those on carbamazepine. Plasma dezinamide concentrations at steady-state were higher than predicted from single-dose pharmacokinetics, with a pattern consistent with dose-dependent Michaelis–Menten kinetics [33].
Because of the large pharmacokinetic variability, a NIH–NINDS sponsored n -of-1 placebo- and concentration-controlled randomized crossover add-on trial in patients with focal seizures uncontrolled by phenytoin was designed [34]. Each of the 15 patients was scheduled to be treated for six 5-week periods (three active paired with three placebo periods in random sequence) with placebo or dezinamide on a q.i.d. regimen. In the active treatment periods, the dose was increased by targeting a plasma dezinamide concentration of 12 µg/mL (dezinamide period 1), 24 µg/mL (dezinamide period 2), and 32 µg/mL (dezinamide period 3), though the concentrations achieved were well below target, particularly at the highest dose. Among the 14 patients completing the trial, seizure frequency was reduced by a median of 37.9% during dezinamide treatment (all doses combined) compared with placebo (P = 0.0025), and 40% of patients had > 50% seizure reduction. Dezinamide was well tolerated, but five patients required dose reductions owing to adverse effects.
Although the results of the trial were regarded as positive, no further clinical studies have been reported. Possible reasons for terminating development include a short half-life requiring multiple daily doses, and a highly variable and dose-dependent pharmacokinetics complicating individualization of dosage.
Valrocemide
Valrocemide (N-valproyl glycinamide, TV-1901, Fig. 4) was selected for development after pharmacokinetic/pharmacodynamic (PK/PD) studies conducted within a drug discovery program aimed at identifying valproyl derivatives of GABA and glycine derivatives with greater potency and reduced teratogenic potential compared with VPA. Valrocemide displays broad spectrum anticonvulsant activity in a wide variety of animal models [35]. After intraperitoneal (i.p.) administration in SWV mice at a dose of 600 mg/kg, it was devoid of teratogenic effects, whereas an identical dose of VPA was highly teratogenic in the same strain [36].
In humans, valrocemide has a half-life of 6.4–9.4 h and exhibits linear pharmacokinetics after single oral doses ranging between 250 and 4000 mg and multiple doses ranging between 250 and 1000 mg t.i.d [37]. Following oral dosing 10–20% of a valrocemide dose is excreted unchanged in urine and about 40% is recovered in urine as the inactive major metabolite valproyl glycine [36]. In rats and dogs, no conversion to VPA was detected. In humans, however, a small fraction of a valrocemide dose was metabolized to VPA [37], and in some individuals, plasma VPA levels were within the range observed in patients treated with low doses of VPA. In preliminary open-label adjunctive-therapy studies in 22 patients with epilepsy, valrocemide was well tolerated at maintenance dosages up to 2000 mg b.i.d., with some patients reporting reduction in seizure frequency [35, 37]. However, eventually, a decision was made to discontinue development, mainly owing to concerns that the presence of low concentrations of VPA in plasma undermined the assumption of valrocemide being also devoid of teratogenicity in humans (Bialer, unpublished).
Losigamone
Losigamone (AO-33, Fig. 5) emerged from the pharmacological screening of derivatives of β-methoxy-butenolides, which are found in various plants. Losigamone is a racemic mixture of two threo-enantiomers, with the (S)-enantiomer being more active than the (R)-enantiomer in protecting against seizures induced by electrical, physical and chemical stimuli in rodents [38, 39]. Losigamone acts by blocking sodium channels, potentiating GABAergic transmission and probably other mechanisms [40].
In healthy individuals, the half-life of (S)-losigamone (about 5 h) is twice as long as that of the (R)-enantiomer (about 2 h). Oral clearance is tenfold greater for the (R)-enantiomer than for the (S)-enantiomer [40].
Two double-blind, placebo-controlled adjunctive-therapy trials in patients with focal seizures have been reported [41, 42]. In the first trial, conducted in 203 patients, participants randomized to losigamone (500 mg t.i.d. for 8 weeks) had a median change in seizure frequency of 14.9% compared with 6.7% for those randomized to placebo (P = 0.004) [41]. Responder rates were 22.3% for the losigamone group versus 14.6% for the placebo group (P = 0.13) Almost twice as many patients in the losigamone group reported CNS-related adverse events (40%) than with placebo (24%). The second trial was a 12-week trial in a total of 264 patients and assessed two doses of losigamone (400 mg t.i.d. and 500 mg t.i.d.) [42]. Median reduction in seizure frequency was 3.3% for placebo compared with 19.7% for losigamone 1200 mg/day and 25.3% for losigamone 1500 mg/day (P < 0.01 for both losigamone groups versus placebo). Responder rates were 11.8% for placebo, 17.2% for losigamone 1200 mg/day, and 29.3% for losigamone 1500 mg/day (P = 0.004 for the highest-dose group versus placebo). The most common adverse event was dizziness, which was reported by 22% of patients in the 1500 mg/day group, 13% in the 1200 mg/day group, and 2.4% in the placebo group. In both trials, increases in γ-GT were more common in losigamone-treated patients than in those treated with placebo.
The reasons for discontinuing losigamone development are likely to be multiple and to include modest efficacy even at the highest dose tested (particularly in the first trial), suboptimal tolerability, a short half-life requiring t.i.d. dosing, and possibly, regulatory concerns in justifying development of a racemate rather than a single individual enantiomer [43].
Soretolide
In experimental models, soretolide (D-2916, Fig. 6) shows an antiseizure profile similar to that of carbamazepine [44]. Soretolide has a half-life of 5–6 h and is metabolized by CYP1A2 to an active metabolite with a slightly longer half-life that is found in serum at concentrations five to ten times higher than those of the parent drug. Soretolide inhibits CYP2C19 and increases the plasma concentration of concomitantly administered phenytoin [45].
Soretolide was well tolerated in an initial safety study in 18 patients with epilepsy at doses up to 2500 mg/day on a b.i.d. regimen. A double-blind, placebo-controlled adjunctive-therapy enrichment trial was initiated in 33 patients with focal epilepsy. Of the 33 patients who received soretolide in the initial open-label enrichment phase, only 2 were responders, 16 withdrew owing to lack of efficacy (n = 16) or intolerance (n = 1), and 10 were under evaluation [45]. Although final outcomes were not reported, these less than encouraging preliminary data might explain why no further clinical studies appear to have been conducted.
Safinamide
The discovery of safinamide (NW 1015, PNU-151774E, Fig. 7) resulted from a screening program of α-amino amide derivatives designed as follow-up compounds to milacemide [46]. Safinamide shows broad spectrum activity in animal models of seizures and epilepsy and is also active in models of Parkinson’s disease and neuroprotection [37, 47, 48]. Actions contributing to this broad activity profile include blockade of voltage-gated sodium and calcium channels, inhibition of monoamine oxidase type B (MAO-B), as well as other mechanisms [46, 48].
In humans, safinamide is rapidly and completely absorbed after oral administration and has linear pharmacokinetics at doses up to 300 mg/day [37, 49, 50]. It is extensively metabolized, with a half-life of about 22 h [37, 49]. Safinamide plasma concentrations appear to be lowered by concomitantly administered enzyme-inducing ASMs [37].
In an exploratory uncontrolled adjunctive-therapy 12-week safety study in 43 patients, mostly with focal seizures, safinamide was well tolerated. Among the 38 patients completing the study, 41% had a ≥ 50% seizure reduction compared with seizure frequency during a prospective baseline [37, 49]. Since then, safinamide development focused on the adjunctive treatment of Parkinson’s disease, an indication subsequently approved by the European Medicines Agency (EMA) and the US Food and Drug Administration (FDA), and no further studies in epilepsy have been reported [46]. The reasons for not pursuing further development in epilepsy are unclear but likely to be related to prioritization of other indications by the sponsor.
Talampanel
Talampanel (GYKI-53773, LY-300164, Fig. 8) is a selective noncompetitive antagonist of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) glutamate receptors [51]. Talampanel is a chiral compound that was developed as the pure (R)-enantiomer, since the (S)-enantiomer is inactive [35]. Talampanel shows broad spectrum antiseizure activity in rodent models and is also active in various models of neuroprotection [37, 51].
After a single oral dose in healthy subjects, talampanel is rapidly absorbed and eliminated with a half-life of about 4 h, which is reduced to about 3 h in patients taking enzyme-inducing ASMs. After repeated doses, the half-life is longer (about 5.6 h in patients comedicated with enzyme inducers), and the oral clearance decreases [37, 52]. The oral clearance of talampanel in patients on enzyme-inducing ASMs is about three times higher than that reported in healthy subjects, whereas patients on VPA tend to have a lower clearance than individuals not comedicated with ASMs [37].
A small double-blind, 28-week placebo-controlled cross-over trial enrolled 49 patients with focal seizures, using target dosages of between 25 and 75 mg t.i.d., depending on type of ASM comedication (enzyme inducers, VPA alone, or others) [53]. Lower doses could be used if the target dose was not tolerated, and the final mean dose for the enzyme-induced group (90% of patients) was 60 mg daily. Compared with the placebo period, median seizure frequency was reduced by 21% during talampanel treatment (P = 0.001). Adverse events were more common while on talampanel than on placebo, the difference being especially marked for dizziness (52% versus 16%) and ataxia (26% versus 2%). A randomized parallel-group, double-blind, placebo-controlled phase 2 trial in approximately 250 patients with focal seizures was reported to be “nearing completion” in 2007 [54] and was confirmed to have been completed with 190 patients enrolled in a 2011 update on ClinicalTrials.gov (NCT00034814), but results do not appear to have been published. At an interim analysis of data from 100 patients who entered an open-label extension study, median seizure reduction from baseline was 27% [51], a rather modest response for uncontrolled follow-up studies. No subsequent trials in epilepsy have been reported. The reasons for terminating talampanel development are open to speculation, but it is plausible that the results of the large unpublished phase 2, double-blind trial were not promising [55]. It has been suggested that the decision not to continue development in epilepsy was motivated by “an unfavorable ratio between efficacy and tolerability” due in part to its short half-life [56]. Additional unfavorable features of talampanel include the need for a t.i.d. dosing scheme, and susceptibility to drug–drug interactions requiring different dosing regimens depending on type of concomitant ASMs.
Carabersat
Carabersat (SB-204269, Fig. 9) is a fluorobenzoylamino benzopyran derivative that is active in rodent models of focal and generalized tonic–clonic seizures [57]. Its antiseizure effects appear to be mediated by interaction with a stereospecific binding site in the brain, the nature of which has not been fully clarified [58] but may be related to tonabersat’s binding site, i.e., connexin 43 hemichannels. Carabersat is extensively metabolized and is eliminated with a terminal half-life of about 24 h [35].
Results of clinical studies with carabersat have not been published in detail [35]. In phase 1 studies, carabersat was found to be cleared metabolically with a half-life of about 24 h. Its oral bioavailability is enhanced by food [35]. A randomized double-blind, placebo-controlled, adjunctive-therapy parallel-group trial that evaluated three doses (400, 800, or 1200 mg/day) of carabersat in 305 patients with focal seizures showed good tolerability and a “mean reduction in seizure frequency of 20–30% in the 1200 mg/day dose group compared with placebo (P < 0.05),” but no further detail was provided [35]. A novel formulation with improved pharmacokinetic profile was developed for subsequent studies, but no further trials were reported. Based on information in the public domain, the reasons for terminating carabersat development are unclear.
VX-765
VX-765 or belnacasan (Fig. 10) is a prodrug that undergoes esterase-mediated cleavage to VRT-043198, a selective and reversible inhibitor of the interleukin-1-converting enzyme caspase 1 [59, 60]. The rationale for its development stems from evidence that brain inflammation plays an important role in epileptogenesis and maintenance of seizure susceptibility [61]. VX-765 is effective in protecting against seizures in acute and chronic seizure models where inflammatory mechanisms are considered to have a pathogenic role [62]. VRT-043198 cannot be administered orally owing to its metabolic instability, but VX-765 is adequately absorbed and is effective in inhibiting caspase 1 activity in the brain [59, 62].
VX-765 (900 mg t.i.d.) was evaluated in a 6-week, phase 2a randomized, double-blind, parallel-group adjunctive-therapy trial in patients with focal seizures [60]. VX-765 was well tolerated. The mean percent reduction in seizure frequency was 15.6% in the VX-765 group (n = 48) compared with 7.9% in the placebo group (n = 12). Although the difference was not statistically significant, the trial was not powered for efficacy and a post hoc analysis revealed a trend for a delayed onset of antiseizure activity in the VX-765 group. A subsequent double-blind, placebo-controlled 13-week dose-ranging efficacy and safety study in patients with treatment-resistant focal seizures was terminated prematurely for “business-related” reasons [60]. Presumably, this decision was influenced by the rather modest efficacy signals emerged from the phase 2a study.
A possible reason for the less than impressive efficacy data from the phase 2a trial is the lack of robust clinical biomarkers to identify patients whose focal epilepsy is primarily caused by inflammatory mechanisms and is therefore more likely to benefit from caspase 1 inhibition. Preclinical studies have also suggested that effective targeting of brain inflammation in epilepsy may require combined administration of drugs acting at different levels in the inflammatory cascade [63].
Cannabidavarin
Cannabidavarin (Fig. 11) is a naturally occurring propyl analogue of cannabidiol that possesses anticonvulsant activity in a variety of preclinical seizure models [64–69]. The mechanisms responsible for these effects are unclear but they resemble at least in part those described for cannabidiol [70, 71].
Cannabidavarin pharmacokinetics has been investigated in healthy subjects following single (25–800 mg) and multiple (800 mg/day for 5 days) oral doses and was found to be dose-proportional [68]. Because of similarities in physical–chemical characteristics with cannabidiol, cannabidivarin’s oral bioavailability is likely to be low and highly variable [70]. In a study in children with epilepsy, peak plasma concentrations of the 7-hydroxy- and 7-carboxy-metabolites of cannabidivariin were much higher than those of the parent compound [71].
The efficacy and safety of cannabidivarin (400 mg b.i.d. for 2 weeks, followed by 800 mg b.i.d. for 6 weeks) have been evaluated in a double-blind, parallel-group adjunctive-therapy trial in a total of 162 patients with focal seizures [72]. Cannabidivarin was considered to have an acceptable safety profile in this study, but reductions in seizure frequency from baseline in the cannabidivarin group (40.5%) did not differ significantly from that reported for the placebo group (37.7%) [72]. Presumably owing to disappointing efficacy results, the development of cannabidivarin in epilepsy appears to have been halted. The compound, however, is still being evaluated for its potential as a treatment for autism spectrum disorder [73].
Beprodone
Beprodone (VLB-01, Fig. 12) is a melatonin receptor agonist that is active in a variety of models of focal and generalized seizures [67]. Phase 1 studies in healthy subjects showed that beprodone is rapidly absorbed and eliminated with a half-life of 8–9 h [67]. In a phase 2a double-blind, adjunctive-therapy safety trial, 60 patients with focal seizures were randomized to receive beprodone (500, 750, or 1000 mg/day) or placebo on a b.i.d. schedule for 9 weeks (5 weeks titration and 4 weeks maintenance). Beprodone was well tolerated at these doses. Patients allocated to the 750 and 1000 mg/day doses had a greater reduction in seizure frequency than those allocated to placebo [67]. The difference was not statistically significant, but the trial was not powered for efficacy.
A phase 3 randomized, 12-week placebo-controlled trial of beprodone (1500 mg/day) in patients with focal seizures aimed at enrolling 224 patients at nine centers in Russia was apparently completed in 2013, but final results were not reported [74]. An interim analysis limited to 112 patients from what appears to be the same trial reported a median percentage seizure reduction of 57.8% in the beprodone group compared with 20.7% in the placebo group (P = 0.002) [67]. The proportion of patients with ≥ 50% seizure reduction was also greater in patients allocated to beprodone than in those allocated to placebo (70.4% versus 38.5%, P = 0.001). Beprodone was well tolerated in this trial. No further studies appear to have been reported for this compound [75], and the reasons for terminating its clinical development remain open to speculation.
Naluzotan
Naluzotan (PRX-00023, Fig. 13) is a serotonin 5-HT1A receptor agonist. In phase 1 pharmacokinetic studies, naluzotan was found to be rapidly absorbed and eliminated metabolically with a half-life of about 12 h [76]. Naluzotan was evaluated initially as a potential treatment for anxiety and depression, but development in these indications was discontinued after the first controlled clinical trials failed to provide evidence of adequate efficacy [67, 77, 78]. On the basis of evidence that 5-HT1A receptor activation can result in antiseizure effects [67], a pilot placebo-controlled, two-period cross-over trial of naluzotan (120 mg b.i.d.) was conducted by NIH–NINDS in 12 patients with focal seizures. Duration of each treatment period was 12 weeks, with a 2–3-week wash-out between periods. No efficacy signals were identified in the nine assessable patients, and the study was terminated. No further clinical trials appear to have been conducted in epilepsy or other indications.
Pitolisant
Pitolisant (tripolisant, BF2.649, Fig. 14) is an antagonist/inverse agonist of the histamine 3 (H3) receptor. It is absorbed rapidly from the gastrointestinal tract and has a half-life of 10–12 h [79]. Pitolisant has been developed primarily for the treatment of excessive daytime sleepiness in individuals with narcolepsy, and it is marketed for this indication in Europe and the USA since 2016 and 2019, respectively [80].
Because pitolisant shows broad-spectrum activity in rodent models of focal, generalized tonic–clonic, and absence seizures [67], it was also considered for development as a potential ASM. In a proof-of-concept study in 14 patients with photosensitive epilepsy, pitolisant (single oral doses of 20, 40, and 60 mg) reduced or suppressed dose-dependently the photoparoxismal electroencephalographic (EEG) response [81]. These promising findings led to conduction of an open-label, noncomparative exploratory adjunctive-therapy trial in which pitolisant (titrated at a dose of 40 mg once daily) was administered for 3 months to patients with focal seizures. A predetermined interim analysis (after inclusion of 23 patients) showed no robust signals of efficacy, and the study was terminated. Presumably, the disappointing results of this study (conducted in 2005–2006 but published 10 years later) led the sponsor to prioritize development in narcolepsy [82]. In retrospect, it could be questioned whether an antiseizure effect would have been seen using a divided daily dosing scheme.
JNJ-40411813
JNJ-40411813 (ADX71149, Fig. 15) acts as a positive allosteric modulator at the presynaptic metabotropic glutamate receptor type 2 (mGlu2). In rodents, JNJ-40411813 is devoid of anticonvulsant activity in the MES, s.c. PTZ, and corneal-kindling model at doses up to 100 mg/kg s.c., but it does dose-dependently inhibit seizure activity in the 6 Hz mouse model [83]. A notable property of JNJ-40411813 is its ability to markedly potentiate the antiseizure activity of levetiracetam in experimental models, even when tested at doses where JNJ-40411813 is virtually inactive [84, 85]. This is consistent with broader evidence of synergistic antiseizure activity when mGlu2-positive allosteric modulators are combined with SV2A modulators [85].
In phase 1 studies in healthy subjects, JNJ-40411813 was well tolerated after single oral doses of 5–1000 mg and multiple doses of 50–225 mg b.i.d. JNJ-40411813 has an oral bioavailability of 20–50% and is metabolized by CYP3A4. Its half-life is about 24 h after multiple dosing [83], and its clearance is increased by concomitant treatment with enzyme-inducing ASMs.
The potential antiseizure efficacy of JNJ-40411813 was evaluated in a phase 2 randomized, double-blind, placebo-controlled, parallel-group adjunctive-therapy trial in a total of 110 patients with focal seizures receiving levetiracetam or brivaracetam comedication [86, 87]. Two cohorts were assessed. Depending on presence or absence of enzyme-inducing comedication, the JNJ-40411813 dose was 50 or 100 mg b.i.d., respectively, in cohort 1, and 100 or 200 mg b.i.d., respectively, in cohort 2. A time-to-event design was used, with time to baseline monthly seizure count as primary endpoint and a treatment duration up to 12 weeks. In neither cohort 1 nor cohort 2 JNJ-40411813 could be differentiated from placebo for the primary endpoint, and there were also no clinical benefits observed for any of the secondary efficacy endpoints. Although the treatment was well tolerated, the authors concluded that “the totality of the data within the dose range tested do not support any further studies of JNJ-40411813 in focal seizures” [87].
Padsevonil
Padsevonil (UCB-0942, Fig. 16) was designed to have a dual action, i.e., to act as an SV2 modulator (active on all SV2 isoforms, i.e., SV2A, SV2B, and SV2C) and as a GABAA receptor-positive allosteric modulator with a partial agonist profile at the benzodiazepine site [88–90].The rationale for its development was to combine in a single molecular entity properties associated with synergistic antiseizure activity, and the overall aim was to achieve superior efficacy and tolerability compared with existing ASMs [69, 88, 91]. In animal models, padsevonil exhibited broad spectrum antiseizure activity with greater potency than levetiracetam and brivaracetam. It was also effective in seizure models where levetiracetam and brivaracetam show little or no activity [69, 88].
In studies in healthy subjects, padsevonil was rapidly absorbed from the gastrointestinal tract and eliminated with a half-life of 6–7 h [92]. Plasma padsevonil exposure (area under the plasma drug concentration-time curve [AUC]) after multiple dosing was higher than that predicted from single-dose pharmacokinetics. The compound is eliminated primarily by CYP3A4-mediated metabolism, with CYP2C19 playing a secondary role. Padsevonil is a moderate inhibitor of CYP2C19 and a weak inducer of CYP3A4. Importantly, padsevonil metabolism is highly vulnerable to CYP3A4 induction and inhibition. In particular, plasma padsevonil exposure (AUC) is decreased by 83% by comedication with carbamazepine and increased approximately twofold by coadministration of erythromycin [92]. As a result, patients comedicated with carbamazepine or other strong CYP3A4 inducers as well as strong CYP3A4 inhibitors, were excluded from participation in phase 2 and 3 trials.
Dose selection for initial clinical trials of padsevonil aimed at achieving optimal target receptor occupancy on the basis of results of preclinical studies [93]. A phase 2a randomized, double-blind, proof-of-concept adjunctive-therapy trial was conducted in 55 inpatients with frequent and highly drug-resistant focal seizures, using ≥ 75% reduction in seizure frequency as primary endpoint [93]. The enrolled patients had a median baseline seizure frequency of ten seizures per week, and three-quarters had failed to respond to ≥ 8 ASMs. At the end of the short (3 weeks) inpatient evaluation period, 31% of patients on padsevonil (400 mg b.i.d.) and 11% of those on placebo met the primary endpoint (odds ratio 4.14; P = 0.067). Padsevonil also showed a favorable safety profile in this study. The promising efficacy signal from the phase 2a investigation led to the conduction of two further randomized parallel-group, double-blind, placebo-controlled adjunctive-therapy trials in patients with focal seizures. The first trial assessed padsevonil doses of 50, 100, 200, and 400 b.i.d. in a total of 410 patients, whereas the second trial evaluated doses of 100, 200, and 400 mg b.i.d. in a total of 210 patients [94]. Both trials included a 4-week titration and a 12-week maintenance period. In both trials, neither of the primary efficacy outcomes (change in in observable seizure frequency from baseline and proportion of patients with ≥ 75% reduction in observable seizure frequency from baseline) differed statistically from placebo for any of the padsevonil dose group. Although numerically more patients improved on padsevonil than on placebo, the data indicated that padsevonil was unlikely to produce greater therapeutic benefit than existing ASMs, and its clinical development was terminated [94].
The discrepancy between padsevonil’s excellent activity profile in animal models and the disappointing results of clinical trials is intriguing, particularly in view of the careful approach applied in trial design and dose selection. Padsevonil probably represents a late example of the increasing difficulties encountered in differentiating active ASMs from placebo using traditional trial designs [95]. These difficulties appear to originate from closely related factors such as increasing placebo responses over time, recruitment difficulties, geographical variation in response, and involvement of a large number of investigators with variable expertise in ASM trials [95–97]. The list of investigators involved in padsevonil efficacy trials was so long as to require 17 pages of a supplementary file, and involved centers in Europe, North America, Central America, Asia, and Oceania. Padsevonil’s pharmacokinetic profile had several intrinsic weaknesses, including a high interindividual variability in clearance and extreme sensitivity to enzyme induction, which prevented enrollment of patients taking strong CYP3A4 inducers such as carbamazepine and phenytoin [92]. Because of its modes of action, padsevonil’s efficacy could be expected to be reduced in patients taking other SV2 modulators such as levetiracetam and brivaracetam. Results of subgroups analysis of clinical trial data were consistent with this expectation, yet exclusion of patients on SV2 modulators would have severely hampered recruitment and negatively affected the compound’s market potential. Clearly, the sponsor regarded padsevonil’s commercial viability as being dependent on a large effect size in a difficult-to-treat patient population, a scenario that was not supported by the clinical trial data.
Zandatrigine
Zandatrigine (NBI-921352, XEN901, Fig. 17) is a heterocyclic sulfonamide that acts as a selective blocker of the voltage-gated sodium channels Nav1.6 encoded by the SCN8A gene [85, 98]. Zandatrigine is more potent than commonly used sodium-channel-blocking ASMs in focal seizure models [99]. In humans, it shows linear pharmacokinetics after single oral doses up to 80 mg or multiple doses up to 75 mg once daily or 45 mg b.i.d. for 7 days. Zandatrigine is metabolized by CYP3A4, CYP2D6, and CYP2C9, and has a half-life of 7.5–10.6 h.
In a first-in-human transcranial magnetic simulation (TMS) study, zandatrigine reduced corticospinal and cortical excitability at doses of 50–75 mg once daily [98]. A phase 2 proof-of-concept trial (NCT05159908) in 101 adults with focal seizures did not show meaningful reduction in seizure frequency, but details of this study were not reported [100]. On the basis of these results, the sponsor announced that development of zandatrigine for focal seizures was no longer pursued. A predominantly pediatric trial of zandatrigine in patients with SCN8A-DEE was terminated prematurely after only eight participants had been enrolled [101].
On the basis of available information, termination of development for the treatment of focal seizures appears to have been due to lack of efficacy. To our knowledge, the reasons for terminating development for SCN8A-DEE have not been disclosed.
Other Compounds
A summary of compounds for which development in epilepsy was terminated at an earlier stage is provided in Table 2. For most of these compounds, development was either terminated after phase 1 studies in healthy subjects or underwent early diversion to non-epilepsy indications. The Table also lists putative reasons for terminating (or not pursuing) further clinical development in epilepsy.
| Compound | Putative mechanism of action | Summary of available data | Probable reasons for halting/terminating clinical development | References |
|---|---|---|---|---|
| Atipamezole | α2 adrenoreceptor antagonist | Atipamezole is devoid of antiseizure activity in most models but may have disease-modifying effects after epileptogenic insults in animal models. Well tolerated in single-dose phase 1 studies after oral and parenteral administration. No studies reported in patients with epilepsy | Sponsor prioritized development as a drug to reverse the sedative and analgesic effects of α2 adrenoreceptor agonists in dogs | [37, 143, 144] |
| Conantokin-G (CGX-1007) | NMDA receptor antagonist | A cone snail peptide with broad spectrum antiseizure activity. Reported to be safe after i.v. dosing, but studies with intrathecal use (the intended therapeutic use) were not reported | Possibly, challenges involved in developing a compound that requires intrathecal administration | [35] |
| Elpetrigine (JZP-4) | Sodium and calcium channel blocker | Structurally related to lamotrigine, with greater potency than lamotrigine in preclinical studies. Elpetrigine has a half-life of 10 h, and its clearance is not significantly affected by valproic acid. Evidence of antiseizure activity was reported in a single-dose study in patients with photosensitive epilepsy | Reasons for terminating elpetrigine’s development are unclear | [37, 54, 145, 146] |
| Fluorofelbamate | Multiple (similar to felbamate) | Designed to retain the antiseizure activity of felbamate, with metabolic pathways not involving formation of a reactive metabolite assumed to cause felbamate’s idiosyncratic toxicity. Half-life is 16.7 h. No studies reported in patients with epilepsy | Possibly, challenges involved in collecting sufficient clinical data to ensure broad acceptance as a safer alternative to felbamate | [37, 54, 147] |
| ICA-105665 (PF-04895162) | Selective opener of neuronal Kv7.2/ 7.3/7.5 channels | Endowed with broad-spectrum activity in rodent seizure models. Half-life is 5–9.5 h. Single doses of 100–500 mg dose-dependently inhibited the photoparoxismal EEG response in patients with photosensitive epilepsy | Unexpected liver toxicity in a phase 1 study in healthy subjects, due to an off-target effect not identified by toxicology studies in rodents and primates | [60, 148–150] |
| Isovaleramide (NPS 1776) | Positive allosteric GABAA receptor modulator | Extensively metabolized in healthy subjects, with a very short half-life (2.4 h). No studies reported in patients with epilepsy | Unfavorable pharmacokinetics (short half-life) | [37, 151] |
| Ralitoline | Sodium channel blocker | Well tolerated at doses up to 300 mg t.i.d. Short half-life (4–6 h), extensively metabolized. Plasma ralitoline levels were reduced by 50% by enzyme-inducing ASMs. In ten patients with focal epilepsy, epileptiform EEG activity tended to decrease after a single 70 mg oral dose | Unfavorable pharmacokinetics (short half-life requiring multiple daily dosing). Possibly, also prioritization considerations because the sponsor was developing pregabalin at the same time | [32, 152, 153] |
| Tonabersat (SB-220453, XiflamTM), | Inhibition of connexin 43 hemichannels | Active in animal models of focal and generalized tonic–clonic seizures, migraines, and other diseases. Rapidly absorbed and eliminated with a half-life of 24–40 h. No studies reported in patients with epilepsy | Prioritization of other indications, including migraine. Currently being repurposed for other inflammasome-related diseases, in oncology, ophthalmology, and other therapeutic areas | [60, 154–160] |
| T2000 | Potentiation of GABAergic transmission | A prodrug of diphenylbarbituric acid, an analogue of eterobarbital, T2000 was assumed to be a less sedating barbiturate than phenobarbital. Development in essential tremor was terminated because of inadequate efficacy and poor tolerability. Clinical studies in epilepsy have not been reported for T2000 | Possibly, suboptimal outcome in essential tremor studies, as well as a high drug interaction potential. Development might have been deprioritized when the initial sponsor was acquired by another company in 2010 | [60, 154, 161, 162] |
| T2007 | Potentiation of GABAergic transmission | T2007 is the sodium salt of diphenylbarbituric acid, the active entity of T2000. Apart from phase 1 studies in healthy volunteers, clinical trials have not been reported for T2007 | Same as for T2000 | [148] |
| Valnoctamide | Unknown, possibly similar to valproic acid | A chiral compound structurally related to valproic acid but devoid of teratogenic effects in animal models. The half-lives of its stereoisomers range from 5 to 10 h (3–4 h in patients on enzyme-inducing ASMs). Initial promising results as an add-on treatment of mania were not confirmed in a subsequent larger monotherapy trial. No studies reported in patients with epilepsy | Weak intellectual property protection because the chemical structure was described in a 1934 Russian patent, and antiseizure activity was first reported in the 1980s. A short half-life in patients on enzyme-inducing ASMs | [32, 69, 163–168] |
| Zuranolone (SAGE217) | Positive allosteric GABAA receptor modulator | An orally active neurosteroid, eliminated by CY3A4-mediated metabolism with a half-life of 20–24 h. Although it is active in a wide range of seizure models, it has been developed as a potential treatment for mood disorders. No studies reported in patients with epilepsy | Prioritization of non-epilepsy indications. It is currently FDA approved for the treatment of postpartum depression | [68, 169, 170] |
| 534U87 | Unknown | Phase 1 studies completed successfully. Its half-life is 12–19 h, and its clearance is increased twofold by carbamazepine comedication. Of 89 patients with epilepsy who were treated, 17 developed edema of the face or other parts of the body | Unexpected adverse events | [45] |
Discussion
We reviewed available information on 30 different potential ASMs that reached clinical development since the early 1990s but did not progress to marketing approval for an epilepsy-related indication. The development of many of these compounds was stimulated by the US NIH–NINDS, which conducted anticonvulsant screening tests (and sometimes further investigations) in preclinical models, and for some compounds (e.g., flunarizine, dezinamide, and naluzotan), also sponsored pilot clinical trials. The establishment in 1975 of the NIH–NINDS Anticonvulsant Screening Program (ASP, subsequently renamed Epilepsy Therapy Screening Program or ETSP), played a fundamental role in overcoming stagnation in epilepsy drug development, and led to successful development of many valuable second-generation ASMs [102]. Drug development, however, is fraught with hurdles, and not all compounds for which preclinical data were considered sufficiently promising to justify initiation of clinical studies ultimately reached the market. Exploring the reasons for their lack of success can be instructive and allow improved decision-making when developing other ASM candidates.
For most of the compounds included in the present review, termination of development for an epilepsy indication occurred during phase 1 or phase 2, in agreement with results from surveys in other therapeutic areas [103]. For 13 compounds, development was terminated soon after phase 1 (Table 2). Unfavorable pharmacokinetics, particularly a short half-life, in some cases associated with a high drug–drug interaction potential, was a likely cause of early termination of development for many compounds (Table 3). Pharmacokinetic shortcomings seem to play a less important role in recent years [104], probably owing to greater awareness of the limitations of allometric scaling in predicting human pharmacokinetics [105] and availability of improved tools to predict drug interactions [106].
| Unfavorable pharmacokinetics | Lack of efficacy | Adverse effects | Prioritization of other indications | Other or unknown |
|---|---|---|---|---|
| Dezinamide | Cannabidivarin | ICA-105665 | Atipamezole | Beprodone |
| Flunarizine | Flunarizinea | 534U87 | Safinamide | Carabersat |
| Isovaleramide | JNJ-40411813 | Tonabersat | Conantokin-G | |
| Losigamone | Losigamonea | T2000 | Elpetrigine | |
| Ralitoline | Naluzotan | T2007 | Fluorofelbamate | |
| Remacemide | Padsevonil | Zuranolone | Talampanel | |
| Talampanel | Pitolisant | T2000 | ||
| Valrocemide | Remacemidea | T2007 | ||
| Soretolide | Valnoctamide | |||
| Valnoctamide (in acute mania) | Valrocemide | |||
| VX-765 | VX-765 | |||
| Zandatrigine |
Overall, the most common cause for development failure was lack of efficacy. While this may be surprising in view of the wide range of preclinical models that are available to predict antiseizure efficacy in patients with epilepsy [107], failure to demonstrate a sufficient degree of seizure protection in clinical trials can be due to many factors such as suboptimal pharmacokinetics, inadequate exploration of the potentially effective dose range, dose-limiting adverse effects, or a suboptimal trial design. Available information did not allow us to determine consistently whether an optimal dose had been tested, or whether dose-limiting toxicity (or the highest dose permitted by preclinical toxicology findings) was reached. Assessment was complicated by the fact that many studies, including relatively large trials of carabersat [35], talampanel [54], beprodone [74], and zandatrigine [100], do not appear to have been published in full. In any case, it was often difficult to determine whether testing higher doses could have been justified: For example, in the pivotal trial of losigamone, the authors commented that “it might be promising to study higher dosages in the future,” yet the proportion of patients discontinuing study medication prematurely (mostly owing to adverse events) was as high as 24% (24/92) in the 1500 mg/day dose group, compared with 15% (12/87) in the 1200 mg/day group, and 8% (7/85) in the placebo group [42]. The development of flunarizine and remacemide provides meaningful examples of suboptimal trial designs, as discussed earlier in this article.
In recent years, concerns have also been raised about the appropriateness, including statistical power limitations, of the traditional parallel-group adjunctive-therapy trial design to assess drugs for focal epilepsies [95]. Investigators’ increasing unease with exposing patients to serious risks associated with prolonged placebo treatment [108], and patients’ reluctance to take part in these trials, have resulted over the years in a progressive decrease in number of participants recruited per center, from > 10 in 1990 to ≤ 4 after 2020 [96]. At the same time, responder rates on placebo have increased steadily [96, 109, 110], from 10% in 1990 to over 22% after 2022 [97], resulting in the need to recruit larger number of patients to retain statistical power [96, 97, 109, 110]. Inclusion of larger numbers of study sites, with fewer patients enrolled at each site, has hampered statistical assessment of site-level sources of variability, which is problematic owing to evidence that placebo response differs across countries, regions, and patient characteristics [97]. The extent by which these confounders influenced trial outcomes for recently investigated compounds, such as padsevonil, is open to speculation. It should be noted, however, that the recent adjunctive-therapy trial that failed to demonstrate efficacy for JNJ-40411813 did not use the traditional design, but a time-to-event approach, which is especially designed to minimize exposure to placebo or ineffective treatments [87].
Some compounds that did not demonstrate adequate efficacy act with satisfactory potency on well-validated targets of established ASMs, such as ion channels and GABAergic transmission [107], suggesting that their suboptimal efficacy is unlikely to be primarily explained by their mechanism of action. In fact, efficacy, as assessed in a clinical trial, can be influenced by other variables such as target engagement at the assessed doses, other actions impacting on tolerability, pharmacokinetic factors, patient selection, and, as discussed above, clinical trial design. Because of these and other potential confounders, failure to demonstrate efficacy for a compound acting on a novel target does not necessarily negate the validity of that target. The efficiency of target-directed drug discovery and development is likely to improve in the future, owing to advances in drug design based on molecular modeling for receptor/target engagement [111, 112], advances in drug delivery techniques [113], availability of tools to assess target engagement in vivo [93], and, not least, the increasing focus on precision medicines targeting the molecular defects underlying specific epilepsy syndromes [114]. Some compounds discussed in this article also share mechanisms of action of drugs approved for non-epilepsy indications, which led the sponsor to prioritize development in other neurological or psychiatric conditions. This may reflect the perception of the drug market for common epilepsies being relatively crowded, and only attractive for compounds with outstanding safety or efficacy advantages over existing ASMs [115].
Several limitations of our work need to be acknowledged. We only considered compounds targeting common epilepsies, which represented the vast majority of treatments developed prior to 2015, because the market for rare genetic epilepsies was perceived at the time as being too small. This scenario changed completely in the last decade, largely owing to regulatory incentives to develop treatments for orphan diseases, coupled with an improved understanding of the pathophysiology of DEEs and other rare epilepsies [116]. Of the 6 ASMs approved by the FDA since 2018, five (everolimus, cannabidiol stiripentol, fenfluramine and ganaxolone) target orphan indications and only one (cenobamate) is approved for the treatment of focal epilepsies. Importantly, many more treatments targeting mostly DEEs are now in clinical development, including precisions treatments designed to correct the underlying genetic defect or its consequences [114, 116]. Factors contributing to success or failure in the development of these treatments may differ from those applicable to common epilepsies. We limited our assessment to compounds presented at EILAT Conferences, which may not be fully representative of the epilepsy drug development scenario ongoing at the time. However, the investigational treatments presented at the EILAT Conferences included almost all ASMs that subsequently received marketing approval over the assessed period. Although we based our assessment on publicly accessible information, the reason(s) for terminating clinical development was not always disclosed, and in some cases had to be inferred from available data, a process that required some subjective judgment. Other reasons, such as unpublished negative findings, financial constraints, and strategic considerations, could also have played a role. We are fully aware of the ancient proverb “When there is no flour for bread there is no Torah” and for some compounds, development may have stopped owing to lack of funds. As an indirect measure of financial resources available for development, we reviewed the development history of each compound. As presented in Table 4, about one-half of compounds with a terminated development (versus about two-thirds of compounds that reached the epilepsy market) were developed by medium- to large-sized companies. A few compounds that did not reach the market were developed by new venture companies, for which financial constraints are more likely to have contributed to a decision not to proceed with development. Of note, drug development decisions are influenced not only by the amount of resources available but also by prioritization of different investment options.
| Development for epilepsy halted or terminated | Reached the epilepsy drug market | ||||
|---|---|---|---|---|---|
| Compound | Company responsible for initial development | Notes | Compound | Company responsible for initial development | Notes |
| Atipamezole | Orion Pharma | First approved in Europe in 1966 for veterinary use to reverse the effects of certain sedative drugs | Brivaracetam | UCB Pharma | Approved in the USA and Europe in 2016 |
| Beprodone | Marco Polo Therapeutics | Cenobamate | SK Biopharmaceuticals | First approved in the USA in 2019 | |
| Cannabidivarin | GW Pharma | Eslicarbaze-pine acetate | Bial | Approved in Europe in 2009 and in the USA in 2013 | |
| Carabersat | SmithKline Beecham | SmithKline Beecham merged with Glaxo into GSK in 2000 | Felbamate | Carter Wallace | First approved in the USA in 1993 for the treatment of focal seizures and the treatment of seizures associated with Lennox–Gastaut syndrome |
| Conantokin-G | Cognetix | ||||
| Dezinamide | A.H Robins | A.H. Robins was acquired in 1989 by Wyeth, which in turn was acquired by Pfizer in 2009 | Gabapentin | Gödecke | Development was continued in the 1980s by Parke-Davis. First approved in the UK and the USA in 1993 |
| Flunarizine | Janssen | Approved in Europe in the 1980s for the treatment of migraine and vestibular vertigo | Ganaxolone | CoCensys | From 1999, development was pursued by Purdue Pharma and from 2004 by Marinus Pharmaceuticals. US approval was in 2022 |
| Fluorofelbamate | Medpointe | Lacosamide | Schwarz Pharma | Developed until the late 1990s by researchers at the University of Houston. Schwarz Pharma was acquired in 2006 by UCB Pharma. US approval was in 2009 | |
| ICA-105665 | Icagen | Pfizer collaborated with development since 2007 | |||
| Isovaleramde | NPS Pharmaceuticals | Levetiracetam | UCB Pharma | First approved in the USA in 2019 | |
| JJ-40411813 | Janssen | Janssen has been part of the Johnson & Johnson group since 1961. Since 2004, development was carried out in partnership with Addex Therapeutics | Perampanel | Eisai | Approved in the USA and Europe in 2012 |
| JZP-4 | Jazz Pharmaceuticals | Pregabalin | Parke-Davis | Parke-Davis (Warner Lambert group) was acquired by Pfizer in 2000. Pregabalin was approved in the USA and Europe in 2004 | |
| Losigamone | Dr. Willmar Schwabe | ||||
| Naluzotan | Eli Lilly | Development was continued by Predix Pharmaceuticals since 2004, by EPIX Pharmaceuticals since 2006 and by Proximagen since 2009 | Retigabine | Asta Medica | Development was continued by Wyeth since 1997 and since 2000 by Valeant and GlaxoSmithKline. Approved in Europe and the USA in 2011. Voluntarily withdrawn from market in 2017 |
| Padsevonil | UCB Pharma | ||||
| Pitolisant | Bioprojet | Initially developed by academic researchers. In 2017, Harmony Biosciences joined Bioprojet as co-developer. First approved in Europe in 2016 as a treatment for narcolepsy | Rufinamide | Ciba-Geigy | Ciba-Geigy merged with Sandoz into Novartis in 1996. Development rights were ceded to Eisai in 2004. Although initially developed in focal epilepsy, it was approved in Europe (2007) and the USA (2008) for the treatment of seizures associated with Lennox–Gastaut syndrome |
| Ralitoline | Gödecke | ||||
| Remacemide | Fisons | Since 1995, development was pursued by Astra, which merged into AstraZeneca in 1999 | Stiripentol | Biocodex | Though initially developed for focal epilepsy, it was approved in Europe (2007) and the USA (2018) for the treatment of seizures associated with Dravet syndrome |
| Safinamide | Farmitalia-Carlo Erba | Farmitalia-Carlo Erba was acquired by Pharmacia in 1993. Its license was acquired by Newron in 1998. Approved in Europe in 2015 for the treatment of Parkinson’s disease | |||
| Soretolide | Biocodex | Tiagabine | Novo-Nordisk | Co-developed with Abbott. Initially marketed in Denmark in 1996 | |
| T2000 | Taro Pharmaceuticals | Discovered by academic researchers in the 1980s | |||
| T2007 | Taro Pharmaceuticals | Development stopped in 2010 when Taro Pharmaceuticals was acquired by Sun Pharma | Topiramate | Johnson & Johnson | Initially approved in the UK in 1995 and in the USA in 1996 |
| Talampanel | IVAX | Development was pursued since 2001 by Eli Lilly and since 2006 by Teva, which terminated its development in 2010 | |||
| Tonabersat | SmithKline Beecham | Development was continued by Minster Pharmaceuticals (2001–2009) and by Proximagen and Upsher-Smith (since 2010). InflammX Therapeutics is pursuing development for ophthalmic indications | |||
| Valnoctamide | Neurocrine Biosciences | Initially marketed in France in 1964 as an anxiolytic. Developed for epilepsy and psychiatric indications by academic researchers in the 1990s and subsequently by Neurocrine Biosciences (2006–2009) | |||
| Valrocemide | Teva | Development was initiated by an academic group, pursued by Teva (1995–2005) and subsequently by Shire, which terminated development in 2008 | |||
| VX-765 | Vertex Phar maceuticals | ||||
| Zandatrigine | Xenon Pharmaceuticals | Development was continued by Neurocrine Biosciences in 2019 | |||
| Zuranolone | Sage Therapeutics | Since 2020, Biogen collaborated with Sage. First approved in 2020 in the USA for the treatment of postpartum depression | |||
| 534U87 | Glaxo Wellcome | Glaxo Wellcome and SmithKlineBeecham merged into GSK in 2000 | |||
For some compounds, a formal decision to terminate development was not made public, and evidence that development was no longer being pursued was based on lack of further published reports, press releases, or active protocols on ClinicalTrials.gov. We acknowledge that clinical development of these drugs could eventually resume, and there have been examples of compounds that re-entered clinical development after a hiatus of many years. Examples include DP-VPA, a valproic acid prodrug that has been in development since the 1990s [45, 117] and has recently re-entered clinical development in China [118], and carisbamate (RWJ-333369), a carbamate derivative that over a decade ago failed to demonstrate consistent efficacy in three phase 3 trials in focal seizures [119, 120] and is now being investigated by a different sponsor for the treatment of Lennox–Gastaut syndrome [121].
Conclusions
We reviewed clinical development pathways for 30 promising antiseizure compounds that failed to reach the epilepsy market. Not surprisingly, termination of development occurred mostly in phase 1 or early phase 2. However, there were several instances of development being stopped after completion of phase 2b or phase 3 trials involving a relatively large number of patients, as in the case of remacemide [26, 27], losigamone [41, 42], carabersat [35], beprodone [74], and padsevonil [94]. Terminating development at an advanced stage is a major setback, not least because of the cost implications involved, and reinforces the importance of critically evaluating underlying reasons and whether those could be predicted.
Our review suggests that lack of efficacy and unfavorable pharmacokinetics, including drug interactions, were the most common findings leading to development failure. Pharmacokinetic shortcomings such as poor bioavailability, short half-life, and drug interactions are generally identified early during development, and to some extent can be predicted on the basis of preclinical data [105, 106]. The compounds reviewed generally showed a promising antiseizure profile in preclinical models, including models incorporated in the ETSP. However, these models and their application have limitations, and suggestions have been made on how to improve their predictive value and reliability [122, 123]. Over time, the ETSP itself has greatly expanded their screening program, which now includes models of drug-resistant epilepsy, etiology-targeted models, syndrome-specific models, and models of disease modification [124–128]. As discussed in Sect. 4, the efficiency of drug discovery has also been improved by designing precision treatments that target the etiology of specific epilepsy syndromes, particularly monogenic epilepsies [114]. These advances, however, have not been matched by comparable progress in the efficiency of clinical drug designs. We still lack fully satisfactory proof-of-concept paradigms to assess antiseizure activity in humans after single-dose administration or short duration of treatment. Cortical excitability measures during transcranial magnetic stimulation in healthy individuals [129] or assessment of the photoparoxismal EEG response in patients with photosensitive epilepsy [130] have been advocated for this purpose. However, it is unclear whether these models are applicable across all drugs irrespective of mechanisms of action, and whether they can predict effectiveness against specific types of drug-resistant seizures or epilepsies. At present, assessment of antiseizure efficacy continues to be largely based on the traditional adjunctive-therapy RCT design, which, as discussed in section 4, is fraught with recruitment difficulties and appears to have become less efficient in differentiating active treatments from placebo [130]. Moreover, demonstration of the efficacy of ASMs remains based on use of seizure diaries compiled by patients or caregivers. Several studies using video-EEG monitoring have shown that a large proportion of seizures, including tonic–clonic seizures, are unreported, partly owing to patients being unaware of their occurrence [131–134]. Overreporting also occurs, further questioning the accuracy of seizure counts as assessed in traditional clinical trials [133–135]. Novel approaches based on objective counting of seizures with improved seizure detection devices [136–139] or use of intracranial EEG recordings in patients with chronically implanted electrodes [140, 141] could address some of these shortcomings. Efforts to improve ASM trial methodology should go in parallel with a critical re-assessment of clinical trial experience acquired to date, including an analysis of why many antiseizure compounds that entered clinical development failed to reach the market. We believe that our work went some way in addressing the latter question. As stated by Bill Gates, “it’s fine to celebrate success but it is more important to heed the lessons of failure” [142]. Drug development entails hurdles and risks. Some cannot be predicted or overcome, but others can be avoided or minimized by learning from the past.
Acknowledgements
The authors gratefully acknowledge Ms. Yara Sheeni for assistance with the drawing of chemical structures.
Funding
Open access funding provided by Hebrew University of Jerusalem.
Declarations
Funding
This work received no external funding.
Conflict of Interest
M.B. received consultancy fees from Angelini Pharma, Clexio Therapeutics, Guidepoint, Meditec (Sam On), NeuroSense Therapeutics, Pharma2B, Selene Therapeutics, Shackelford Pharma, and Xenon Pharmaceuticals. E.P. received speaker’s or consultancy fees from Eisai, GRIN Therapeutics, SK Life Science, Sun Pharma, Takeda, UCB Pharma, and Xenon Pharmaceuticals, and royalties from Wiley and Elsevier.
Data Availability
No datasets were generated or analyzed as part of the present work.
Ethics Approval
Not applicable.
Consent to Participate
Not applicable.
Consent to Publish
Not applicable.
Code Availability
Not applicable.
References
Untitled section
References
- 1.Kola I. The state of innovation in drug development. Clin Pharmacol Ther. 2008;83:227–30. 10.1038/sj.clpt.6100479.
- 2.Wang R, Karpinski E, Wu LY, Pang PK. Flunarizine selectively blocks transient calcium channel currents in N1E–115 cells. J Pharmacol Exp Ther. 1990;254(3):1006–11 (PMID: 1697620).
- 3.Rogawski MA, Porter RJ. Antiepileptic drugs: pharmacological mechanisms and clinical efficacy with consideration of promising developmental stage compounds. Pharmacol Rev. 1990;42(3):223–86 (PMID: 2217531).
- 4.Vanden Bussche G, Wauquier A, Ashton D, de Beukelaar F. Flunarizine. In: Meldrum BS, Porter RJ, editors. New anticonvulsant drugs. London: John Libbey; 1986. p. 125–33.
- 5.Amery WK. Flunarizine, a calcium channel blocker: a new prophylactic drug in migraine. Headache. 1983;23:70–4. 10.1111/j.1526-4610.1983.hed2302070.x.
- 6.Heykants J, De Cree J, Hoerig Ch. Steady-state plasma levels of flunarizine in chronically treated patients. Arzneimittel Forsch. 1979;29:1168–71 (PMID: 387049).
- 7.Pledger GW, Sackellares JC, Treiman DM, Pellock JM, Wright FS, Mikati M, et al. Flunarizine for treatment of partial seizures: results of a concentration-controlled trial. Neurology. 1994;44(10):1830–6. 10.1212/wnl.44.10.1830.
- 8.Treiman DM, Pledger GW, DeGiorgio C, Tsay JY, Cereghino JJ. Increasing plasma concentration tolerability study of flunarizine in comedicated epileptic patients. Epilepsia. 1993;34(5):944–53. 10.1111/j.1528-1157. (1993.tb02116.x.).
- 9.Cavazzuti GB, Galli V, Benatti A. The use of flunarizine in pediatric epilepsy. Funct Neurol. 1986;1(4):551–4 (PMID: 3609882).
- 10.Alving J, Kristensen O, Tsiropoulos I, Mondrup K. Double-blind placebo-controlled evaluation of flunarizine as adjunct therapy in epilepsy with complex partial seizures. Acta Neurol Scand. 1989;79(2):128–32. 10.1111/j.1600-0404.1989.tb03724.x.
- 11.Battaglia A, Ferrari AR, Guerrini R. Double-blind placebo-controlled trial of flunarizine as add-on therapy in refractory childhood epilepsy. Brain Dev. 1991;13(4):217–22. 10.1016/s0387-7604(12)80052-0.
- 12.Fröscher W, Bülau P, Burr W, Penin H, Rao ML, de Beukelaar F. Double-blind placebo-controlled trial with flunarizine in therapy-resistant epileptic patients. Clin Neuropharmacol. 1988;11(3):232–40. 10.1097/00002826-198806000-00005.
- 13.Keene D, Whiting S, Humphreys P, Jacob P. Flunarizine as a supplementary medication in refractory childhood epilepsy: a double-blind crossover study. Can J Neurol Sci. 1989;16(2):191–3. 10.1017/s0317167100028882.
- 14.Moglia A, Bergamasco B, Di Perri R, Mancia D. Flunarizine as add-on therapy in epilepsy. Crossover study vs placebo. Funct Neurol. 1986;1(4):547–50.
- 15.Overweg J, Binnie CD, Meijer JW, Meinardi H, Nuijten ST, Schmaltz S, et al. Double-blind placebo-controlled trial of flunarizine as add-on therapy in epilepsy. Epilepsia. 1984;25(2):217–22. 10.1111/j.1528-1157.1984.tb04178.x.
- 16.Starreveld E, de Beukelaar F, Wilson AF, McLean DR, Findlay HP. Double-blind crossover placebo-controlled study of flunarizine in patients with therapy-resistant epilepsy. Can J Neurol Sci. 1989;16(2):187–90. 10.1017/s0317167100028870.
- 17.Hasan M, Pulman J, Marson AG. Calcium antagonists as an add-on therapy for drug-resistant epilepsy. Cochrane Database Syst Rev. 2013;CD002750. 10.1002/14651858.CD002750.pub2.
- 18.Handforth A, Mai T, Treiman DM. Rising dose study of safety and tolerance of flunarizine. Eur J Clin Pharmacol. 1995;49(1–2):91–4. 10.1007/BF00192365.
- 19.Davies JA. Remacemide hydrochloride: a novel antiepileptic agent. Gen Pharmacol. 1997;28(4):499–502. 10.1016/s0306-3623(96)00280-7.
- 20.Nehlig A, Boehrer A. Effects of remacemide in two models of genetically determined generalized epilepsy, the GAERS and the audiogenic Wistar AS. Epilepsy Res. 2003;52(3):253–61. 10.1016/s0920-1211(02)00236-x.
- 21.Schachter SC, Tarsy D. Remacemide: current status and clinical applications. Expert Opin Investig Drugs. 2000;9(4):871–83. 10.1517/13543784.9.4.871.
- 22.Leach JP, Blacklaw J, Stewart M, Jones T, Richens A, Brodie MJ. Mutual interaction between remacemide hydrochloride and carbamazepine: two drugs with active metabolites. Epilepsia. 1996;37:1100–6. 10.1111/j.1528-1157.1996.tb01031.x.
- 23.Mawer GE, Jamieson V, Lucas SB, Wild JM. Adjustment of carbamazepine dose to offset the effects of the interaction with remacemide hydrochloride in a double-blind, multicentre, add-on drug trial (CR2237) in refractory epilepsy. Epilepsia. 1999;40:190–6. 10.1111/j.1528-1157.1999.tb02074.x.
- 24.Chadwick D, Smith D, Crawford P, Harrison B. Remacemide hydrochloride: a placebo-controlled, one-month, double-blind assessment of its safety, tolerability and pharmacokinetics as adjunctive therapy in patients with epilepsy. Seizure. 2000;9(8):544–50. 10.1053/seiz.2000.0448.
- 25.Leach JP, Girvan J, Jamieson V, Jones T, Richens A, Brodie MJ. Mutual interaction between remacemide hydrochloride and phenytoin. Epilepsy Res. 1997;26:381–8. 10.1016/s0920-1211(96)01005-4.
- 26.Jones MW, Blume WT, Guberman A, Lee MA, Pillay N, Weaver DF, et al. Remacemide hydrochloride as an add-on therapy in epilepsy: a randomized, placebo-controlled trial of three dose levels (300, 600 and 800 mg/day) in a B.I.D. regimen. Seizure. 2002;11(2):104–13. 10.1053/seiz.2002.0589.
- 27.Chadwick DW, Betts TA, Boddie HG, Crawford PM, Lindstrom P, Newman PK, et al. Remacemide hydrochloride as an add-on therapy in epilepsy: a randomized, placebo-controlled trial of three dose levels (300, 600 and 1200 mg/day) in a Q.I.D. regimen. Seizure. 2002;11(2):114–23. 10.1053/seiz.2002.0588.
- 28.Marson AG. Remacemide for drug-resistant localization related epilepsy. Cochrane Database Syst Rev. 2002;4:CD001900. 10.1002/14651858.CD001900.
- 29.Brodie MJ, Wroe SJ, Dean AD, Holdich TA, Whitehead J, Stevens JW. Efficacy and safety of remacemide versus carbamazepine in newly diagnosed epilepsy: comparison by sequential analysis. Epilepsy Behav. 2002;3:140–6. 10.1006/ebeh.2002.0337.
- 30.Donofrio PD, Sackellares C, Madsen JA, Matsuo F, Abou-Khalil B, Berent S, et al. Pilot study of fluzinamide (N-methyl-3-[3-(trifluoromethyl)phenoxy]-1-azetidinecarboxamide) in refractory partial seizures. Neurology. 1985;35(8):1239–41. 10.1212/wnl.35.8.1239.
- 31.Johnson DN, Osman MA, Cheng LK, Swinyard EA. Pharmacodynamics and pharmacokinetics of AHR-11748, a new antiepileptic agent, in rodents. Epilepsy Res. 1990;5(3):185–91. 10.1016/0920-1211(90)90037-v.
- 32.Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Loiseau P, Perucca E. Progress report on new antiepileptic drugs: a summary of the Third Eilat Conference. Epilepsy Res. 1996;25(3):299–319. 10.1016/s0920-1211(96)00081-2.
- 33.Pledger GW, Laxer KD, Sahlroot JT, Taylor MR, Cereghino JJ, McCormick C, et al. Pharmacokinetics and dose tolerability study ADD 94057 in comedicated patients with partial seizures. Epilepsia. 1992;33:112–8. 10.1111/j.1528-1157.1992.tb02292.x.
- 34.Privitera MD, Treiman DM, Pledger GM, Sahlroot JT, Handforth A, Linde MS, et al. Dezinamide for partial seizures: results of an n-of-1 design trial. Neurology. 1994;44:1453–8. 10.1212/wnl.44.8.1453.
- 35.Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Loiseau P, Perucca E. Progress report on new antiepileptic drugs: a summary of the sixth Eilat Conference (EILAT VI). Epilepsy Res. 2002;51(1–2):31–71. 10.1016/s0920-1211(02)00106-7.
- 36.Isoherranen N, Woodhead JH, White HS, Bialer M. Anticonvulsant profile of valrocemide (TV 1901): a new antiepileptic and CNS drug. Epilepsia. 2001;42:831–6. 10.1046/j.1528-1157.2001.042007831.x.
- 37.Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Perucca E, Tomson T. Progress report on new antiepileptic drugs: a summary of the Seventh Eilat Conference (EILAT VII). Epilepsy Res. 2004;61(1–3):1–48. 10.1016/j.eplepsyres.2004.07.010.
- 38.Jones FA, Davies JA. The anticonvulsant effects of the enantiomers of losigamone. Br J Pharmacol. 1999;128:1223–8. 10.1038/sj.bjp.0702919.
- 39.Stein U. Losigamone. In: Antiepileptic drugs. 3rd edition. In: Levy RH. Mattson RH, Meldrum BS, editors. New York: Raven Press; 1995. pp.1025–34.
- 40.Luszczki JJ. Third-generation antiepileptic drugs: mechanisms of action, pharmacokinetics and interactions. Pharmacol Rep. 2009;61:197–216. 10.1016/s1734-1140(09)70024-6.
- 41.Bauer J, Dienel A, Elger CE; Losigamone Study Group. Losigamone add-on therapy in partial epilepsy: a placebo-controlled study. Acta Neurol Scand. 2001;103(4):226–30 (PMID: 11328193).
- 42.Baulac M, Losigamone study group. Efficacy and safety of losigamone in partial seizures: a randomized double-blind study. Epilepsy Res. 2003;55:177–89. 10.1016/s0920-1211(03)00108-6.
- 43.Bialer M, Perucca E. From barbiturates to ganaxolone: the importance of chirality in drug development and in understanding the actions of old and new antiseizure medications. Pharmacol Ther. 2025;268:108808. 10.1016/j.pharmthera.2025.108808.
- 44.Fatope MO. Soretolide (Laboratoires Biocodex). Curr Opin Investig Drugs. 2001;2(6):824–7 (PMID: 11572664).
- 45.Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Loiseau P, Perucca E. Progress report on new antiepileptic drugs: a summary of the fourth Eilat conference (EILAT IV). Epilepsy Res. 1999;34(1):1–41. 10.1016/s0920-1211(98)00108-9.
- 46.Wasan H, Singh D, Kh R. Safinamide in neurological disorders and beyond: evidence from preclinical and clinical studies. Brain Res Bull. 2021;168:165–77. 10.1016/j.brainresbull.2020.12.018.
- 47.Fariello RG, McArthur RA, Bonsignori A, Cervini MA, Maj R, Marrari P, et al. 1. Preclinical evaluation of PNU 151774E as a novel anticonvulsant. J Pharmacol Exp Ther. 1998;285:397–403 (PMID: 9580576).
- 48.Fariello FG. Safinamide. Neurotherapeutics. 2007;4:110–6. 10.1016/j.nurt.2006.11.011.
- 49.Marzo A, Dal Bo L, Monti NC, Crivelli F, Ismaili S, Caccia C, et al. Pharmacokinetics and pharmacodynamics of safinamide, a neuroprotectant with antiparkinsonian and anticonvulsant activity. Pharmacol Res. 2004;50:77–85. 10.1016/j.phrs.2003.12.004.
- 50.Seithel-Keuth A, Johne A, Freisleben A, Kupas K, Lissy M, Krösser S. Absolute bioavailability and effect of food on the disposition of safinamide immediate release tablets in healthy adult subjects. Clin Pharmacol Drug Dev. 2013;2(1):79–89. 10.1002/cpdd.2.
- 51.Howes JF, Bell C. Talampanel. Neurotherapeutics. 2007;4(1):126–9. 10.1016/j.nurt.2006.11.001.
- 52.Langan YM, Lucas R, Jewell H, Toublanc N, Schaefer H, Sander YWAS, et al. Talampanel, a new antiepileptic drug: single- and multiple-dose pharmacokinetics and initial 1-week experience in patients with chronic intractable epilepsy. Epilepsia. 2003;44:46–53. 10.1046/j.1528-1157.2003.128902.x.
- 53.Chappell AS, Sander JW, Brodie MJ, Chadwick D, Lledo A, Zhang D, et al. A crossover, add-on trial of talampanel in patients with refractory partial seizures. Neurology. 2002;58(11):1680–2. 10.1212/wnl.58.11.1680.
- 54.Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Perucca E, Tomson T. Progress report on new antiepileptic drugs: a summary of the Eigth Eilat Conference (EILAT VIII). Epilepsy Res. 2007;73(1):1–52. 10.1016/j.eplepsyres.2006.10.008.
- 55.Rogawski MA. Revisiting AMPA receptors as an antiepileptic drug target. Epilepsy Curr. 2011;11(2):56–63. 10.5698/1535-7511-11.2.56.
- 56.Zaccara G, Giovannelli F, Cincotta M, Iudice A. AMPA receptor inhibitors for the treatment of epilepsy: the role of perampanel. Expert Rev Neurother. 2013;13(6):647–55. 10.1586/ern.13.46.
- 57.Upton N, Blackburn TP, Campbell CA, Cooper D, Evans ML, Herdon HJ, et al. Profile of SB-204269, a mechanistically novel anticonvulsant drug, in rat models of focal and generalized epileptic seizures. Br J Pharmacol. 1997;121(8):1679–86. 10.1038/sj.bjp.0701330.
- 58.Herdon HJ, Jerman JC, Stean TO, Middlemiss DN, Chan WN, Vong AK, et al. Characterization of the binding of [3H]-SB-204269, a radiolabelled form of the new anticonvulsant SB-204269, to a novel binding site in rat brain membranes. Br J Pharmacol. 1997;121(8):1687–91. 10.1038/sj.bjp.0701331.
- 59.Boxer BB, Quinn AMN, Shen M, Jadhav A, Leister W, Simeonov A, et al. A highly potent and selective caspase 1 inhibitor that utilizes a key 3-cyanopropanoic acid moiety. ChemMedChem. 2010;5:730–9. 10.1002/cmdc.200900531.
- 60.Bialer M, Johannessen SI, Levy RH, Perucca E, Tomson T, White HS. Progress report on new antiepileptic drugs: a summary of the Eleventh Eilat Conference (EILAT XI). Epilepsy Res. 2013;103(1):2–30. 10.1016/j.eplepsyres.2012.10.001.
- 61.Vezzani A, Balosso S, Ravizza T. The role of cytokines in the pathophysiology of epilepsy. Brain Behav Immun. 2008;22:797–803. 10.1016/j.bbi.2008.03.009.
- 62.Maroso M, Balosso S, Ravizza T, Iori V, Wright CI, French J, et al. Interleukin-1β biosynthesis inhibition reduces acute seizures and drug resistant chronic epileptic activity in mice. Neurotherapeutics. 2011;8(2):304–15. 10.1007/s13311-011-0039-z.
- 63.Noe FM, Polascheck N, Frigerio F, Bankstahl M, Ravizza T, Marchini S, et al. Pharmacological blockade of IL-1β/IL-1 receptor type 1 axis during epileptogenesis provides neuroprotection in two rat models of temporal lobe epilepsy. Neurobiol Dis. 2013;59:183–93. 10.1016/j.nbd.2013.07.015.
- 64.Hill AJ, Mercier MS, Hill TD, Glyn SE, Jones NA, Yamasaki Y, et al. Cannabidivarin is anticonvulsant in mouse and rat. Br J Pharmacol. 2012;167(8):1629–42. 10.1111/j.1476-5381.2012.02207.x.
- 65.Hill TD, Cascio MG, Romano B, Duncan M, Pertwee RG, Williams CM, et al. Cannabidivarin-rich cannabis extracts are anticonvulsant in mouse and rat via a CB1 receptor-independent mechanism. Br J Pharmacol. 2013;170(3):679–92. 10.1111/bph.12321.
- 66.dos Santos RG, Hallak JE, Leite JP, Zuardi AW, Crippa JA. Phytocannabinoids and epilepsy. J Clin Pharm Ther. 2015;40(2):135–43. 10.1111/jcpt.12235.
- 67.Bialer M, Johannessen SI, Levy RH, Perucca E, Tomson T, White HS. Progress report on new antiepileptic drugs: a summary of the Twelfth Eilat Conference (EILAT XII). Epilepsy Res. 2015;111:85–141. 10.1016/j.eplepsyres.2015.01.001.
- 68.Bialer M, Johannessen SI, Levy RH, Perucca E, Tomson T, White HS. Progress report on new antiepileptic drugs: a summary of the Thirteenth Eilat Conference on New Antiepileptic Drugs and Devices (EILAT XIII). Epilepsia. 2017;58(2):181–221. 10.1111/epi.13634.
- 69.Bialer M, Johannessen SI, Koepp MJ, Levy RH, Perucca E, Tomson T, et al. Progress report on new antiepileptic drugs: a summary of the Fourteenth Eilat Conference on New Antiepileptic Drugs and Devices (EILAT XIV). II. Drugs in more advanced clinical development. Epilepsia. 2018;59(10):1842–66. 10.1111/epi.14555.
- 70.Zamberletti E, Robion T, Parolaro D. Therapeutic potential of cannabidivarin for epilepsy and autism spectrum disorder. Pharmacol Ther. 2021;226:108787. 10.1016/j.pharmthera.2021.107878.
- 71.Hurley EN, Ellaway CJ, Johnson AM, Truong L, Gordon R, Galettis P, et al. Efficacy and safety of cannabidavarin treatment of epilepsy in girls with Rett syndrome: a phase 1 clinical trial. Epilepsia. 2022;63:1736–48. 10.1111/epi.17247.
- 72.Brodie M, Czapinski P, Pazdera L, Sander JW, Toledo M, Napoles M, et al. A phase 2 randomized controlled trial of the efficacy and safety of cannabidavarin as add-on therapy in participants with inadequately controlled focal seizures. Cannabis Cannabinoid Res. 2021;6:528–36. 10.1089/can.2020.0075.
- 73.Cammà G, Verdouw MP, van der Meer PB, Groenink L, Batalla A. Therapeutic potential of minor cannabinoids in psychiatric disorders: a systematic review. Eur Neuropsychopharmacol. 2025;91:9–24. 10.1016/j.euroneuro.2024.10.006.
- 74.Valexfarm. Remegal fixed dose as adjunctive therapy in patients with partial seizures (NCT01749046) (last updated 11 May 2017). https://clinicaltrials.gov/study/NCT01749046. Accessed 15 Nov 2025.
- 75.Anonymous. Drug profile: VLB-01. AdisInsight, Springer, Last updated 2 October 2021. https://adisinsight.springer.com/drugs/800032715. Accessed 15 Nov 2025.
- 76.Iyer GR, John F, Reinhard JF, Oshana S, Kauffman M, Donahue S. Tolerability, pharmacokinetics, and neuroendocrine effects of PRX-00023, a novel 5-HT1A agonist, in healthy subjects. J Clin Pharmacol. 2007;47:817–24. 10.1177/0091270007300953.
- 77.Rickels K, Mathew S, Banov MD, Zimbroff DL, Oshana S, Parsons EC, et al. Effects of PRX-00023, a novel, selective serotonin 1A receptor agonist on measures of anxiety and depression in generalized anxiety disorder: results of a double-blind, placebo-controlled trial. J Clin Psychopharmacol. 2008;28:2359. 10.1097/JCP.0b013e31816774de.
- 78.Kirchhoff VD, Nguyen HT, Soczynska JK, Woldeyohannes H, McIntyre RS. Discontinued psychiatric drugs in 2008. Expert Opin Investig Drugs. 2009;18:1431–43. 10.1517/13543780903184591.
- 79.Schwartz JC. The histamine H3 receptor: from discovery to clinical trials with pitolisant. Br J Pharmacol. 2011;163:713–21. 10.1111/j.1476-5381.2011.01286.x.
- 80.Lamb YN. Pitolisant: a review in narcolepsy with or without cataplexy. CNS Drugs. 2020;34:207–18. 10.1007/s40263-020-00703-x.
- 81.Kasteleijn-Nolst Trenité D, Parain D, Genton P, Masnou P, Schwartz JC, Hirsch E. Efficacy of the histamine 3receptor (H3R) antagonist pitolisant (formerly known as tiprolisant; BF2.649) in epilepsy: dose-dependent effects in the human photosensitivity model. Epilepsy Behav. 2013;28:66–70. 10.1016/j.yebeh.2013.03.018.
- 82.Collart Dutilleul P, Ryvlin P, Kahane P, Vercueil L, Semah F, Biraben A, et al. Exploratory phase II trial to evaluate the safety and the antiepileptic effect of pitolisant (BF2.649) in refractory partial seizures, given as adjunctive treatment during 3 months. Clin Neuropharmacol. 2016;39(4):188–93. 10.1097/WNF.0000000000000159.
- 83.Bialer M, Johannessen SI, Koepp MJ, Levy RH, Perucca E, Tomson T, et al. Progress report on new antiepileptic drugs: a summary of the Fourteenth Eilat Conference on New Antiepileptic Drugs and Devices (EILAT XIV). I. Drugs in preclinical and early clinical development. Epilepsia. 2018;59(10):1811–41. 10.1111/epi.14557.
- 84.Metcalf CS, Klein BD, Smith MD, Pruess T, Ceusters M, Lavreysen H, et al. Efficacy of mGlu2 -positive allosteric modulators alone and in combination with levetiracetam in the mouse 6 Hz model of psychomotor seizures. Epilepsia. 2017;58(3):484–93. 10.1111/epi.13659.
- 85.Bialer M, Johannessen SI, Koepp MJ, Levy RH, Perucca E, Perucca P, et al. Progress report on new antiepileptic drugs: a summary of the Fifteenth Eilat Conference on New Antiepileptic Drugs and Devices (EILAT XV). I. Drugs in preclinical and early clinical development. Epilepsia. 2020;61(11):2340–64. 10.1111/epi.16725.
- 86.Janssen Research & Development, LLC. A study to investigate JNJ-40411813 in combination with levetiracetam or brivaracetam in epilepsy. Last updated 3 July 2025. https://clinicaltrials.gov/study/NCT04836559. Accessed 25 Nov 2025.
- 87.French J, Bosak M, Kerr WT, Smekens P, Mistry D, Van Der Geyten S, et al. A randomized, double-blind, placebo-controlled, time to-event study of the efficacy and safety of JNJ-40411813 in combination with levetiracetam or brivaracetam in patients with focal onset seizures. Epilepsia. 2025. 10.1111/epi.18707. (Online ahead of print).
- 88.Leclercq K, Matagne A, Provins L, Klitgaard H, Kaminski RM. Pharmacological profile of the novel antiepileptic drug candidate padsevonil: characterization in rodent seizure and epilepsy models. J Pharmacol Exp Ther. 2020;372(1):11–20. 10.1124/jpet.119.261222.
- 89.Niespodziany I, Ghisdal P, Mullier B, Wood M, Provins L, Kaminski RM, et al. Functional characterization of the antiepileptic drug candidate, padsevonil, on GABA(A) receptors. Epilepsia. 2020;61(5):914–23. 10.1111/epi.16497.
- 90.Wood M, Daniels V, Provins L, Wolff C, Kaminski RM, Gillard M. Pharmacological profile of the novel antiepileptic drug candidate padsevonil: interactions with synaptic vesicle 2 proteins and the GABA(A) receptor. J Pharmacol Exp Ther. 2020;372(1):1–10. 10.1124/jpet.119.261149.
- 91.Wu PP, Cao BR, Tian FY, Gao ZB. Development of SV2A ligands for epilepsy treatment: a review of levetiracetam, brivaracetam, and padsevonil. Neurosci Bull. 2024;40:594–608. 10.1007/s12264-023-01138-2.
- 92.Chanteux H, MacPherson M, Kramer H, Otoul C, Okagaki T, Rospo C, et al. Overview of preclinical and clinical studies investigating pharmacokinetics and drug-drug interactions of padsevonil. Expert Opin Drug Metab Toxicol. 2024;20(8):841–55. 10.1080/17425255.2024.2373108.
- 93.Muglia P, Hannestad J, Brandt C, DeBruyn S, Germani M, Lacroix B, et al. Padsevonil randomized phase IIa trial in treatment-resistant focal epilepsy: a translational approach. Brain Commun. 2020;2(2):fcaa183. 10.1093/braincomms/fcaa183.
- 94.Rademacher M, Toledo M, Van Paesschen W, Liow KK, Milanov IG, Esch ML, et al. Efficacy and safety of adjunctive padsevonil in adults with drug-resistant focal epilepsy: results from two double-blind, randomized, placebo-controlled trials. Epilepsia Open. 2022;7(4):758–70. 10.1002/epi4.12656.
- 95.Perucca E. Antiepileptic drugs: evolution of our knowledge and changes in drug trials. Epileptic Disord. 2019;21(4):319–29. 10.1684/epd.2019.1083.
- 96.Kerr WT, Reddy AS, Seo SH, Kok N, Stacey WC, Stern JM, et al. Increasing challenges to trial recruitment and conduct over time. Epilepsia. 2023;64(10):2625–34. 10.1111/epi.17716.
- 97.Kerr WT, Suprun M, Kok N, Reddy AS, McFarlane KN, Kwan P, et al. Factors associated with placebo response rate in randomized controlled trials of antiseizure medications for focal epilepsy. Epilepsia. 2025;66(2):407–16. 10.1111/epi.18197.
- 98.Bialer M, Johannessen SI, Koepp MJ, Levy RH, Perucca E, Perucca P, et al. Progress report on new antiepileptic drugs: a summary of the Sixteenth Eilat Conference on New Antiepileptic Drugs and Devices (EILAT XVI): i. Drugs in preclinical and early clinical development. Epilepsia. 2022;63(11):2865–82. 10.1111/epi.17373.
- 99.Johnson JP, Focken T, Khakh K, Tari PK, Dube C, Goodchild SJ, et al. NBI-921352, a first-in-class, NaV1.6 selective, sodium channel inhibitor that prevents seizures in Scn8a gain-of-function mice, and wild-type mice and rats. Elife. 2022;11:e72468. 10.7554/eLife.72468.
- 100.Neurocrine Biosciences. Neurocrine Biosciences provides development pipeline update. Phase 2 proof-of-concept study of NBI-921352 in patients with focal onset seizures failed to demonstrate meaningful reduction in seizure frequency. Press, release 9 November 2023. https://neurocrine.gcs-web.com/node/20241/pdf. Accessed 15 Nov 2025.
- 101.Neurocrine Biosciences. Study to evaluate NBI-921352 as adjunctive therapy in subjects with SCN8A developmental and epileptic encephalopathy syndrome (SCN8A-DEE). Last updated 17 July 2025. https://www.clinicaltrials.gov/study/NCT04873869. Accessed 15 Nov 2025.
- 102.Rho JM, White HS. Brief history of anti-seizure drug development. Epilepsia Open. 2018;3(Suppl 2):114–9. 10.1002/epi4.12268.
- 103.Hay M, Thomas DW, Craighead JL, Economides C, Rosenthal J. Clinical development success rates for investigational drugs. Nat Biotechnol. 2014;32(1):40–51. 10.1038/nbt.2786.
- 104.Sun D, Gao W, Hu H, Zhou S. Why 90% of clinical drug development fails and how to improve it? Acta Pharm Sin B. 2022;12(7):3049–62. 10.1016/j.apsb.2022.02.002.
- 105.Mahmood I. Misconceptions and issues regarding allometric scaling during the drug development process. Expert Opin Drug Metab Toxicol. 2018;14(8):843–54. 10.1080/17425255.2018.1499725.
- 106.Lohy Das J, Ajavon-Hartmann AT, Schmith VGD. Navigating drug-drug interactions in clinical drug development: a tutorial. Clin Transl Sci. 2025;18(9):e70342. 10.1111/cts.70342.
- 107.Löscher W, Klein P. The pharmacology and clinical efficacy of antiseizure medications: from bromide salts to cenobamate and beyond. CNS Drugs. 2021;35(9):935–63. 10.1007/s40263-021-00827-8.
- 108.Ryvlin P, Cucherat M, Rheims S. Risk of sudden unexpected death in epilepsy in patients given adjunctive antiepileptic treatment for refractory seizures: a meta-analysis of placebo-controlled randomised trials. Lancet Neurol. 2011;10(11):961–8. 10.1016/S1474-4422(11)70193-4.
- 109.Guekht AB, Korczyn AD, Bondareva IB, Gusev EI. Placebo responses in randomized trials of antiepileptic drugs. Epilepsy Behav. 2010;17(1):64–9. 10.1016/j.yebeh.2009.10.007.
- 110.Rheims S, Perucca E, Cucherat M, Ryvlin P. Factors determining response to antiepileptic drugs in randomized controlled trials. A systematic review and meta-analysis. Epilepsia. 2011;52(2):219–33. 10.1111/j.1528-1167.2010.02915.x.
- 111.Garibsingh RA, Schlessinger A. Advances and challenges in rational drug design for SLCs. Trends Pharmacol Sci. 2019;40(10):790–800. 10.1016/j.tips.2019.08.006.
- 112.Rohilla S, Goyal G. Computer-aided drug design for Alzheimer’s disease: recent advancements and future perspectives. Curr Top Med Chem. 2025. 10.2174/0115680266343814250713100224. (Online ahead of print).
- 113.Vargas R, Martinez-Martinez N, Lizano-Barrantes C, Pacheco-Molina JA, García-Montoya E, Pérez-Lozano P, et al. Advancing through the blood-brain barrier: mechanisms, challenges and drug delivery strategies. ADMET DMPK. 2025;13(5):2988. 10.5599/admet.2988.
- 114.Wang S, Perucca E, Berkovic SF, Perucca P. Precision therapies for genetic epilepsies in 2025: Promises and pitfalls. Epilepsia Open. 2025 May 24. Online ahead of print. 10.1111/epi.13273.
- 115.Gohil K, Enhoffer D. Modest growth seen in the epilepsy market. Pharm Ther. 2014;39(11):786–7 (PMID: 25395821).
- 116.Perucca P, Perucca E. Novel pharmacological therapies in development. Curr Opin Neurol 2026 Jan 30. Online ahead of print. 10.1097/WCO.0000000000001456.
- 117.Labiner DM. DP-VPA D-Pharm. Curr Opin Investig Drugs. 2002;3:921–3 (PMID: 12137414).
- 118.Li Y, Zhan H, Wu J, Yu J, Cao G, Wu X, et al. Population pharmacokinetics and exposure-safety of lipophilic conjugates prodrug DP-VPA in healthy Chinese subjects for dose regime exploring. Eur J Pharm Biopharm. 2023;188:153–60. 10.1016/j.ejpb.2023.04.023.
- 119.Sperling MR, Greenspan A, Cramer JA, Kwan P, Kälviäinen R, Halford JJ, et al. Carisbamate as adjunctive treatment of partial onset seizures in adults in two randomized, placebo-controlled trials. Epilepsia. 2010;51(3):333–43. 10.1111/j.1528-1167.2009.02318.x.
- 120.Halford JJ, Ben-Menachem E, Kwan P, Ness S, Schmitt J, Eerdekens M, et al. A randomized, double-blind, placebo-controlled study of the efficacy, safety, and tolerability of adjunctive carisbamate treatment in patients with partial-onset seizures. Epilepsia. 2011;52(4):816–25. 10.1111/j.1528-1167.2010.02960.x.
- 121.Vashi V, Laramy JK, Vossler DG. Carisbamate treatment of adult and pediatric patients with Lennox-Gastaut-syndrome: a phase I pharmacokinetic dose-escalation study. Annual meeting of the American Epilepsy Society, Los Angeles, CA, 6–10 December 2024, abstract 1.403. https://aesnet.org/abstractslisting/carisbamate-treatment-of-adult-and-pediatric-patients-with-lennox-gastaut-syndrome-a-phase-i-pharmacokinetic-dose-escalation-study. Accessed 15 Nov 2025.
- 122.Galanopoulou AS, Kokaia M, Loeb JA, Nehlig A, Pitkänen A, Rogawski MA, et al. Epilepsy therapy development: technical and methodologic issues in studies with animal models. Epilepsia. 2013;54(4):13–23. 10.1111/epi.12295.
- 123.O’Brien TJ, Ben-Menachem E, Bertram EH 3rd, Collins SD, Kokaia M, Lerche H, et al. Proposal for a “phase II” multicenter trial model for preclinical new antiepilepsy therapy development. Epilepsia. 2013;54(4):70–4. 10.1111/epi.12300.
- 124.Wilcox KS, West PJ, Metcalf CS. The current approach of the Epilepsy Therapy Screening Program contract site for identifying improved therapies for the treatment of pharmacoresistant seizures in epilepsy. Neuropharmacology. 2020;166:107811. 10.1016/j.neuropharm.2019.107811.
- 125.Pernici CD, Mensah JA, Dahle EJ, Johnson KJ, Handy L, Buxton L, et al. Development of an antiseizure drug screening platform for Dravet syndrome at the NINDS contract site for the Epilepsy Therapy Screening Program. Epilepsia. 2021;62(7):1665–76. 10.1111/epi.16925.
- 126.Barker-Haliski M, Knox K, Zierath D, Koneval Z, Metcalf C, Wilcox KS, et al. Development of an antiepileptogenesis drug screening platform: effects of everolimus and phenobarbital. Epilepsia. 2021;62(7):1677–88. 10.1111/epi.16955.
- 127.Metcalf CS, Vanegas F, Underwood T, Johnson K, West PJ, Smith MD, et al. Screening of prototype antiseizure and anti-inflammatory compounds in the Theiler’s murine encephalomyelitis virus model of epilepsy. Epilepsia Open. 2022;7(1):46–58. 10.1002/epi4.12550.
- 128.Mensah JA, Thomson KE, Huff JL, Freeman T, Reilly CA, Rower JE, et al. Development of a preclinical testing platform for clinically relevant therapy for Dravet syndrome. Epilepsia. 2025;66(10):4023–34. 10.1111/epi.18516.
- 129.Premoli I, Rossini PG, Goldberg PY, Posadas K, Green L, Yogo N, et al. TMS as a pharmacodynamic indicator of cortical activity of a novel anti-epileptic drug, XEN1101. Ann Clin Transl Neurol. 2019;6(11):2164–74. 10.1002/acn3.50896.
- 130.Löscher W, Trenité DK. The human photosensitive epilepsy model for clinical proof-of-principle trials of novel antiseizure medications: 2. Analysis of drug trials and predictive value of the model. Epilepsia. 2025;66(8):2619–38. 10.1111/epi.18444.
- 131.Hoppe C, Poepel A, Elger CE. Epilepsy: accuracy of patient seizure counts. Arch Neurol. 2007;64(11):1595–9. 10.1001/archneur.64.11.1595.
- 132.Poochikian-Sarkissian S, Tai P, del Campo M, Andrade DM, Carlen PL, Valiante T, et al. Patient awareness of seizures as documented in the epilepsy monitoring unit. Can J Neurosci Nurs. 2009;31(4):22–3 (PMID: 20085117).
- 133.Nurse ES, Cook MJ, Dalic LJ. Seizure reporting in older patients with developmental and epileptic encephalopathies: a retrospective review of ambulatory video-EEG reports. Seizure. 2025;131:50–6. 10.1016/j.seizure.2025.06.003.
- 134.Hannon T, Fernandes KM, Wong V, Nurse ES, Cook MJ. Over- and underreporting of seizures: how big is the problem? Epilepsia. 2024;65(5):1406–14. 10.1111/epi.17930.
- 135.Wong V, Hannon T, Fernandes KM, Cook MJ, Nurse ES. Unseen yet overcounted: the paradox of seizure frequency reporting. Epilepsy Behav. 2025;165:110335. 10.1016/j.yebeh.2025.110335.
- 136.Shum J, Friedman D. Commercially available seizure detection devices: a systematic review. J Neurol Sci. 2021;428:117611. 10.1016/j.jns.2021.117611.
- 137.Jeppesen J, Christensen J, Ahrenfeldt Petersen O, Fenger S, Armand Larsen S, Wüstenhagen S, et al. Seizure detection using wearable electrocardiogram connected to a smartphone: a phase 3 clinical validation study. EBioMedicine. 2025;120:105952. 10.1016/j.ebiom.2025.105952.
- 138.Halliday AJ, Gillinder L, Lai A, Seneviratne U, Fontenot H, Cameron T, et al. The UMPIRE study: a first-in-human multicenter trial of bilateral subscalp monitoring for epileptic seizure detection. Epilepsia. 2025;66(9):3426–39. 10.1111/epi.18458.
- 139.Vakilna YS, Li X, Hampson JS, Huang Y, Mosher JC, Dabaghian Y, et al. Reliable detection of generalized convulsive seizures using an off-the-shelf digital watch: a multisite phase 2 study. Epilepsia. 2024;65(7):2054–68. 10.1111/epi.17974.
- 140.Goldenholz DM, Tharayil JJ, Kuzniecky R, Karoly P, Theodore WH, Cook MJ. Simulating clinical trials with and without intracranial EEG data. Epilepsia Open. 2017;2(2):156–61. 10.1002/epi4.12038.
- 141.Aboumatar S, Gavvala JR, Jaisani Z, Kuzniecky R, Privitera M, Ring M, et al. Electrographic seizures on responsive neurostimulation: an early and objective measure of response to cenobamate. Epilepsy Res. 2025;218:107647. 10.1016/j.eplepsyres.2025.107647.
- 142.The Economic Times of India. Quote of the day by Bill Gates. 2025. https://m.economictimes.com. Accessed 20 Nov 2025.
- 143.Pertovaara A, Haapalinna A, Sirviö J, Virtanen R. Pharmacological properties, central nervous system effects, and potential therapeutic applications of atipamezole, a selective alpha2-adrenoceptor antagonist. CNS Drug Rev. 2005;11(3):273–88. 10.1111/j.1527-3458.2005.tb00047.x.
- 144.Nissinen J, Andrade P, Natunen T, Hiltunen M, Malm T, Kanninen K, et al. Disease-modifying effect of atipamezole in a model of post-traumatic epilepsy. Epilepsy Res. 2017;136:18–34. 10.1016/j.eplepsyres.2017.07.005.
- 145.Foreman MM, Hanania T, Stratton SC, Wilcox KS, White HS, Stables JP, et al. In vivo pharmacological effects of JZP-4, a novel anticonvulsant, in models for anticonvulsant, antimania and antidepressant activity. Pharmacol Biochem Behav. 2008;89(4):523–34. 10.1016/j.pbb.2008.02.007.
- 146.Kasteleijn-Nolst Trenité D, Rosenfeld W, Wang S, De Colle C, Isojarvi J. Evidence of activity for JZP-4, a novel sodium and calcium channel inihibitor in patients with photosensitive epilepsy patients, an exploratory study. 62nd annual meeting of the American Epilepsy Society, Seattle, WA, 5–9 December 2008, abstract 3.221. Available at: https://aesnet.org/abstractslisting/evidence-of-activity-for-jzp-4--a-novel-sodium-and-calcium-channel-inhibitor-in-photosensitive-epilepsy-patients--an-exploratory-study. Accessed 15 Nov 2025.
- 147.Roecklein BA, Sacks HJ, Mortko H, Stables J. Fluorfelbamate. Neurotherapeutics. 2007;4:97–101. 10.1016/j.nurt.2006.11.015.
- 148.Bialer M, Johannessen SI, Levy RH, Perucca E, Tomson T, White HS. Progress report on new antiepileptic drugs: a summary of the Tenth Eilat Conference (EILAT X). Epilepsy Res. 2010;92(2–3):89–124. 10.1016/j.eplepsyres.2010.09.001.
- 149.Kasteleijn-Nolst Trenite DG, Biton V, French JA, Abou-Khalil B, Rosenfeld WE, Diventura B, et al. Kv7 potassium channel activation with ICA-105665 reduces photoparoxysmal EEG responses in patients with epilepsy. Epilepsia. 2013;54(8):1437–43. 10.1111/epi.12224.
- 150.Aleo MD, Aubrecht J, Bonin PD, Burt DA, Colangelo J, Luo L, et al. Phase I study of PF-04895162, a Kv7 channel opener, reveals unexpected hepatotoxicity in healthy subjects, but not rats or monkeys: clinical evidence of disrupted bile acid homeostasis. Pharmacol Rev Prospect. 2019;7:e00467. 10.1002/prp2.467.
- 151.Eadie MJ. Could valerian have been the first anticonvulsant? Epilepsia. 2004;45:1338–43. 10.1111/j.0013-9580.2004.27904.x.
- 152.Löscher W, von Hodenberg A, Nolting B, Fassbender CP, Taylor C. Ralitoline: a reevaluation of anticonvulsant profile and determination of “active” plasma concentrations in comparison with prototype antiepileptic drugs in mice. Epilepsia. 1991;32(4):560–8. 10.1111/j.1528-1157.1991.tb04693.x.
- 153.Anhut H, Satzinger G, von Hodenberg A. Ralitoline. In: Pisani F, Perucca E, Avanzini G, Richens A, editors. New antiepileptic drugs. Amsterdam: Elsevier; 1991. p. 141–215.
- 154.Bialer M, Johannessen SI, Levy RH, Perucca E, Tomson T, White HS. Progress report on new antiepileptic drugs: a summary of the Ninth Eilat Conference (EILAT IX). Epilepsy Res. 2009;83(1):1–43. 10.1016/j.eplepsyres.2008.09.005.
- 155.Acosta ML, Mat Nor MN, Guo CX, Mugisho OO, Coutinho FP, Rupenthal ID, et al. Connexin therapeutics: blocking connexin hemichannel pores is distinct from blocking pannexin channels or gap junctions. Neural Regen Res. 2021;16(3):482–8. 10.4103/1673-5374.290097.
- 156.Bhujbal S, Rupenthal ID, Agarwal P. Evaluation of ocular tolerability and bioavailability of tonabersat transfersomes ex vivo. Drug Deliv Transl Res. 2025. 10.1007/s13346-025-01872-2(Online ahead of print).
- 157.Cliff CL, Squires PE, Hills CE. Tonabersat suppresses priming/activation of the NOD-like receptor protein-3 (NLRP3) inflammasome and decreases renal tubular epithelial-to-macrophage crosstalk in a model of diabetic kidney disease. Cell Commun Signal. 2024;22(1):351. 10.1186/s12964-024-01728-1.
- 158.Zoteva V, De Meulenaere V, Vanhove C, Leybaert L, Raedt R, Pieters L, et al. Integrating and optimizing tonabersat in standard glioblastoma therapy: a preclinical study. PLoS ONE. 2024;19(3):e0300552. 10.1371/journal.pone.03005.
- 159.Kim Y, Griffin JM, Nor MNM, Zhang J, Freestone PS, Danesh-Meyer HV, et al. Tonabersat prevents inflammatory damage in the central nervous system by blocking connexin43 hemichannels. Neurotherapeutics. 2017;14(4):1148–65. 10.1007/s13311-017-0536-9.
- 160.Silberstein SD, Schoenen J, Gobel H, Diener HC, Elkind AH, Klapper JA, et al. Tonabersat, a gap-junction modulator: efficacy and safety in two randomized, placebo-controlled, dose-ranging studies of acute migraine. Cephalalgia. 2009;29(Suppl. 2):17–27. 10.1111/j.1468-2982.2009.01974.x.
- 161.Melmed C, Moros D, Rutman H. Treatment of essential tremor with the barbiturate T2000 (1,3-dimethoxymethyl-5,5-diphenyl-barbituric acid). Mov Disord. 2007;22:723–7. 10.1002/mds.21321.
- 162.Hadji Tahar A, Chuang R, Steeves T, Jog M, Lang A, Grimes D. Efficacy and safety of T2000 in older patients with essential tremor. Parkinson Relat Disord. 2013;19:485–6. 10.1016/j.parkreldis.2012.11.014.
- 163.Katznelson NM, Kondakova MS. Method for the synthesis of the amide of 1-ethyl-2-methy valeric acid. Russ Pat; 1934:40850.
- 164.Freifelder M, Geiszler AO, Stone GR. Hydrolysis of 5,5-disubstituted barbituric acids. J Org Chem. 1961;26:203–6. 10.1021/jo01060a048.
- 165.Chambon P, Perio A. Valnoctamide: pharmacological data suggesting anticonvulsant activity. Neurosci Lett. 1980;S5:327.
- 166.Bialer M, Haj Yehia A, Badir K, Hadad S. Can we develop improved derivatives of valproic acid? Pharm World Sci. 1994;16:2–6. 10.1007/BF01870931.
- 167.Barel S, Yagen B, Schurig V, Soback S, Pisani F, Perucca E, et al. Stereoselective pharmacokinetic analysis of valnoctamide in healthy subjects and in patients with epilepsy. Clin Pharmacol Ther. 1997;61:442–9. 10.1016/S0009-9236(97)90194-6.
- 168.Weiser M, Levi L, Levine SZ, Bialer M, Shekh-Ahmad T, Matei V, et al. A randomized, double-blind, placebo and risperidone-controlled study on valnoctamide for acute mania. Bipolar Disord. 2017;19:285–94. 10.1111/bdi.12506.
- 169.Althaus AL, Ackley MA, Belfort GM, Gee SM, Dai J, Nguyen DP, et al. Preclinical characterization of zuranolone (SAGE-217), a selective neuroactive steroid GABAA receptor positive allosteric modulator. Neuropharmacology. 2020;181:108333. 10.1016/j.neuropharm.2020.108333.
- 170.Barnes KN, Vogl CM, Nelson LA. Zuranolone: the first FDA-approved oral treatment option for postpartum depression. Ann Pharmacother. 2024;58(7):728–34. 10.1177/10600280231204953.
Associated Data
Data Availability Statement
No datasets were generated or analyzed as part of the present work.