Exploring the Role of Early Add-On Cenobamate in Uncontrolled Focal Epilepsy: A Critical Appraisal of the Evidence

Abstract

Introduction

Cenobamate is a third-generation antiseizure medication (ASM) for the treatment of uncontrolled focal seizures. As evidence accumulated across randomized clinical trials and real-world studies, clinically relevant questions have emerged regarding its comparative efficacy, optimal timing of introduction, safety and interaction profile, long-term retention, and outcomes beyond seizure control.

Methods

This structured expert narrative review addressed four predefined domains—efficacy, safety and retention, pharmacological interactions, and patient-centered outcomes—using PICO-formulated questions. PubMed searches identified randomized controlled trials, network meta-analyses, and observational studies of adjunctive cenobamate in adults with focal epilepsy. Given heterogeneity in design and reporting, evidence was synthesized narratively and interpreted through multidisciplinary expert discussion.

Results

Across diverse study types, cenobamate consistently demonstrated high efficacy, including clinically relevant rates of seizure freedom and ≥ 50% response. Several real-world analyses suggested that earlier introduction—typically after failure of two prior ASMs—may be associated with higher seizure-freedom rates than early use of other ASMs; these signals derived from non-randomized comparisons and may reflect residual confounding. Safety was dominated by predictable, dose-related central nervous system events, with tolerability strongly dependent on titration and proactive adjustment of interacting ASMs (notably clobazam, phenytoin, phenobarbital, and high-dose sodium-channel blockers). Long-term retention remained high across cohorts. Cognitive performance was generally stable and psychiatric symptoms stable or improved, with quality-of-life gains reported, including in patients who had previously failed resective surgery or neuromodulation, albeit from uncontrolled data.

Conclusions

Cenobamate may offer substantial benefit in adults with uncontrolled focal epilepsy, with a favorable efficacy–tolerability balance, high retention, and possible advantages when introduced earlier (after two ASMs). Its interaction profile underscores the importance of careful management of concomitant ASMs. Because the comparative signals derive largely from non-randomized data, prospective controlled research is needed to confirm any benefit of early use and to better characterize long-term cognitive and psychiatric effects.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40120-026-00997-x.

Key Summary Points

Across randomized, indirect comparative, and real-world studies, adjunctive cenobamate was associated with high responder rates and clinically relevant seizure-freedom rates in adults with uncontrolled focal epilepsy.

Observational studies suggested that use after failure of two prior antiseizure medications (ASMs) may be associated with higher seizure-freedom and retention rates than later introduction or some other early add-on regimens, although these findings are subject to the limitations of non-randomized comparisons.

The safety profile was mainly characterized by dose-related central nervous system adverse events, with tolerability strongly influenced by titration strategy and by management of concomitant ASMs, particularly clobazam, phenytoin, phenobarbital, and high-dose sodium-channel blockers.

Available studies consistently showed reductions in concomitant ASM number or drug load during cenobamate treatment, while long-term retention remained high across multiple cohorts.

Evidence on cognition, psychiatric outcomes, and quality of life was limited and heterogeneous, but overall suggested stability in most patients, with some improvements reported in the context of seizure reduction and polytherapy simplification.

Introduction

Uncontrolled focal epilepsy continues to pose significant challenges. Following the failure of adequate trials of two tolerated, appropriately selected, and effectively used antiseizure medication (ASM) regimens, the probability of attaining seizure freedom with subsequent treatments diminishes markedly. Consequently, many patients experience a prolonged course characterized by incomplete control, resulting in cumulative adverse effects and increasing therapeutic complexity due to polypharmacy [1, 2]. Persistent seizures are associated not only with hospitalizations and increased mortality but also with significant psychosocial burden and reduced autonomy, impairing quality of life. Such unmet needs are thus multidimensional and require the identification of treatment strategies capable of improving seizure outcomes while maintaining acceptable tolerability and minimizing the long-term burden of polytherapy [3–5].

Cenobamate is a third-generation ASM approved in EU for the adjunctive treatment of focal seizures in adults, and as monotherapy in the United States. Its introduction has generated substantial interest due to its pharmacodynamic characteristics and documentation of high efficacy from randomized controlled trials (RCTs) [6–12]. As clinical use has expanded, several lines of evidence have emerged that collectively raise important clinical questions across multiple domains. These include the relative benefit of incorporating cenobamate earlier (after 2 ASMs) rather than later in the therapeutic regimen; its comparative performance with respect to other ASMs; its safety, tolerability, and long-term retention in real-world practice; the nature and clinical relevance of its pharmacological interactions with non-ASMs; and its potential impact on cognitive functions, psychiatric comorbidities, and quality of life. At the same time, the available data are diverse in design, population characteristics, outcome measures, and follow-up duration, making it increasingly necessary to place these findings into a coherent, clinically meaningful framework [13–16].

Given the rapid accumulation of heterogeneous evidence and the growing use of cenobamate in routine clinical practice, the present article aimed at synthesizing the current literature on cenobamate across the domains of efficacy, safety and retention, pharmacological interactions, and non-seizure outcomes. The goal was to provide a clear, comprehensive understanding of the emerging role of cenobamate in the management of uncontrolled focal epilepsy and to support informed decision-making in daily clinical practice.

Methods

This article was developed using an Evidence Synthesis Analysis (ESA) framework, a structured methodological approach designed to identify, evaluate, and integrate heterogeneous clinical data around predefined clinical questions. The process began with a group of epileptologists who defined the most relevant topics for an early clinical adoption of cenobamate. Four thematic domains were identified: efficacy, safety and retention, pharmacological interactions, and outcomes beyond seizures. Specific clinical questions were formulated for each domain using the PICO (Population, Intervention, Comparison, Outcome) model. These questions served as the foundation for the evidence search (Table 1).

Table 1.

PICO questions

PICO 1 – Effectiveness
1.1	Early add-on of cenobamate (as a third therapy) in adults with uncontrolled focal epilepsy is associated with a higher rate of seizure freedom compared with early add-on of other antiseizure medications (ASMs).
1.2	In adults with uncontrolled focal epilepsy, does cenobamate provide greater clinical efficacy compared with other third-generation ASMs?
1.3	Is the combination of cenobamate and clobazam associated with an improved therapeutic response?
1.4	Does the introduction of cenobamate allow a reduction in drug load?
PICO 2 – safety and tolerability
2.1	Is early introduction of add-on cenobamate (as a third therapy) in adults with uncontrolled focal epilepsy associated with better tolerability compared with early use of other ASMs?
2.2	Does cenobamate have a more favorable tolerability profile than other third-generation ASMs in terms of incidence and severity of adverse events?
2.3	Do available data allow comparison of the retention rate of cenobamate with that of other ASMs?
PICO 3 – Drug Interactions and combinations
3.1	Do certain combinations of cenobamate with other ASMs lead to reduced tolerability?
3.2	Is the potential synergism between cenobamate and low-dose clobazam supported by evidence of improved efficacy and tolerability?
3.3	Is there evidence of interactions between cenobamate and non-ASM medications?
PICO 4 – Cognitive, psychiatric, quality of life, post-surgical
4.1	Does the use of cenobamate in adults with focal epilepsy have an impact on cognitive function?
4.2	Does the use of cenobamate in adults with focal epilepsy have an impact on psychiatric status or on the course of psychiatric comorbidities (e.g., depression, anxiety, psychosis)?
4.3	Is the use of cenobamate in adults with focal epilepsy associated with improved quality of life?
4.4	In patients who have failed resective surgery or neuromodulation, is the addition of cenobamate associated with a reduction in seizure frequency?

The detailed characteristics and main results of the articles included in the PICO analysis are provided as Supplementary Material (Supplementary Table S1).

A structured literature search was conducted on PubMed between May and September 2025. This article was conceived and conducted as a structured expert narrative review rather than as a formal systematic review; accordingly, the protocol was not prospectively registered and the synthesis did not incorporate a formal quantitative risk-of-bias assessment. To preserve methodological transparency, the clinical questions were framed using the PICO format and an explicit, reproducible search strategy was applied. The search included randomized controlled trials and their open-label phases (including post hoc analyses), systematic reviews, meta-analyses, network meta-analyses (NMAs), observational studies (e.g., cohort studies, case-control studies) reporting data relevant to at least one of the predefined PICO questions. Studies were eligible for inclusion if they involved adults with focal epilepsy treated with cenobamate as adjunctive therapy and provided extractable information corresponding to at least one of the targeted outcomes. Case reports, case series, narrative reviews, commentaries, and preclinical studies were not eligible as primary evidence for the synthesis. A small number of case-level reports were cited only to illustrate specific mechanistic observations (for example, pharmacogenetic drug–drug interactions) and are explicitly identified as such in the text; they were not used to support effectiveness, safety, or comparative conclusions. Only studies published in English were considered.

For each study meeting the inclusion criteria, data were extracted on characteristics of participants, prior and concomitant ASMs, cenobamate dosing and titration, duration of follow-up, and all available outcomes relevant to efficacy, safety, tolerability, retention, pharmacokinetic or pharmacodynamic interactions, and domains beyond seizure control such as cognition, psychiatric status, and quality of life. Each included record was assigned a Level of Evidence (LoE) based on study design, and the certainty of evidence was appraised according to the GRADE approach [17]. Because the included studies varied widely in design, outcome definitions, and follow-up periods, the synthesis was narrative. Quantitative findings from RCTs and meta-analyses were reported descriptively, while data from observational studies were used to contextualize the external validity and practical applicability of the evidence.

Following completion of the literature review, the evidence was presented to the expert board for collective discussion. The group examined the consistency and clinical relevance of the findings within each thematic area and provided interpretative insights based on shared clinical experience. These considerations were incorporated into the synthesis to facilitate a coherent appraisal of the available data while remaining anchored to the published evidence. The resulting work aimed to offer an integrated and clinically meaningful overview of cenobamate across the four predefined domains and does not constitute formal clinical practice guidelines (Table 2).

Table 2.

Practice points for the use of adjunctive cenobamate in adults with uncontrolled focal epilepsy

Early positioning	Cenobamate may be considered as an early third-line add-on option in adults with uncontrolled focal epilepsy after failure of two adequate ASM trials, as available evidence suggests a higher probability of seizure freedom compared with other alternatives
Dose and retention	At the licensed maintenance dose (200 mg/day), cenobamate appears to have a tolerability profile broadly comparable to other third-generation ASMs. Proactive management of interacting ASMs may minimize the risk of adverse events and contribute to high long-term retention
Polytherapy reduction and combinations	Adjunctive cenobamate may facilitate progressive simplification of background polytherapy, particularly reduction of benzodiazepines and sodium-channel blockers. The cenobamate–clobazam combination may enhance response in selected patients but it may be associated with a higher risk of dose-dependent CNS adverse events
Safety and drug–drug interactions	With current titration protocols, the historical DRESS signal appears substantially mitigated, although continued vigilance remains appropriate. Clinically relevant drug–drug interactions may occur primarily through modulation of CYP2C19 and CYP3A4/2B6 activity, warranting anticipatory dose adjustment, therapeutic drug monitoring when available, and targeted patient counseling
Patient-centered and refractory settings	Available evidence suggests that cenobamate is not generally associated with cognitive or psychiatric deterioration and may support quality-of-life benefits in patients achieving seizure reduction. Clinically relevant responses have also been observed in patients with prior surgical or neuromodulatory treatment failure, although evidence remains uncontrolled

This article is based exclusively on previously published studies and does not report any new study involving human participants or animals performed by any of the authors. Ethics committee approval and informed consent were therefore not required.

Results

Area 1 – Efficacy

PICO 1.1 – Early Use and Seizure Freedom

Across the three complementary search strings, 66 unique records were identified and screened, of which 22 full-text articles were assessed for eligibility [3, 9–12, 14, 15, 18–32]. Indirect comparative evidence from RCTs and NMAs (LoE 1a), although not restricted to early-line use and therefore considered ancillary, consistently ranked cenobamate among the ASMs with the highest probability of achieving seizure freedom [9, 20]. Consequently, only studies stratifying outcomes by treatment line and comparative observational cohort studies specifically examining early adjunctive use were deemed relevant.

Among these, the BLESS multicenter cohort study provided line-specific data: individuals treated with cenobamate as an early add-on after two or three prior failed ASMs achieved a 22.4% sustained seizure-freedom rate at 24 weeks, compared with 8.7% in patients receiving cenobamate in later regimens (LoE 2b) [27, 28]. Additional real-world single-arm cohort studies, reported seizure-freedom rates mostly in the ~ 10–20% range over 6–12 months of treatment, with studies noting numerically higher seizure-freedom proportions in patients with fewer previous treatment failures (LoE 2b–4) [14, 23, 26].

One study fulfilled the criteria for a direct comparative assessment in an early-line context, directly addressing this PICO question. In the matched observational cohort study by Winter et al., early add-on cenobamate—introduced after failure of two or three prior ASMs—was associated with higher 12-month seizure-freedom rates compared with early add-on treatment with other commonly used ASMs (lacosamide, levetiracetam, valproate, or topiramate), analyzed as a pooled comparator (19.5% vs. 8.3%), alongside higher responder and retention rates (LoE 2b) [20].

PICO 1.2 – Comparative Efficacy Versus Other Third-Generation ASMs

In total, 31 unique records were screened and 11 full-text articles were assessed [9, 15, 20, 21, 24, 33–38].

Across quantitative syntheses, cenobamate was consistently placed among the most effective third-generation ASMs. In the NMA by Lattanzi et al., cenobamate ranked best for both ≥ 50% responder rate and seizure freedom, with significantly higher odds of ≥ 50% seizure reduction than brivaracetam, eslicarbazepine acetate, lacosamide, and perampanel; seizure-freedom estimates also favored cenobamate but with wide confidence intervals and no consistent statistical separation from all comparators [9]. The indirect treatment comparison by Privitera et al. showed a higher placebo-adjusted ≥ 50% responder rate for cenobamate without increased discontinuations due to adverse events (AEs) compared with pooled branded third-generation ASMs [36]. The broader NMA by Yang et al. confirmed cenobamate as the drug with the highest probability of achieving a ≥ 50% responder rate among third-generation ASMs [33], while the analysis by Mulheron et al. likewise positioned cenobamate at or near the top of the efficacy hierarchy of newer ASMs [15]. However, these indirect comparisons are inherently limited by differences in trial design, patient populations, baseline seizure frequency, and titration schedules across studies, which may affect the relative ranking of treatments.

Real-world comparative data extended these findings to clinical effectiveness and retention [20, 21, 24, 38]. In the CENKORK study, at 12-month follow-up cenobamate achieved ≥ 50% responder rate of 61%, seizure-freedom rate of 14%, and retention rate of 80.0% in a highly refractory population, and outcomes compared favorably with historical third-generation ASM cohorts [24]. In the matched cohort by Winter et al., early adjunctive cenobamate (after failure of two or three prior ASMs) was associated with higher 12-month effectiveness than early add-on lacosamide, levetiracetam, valproate, and topiramate, including higher ≥ 50% responder rates (71.4% vs. 52.5%), higher seizure-freedom rates (19.5% vs. 8.3%), and better retention (92% vs. 62.5–80%) [20]. As an additional note, economic evaluations based on efficacy estimates derived from the RCTs reported the lowest numbers needed to treat for both ≥ 50% response and seizure freedom with cenobamate compared with other third-generation ASMs and identified cenobamate as a cost-effective or dominant option versus brivaracetam, eslicarbazepine acetate, lacosamide, and perampanel [38].

PICO 1.3 – Cenobamate + Clobazam and Therapeutic Response

For this PICO, 19 unique records were screened and 13 full-text articles were assessed [5, 8, 14, 39–48]. Across the available literature, no randomized trial and no predefined comparative study directly evaluating cenobamate plus clobazam versus cenobamate alone were identified. The evidence base consisted predominantly of real-world observational cohorts (LoE 2b) and small mechanistic or case-based publications (LoE 4–5), providing indirect information on clinical response and safety. Several retrospective cohorts reported responder and seizure-freedom rates in adults treated with adjunctive cenobamate within complex polytherapy regimens, where clobazam was frequently co-administered [5, 14, 43, 44]. In these series, exploratory subgroup or regression analyses suggested that concomitant clobazam or other GABAergic ASMs might be associated with higher probabilities of response or seizure freedom; however, none of the studies included prospectively defined cenobamate-alone versus cenobamate-plus-clobazam arms, and potential confounding by baseline characteristics, dosing, and treatment history was not fully controlled [14, 44]. Therefore, any suggestion of pharmacodynamic synergism remains hypothesis-generating and cannot be considered established based on currently available evidence.

Safety findings were more consistent across pharmacokinetic, therapeutic drug monitoring (TDM), and expert opinion sources. Cenobamate inhibits CYP2C19 and can substantially increase N-desmethylclobazam exposure during co-treatment, increasing the risk of dose-dependent somnolence and other CNS adverse effects. Consensus-based recommendations and Delphi guidance uniformly endorsed proactive clobazam down-titration during cenobamate titration (depending on clobazam dose) to mitigate sedation and optimize tolerability. Case-level pharmacogenetic evidence further illustrated that, in poor CYP2C19 metabolizers, the cenobamate–clobazam interaction can result in prolonged elevation of N-desmethylclobazam and sustained excessive sedation [39, 47].

PICO 1.4 – Drug-Load Reduction with Adjunctive Cenobamate

A total of 57 unique records were screened and 20 full-text articles were assessed [23, 25, 27, 28, 32, 42–44, 47, 49–59]. Across all available evidence—derived largely from real-world observational cohorts and post hoc analyses of open-label phase-3 studies (LoE 2b, with some LoE 4)—adjunctive cenobamate was consistently associated with reductions in the number and total daily dose of concomitant ASMs.

In post hoc analyses of a phase-3 open-label extension, mean total concomitant ASM drug load decreased by 29.4% at 12 months and 31.8% at 24 months of adjunctive cenobamate, with reductions observed across all ASM classes and the largest relative decrease reported for benzodiazepines (≈ 55% at 24 months). Importantly, maintenance of ≥ 50% and 100% response rates was preserved across subgroups despite these reductions, indicating that cenobamate-based regimens allowed drug-load reduction without compromising long-term seizure control [55].

Real-world cohorts provided convergent findings. Multicenter and single-center studies reported substantial decreases in both ASM number and total defined daily dose (DDD) within 3–12 months of cenobamate introduction, with many patients discontinuing at least one concomitant ASM and a sizeable proportion reducing doses across several medication classes [23, 25, 32, 43, 44]. Across series, the burden of polytherapy frequently decreased by one or more ASMs, and median concomitant ASM load (by count or DDD) declined significantly over follow-up. Reductions were most commonly observed for sodium-channel blockers and GABAergic agents—particularly clobazam—as well as perampanel and other drugs with known pharmacokinetic interactions with cenobamate [23, 25, 32, 43, 44]. A recent Spanish real-world retrospective study further confirmed these patterns, showing sustained reductions in concomitant ASM use during long-term cenobamate treatment in routine clinical practice [58].

Several large-scale registries and expanded-access or multicenter programs showed that drug-load simplification often began within the first 12–24 weeks of cenobamate addition and continued over time [27, 55]. In some studies, a minority of patients transitioned to dual therapy or, less frequently, cenobamate monotherapy by the end of follow-up. Improvements in tolerability and resolution of adverse effects were frequently attributed to stepwise de-escalation of background ASMs [23, 28, 43]. Notably, recent long-term real-world data found significant improvements in cognitive scores and a reduction in the concomitant ASM drug load from baseline to month 18, further supporting the clinical relevance of rational ASM de-escalation when cenobamate is added to the therapy [59]. It should be acknowledged that reductions in concomitant ASM load may partly reflect clinician-driven optimization strategies following seizure improvement, rather than a direct pharmacological effect of cenobamate itself.

Area 2 – Safety

PICO 2.1 – Early Use and Tolerability Versus Other ASMs

A total of 45 unique records were screened, 7 full-text articles were evaluated, and 3 additional studies were retrieved through hand-searching [9, 15, 18, 20, 24, 32, 33, 36]. No head-to-head RCTs comparing early adjunctive cenobamate with other ASMs were identified.

One real-world study addressed early add-on use. In the matched cohort by Winter et al. (LoE 2b), early adjunctive cenobamate showed no increase in AEs compared with matched early add-on comparators (lacosamide, levetiracetam, valproate, topiramate), and no discontinuations for AEs occurred in any group over 12 months [20].

PICO 2.2 – Comparative Safety Versus Other Third-Generation ASMs

A total of 67 unique records were screened and 20 full-text articles were assessed for eligibility [5, 9, 11, 12, 15, 18, 24, 26, 31, 33, 36, 37, 50, 60–66]. No head-to-head RCTs comparing cenobamate with other third-generation ASMs were identified. The comparative evidence base was therefore composed of indirect comparisons from NMAs of placebo-controlled RCTs (LoE 1a) and a small number of real-world analyses providing active-comparator data (LoE 2b–3b).

Across four NMAs, cenobamate showed a tolerability profile comparable to other third-generation ASMs [9, 15, 18, 33]. At licensed maintenance doses of cenobamate (200 mg/day), the risk of treatment-emergent adverse events (TEAEs) and TEAE-related discontinuation was broadly similar to brivaracetam, lacosamide, perampanel, and eslicarbazepine acetate, with no statistically significant differences in pairwise comparisons. At higher cenobamate doses (400 mg/day), TEAE incidence and discontinuation increased, mirroring the dose-dependence observed within cenobamate RCTs; analyses comparing maximal doses across ASMs showed no significant tolerability advantage for cenobamate. Rash and other cutaneous adverse events were infrequent across the network. Suicidality was specifically assessed in one pooled analysis, with no increased risk for cenobamate versus placebo, but no head-to-head comparisons with other ASMs [9, 15, 18, 33].

Real-world comparative evidence was limited but consistent. In the CENKORK cohort (LoE 2b), cenobamate showed AE frequencies comparable to lacosamide and brivaracetam and lower than perampanel, with AE-related discontinuation at 12 months of 8% [24]. In the early-line matched cohort by Winter et al. (LoE 2b), cenobamate demonstrated similar rates of adverse events and no AE-related discontinuations compared with matched early-add-on lacosamide, levetiracetam, valproate, and topiramate [20]. A pharmacovigilance analysis by Shang et al. using the FDA Adverse Event Reporting System (FAERS) (LoE 3b) suggested differing patterns of AE reporting—predominantly CNS-related for cenobamate and additional cardiac signals for lacosamide—although true incidence, causality, and comparative risk cannot be established due to underreporting, reporting bias, and lack of denominator data inherent to spontaneous reporting systems [60]. Recent large-scale pharmacovigilance analyses further contextualized psychiatric safety. A disproportionality study from Li et al. of newer ASMs using FAERS data (LoE 3b) identified heterogeneous psychiatric AE reporting patterns across agents, with the evidence that cenobamate was not associated with severe psychiatric outcomes compared with other newer ASMs. Although limited, such analyses support the absence of a strong psychiatric safety concern for cenobamate at the population-reporting level [67].

Finally, although not designed as safety studies per se, systematic evaluations of cognitive-monitoring tools such as EpiTrack underscored the importance of structured neuropsychological surveillance when interpreting CNS tolerability across ASMs. These data did not identify cenobamate-specific cognitive safety concerns but highlighted the challenges of attributing subtle executive or memory changes to individual ASMs in polytherapy settings [68].

PICO 2.3 – Retention Rate vs. Other ASMs

A total of 35 unique records were screened and 3 full-text articles were assessed for eligibility, with 2 additional studies retrieved by hand-searching [20, 24, 35, 69, 70]. Only 3 studies provided active-comparator data suitable for evaluating retention rates versus other ASMs, all based on real-world observational design (LoE 2b). No randomized or prospective head-to-head trials were identified.

In the matched cohort by Winter et al. (LoE 2b), early adjunctive cenobamate demonstrated a significantly higher 12-month retention rate compared with other ASMs used at similar treatment lines. Retention was 92.0% for cenobamate versus 80.0% for lacosamide, 73.3% for levetiracetam, 68.2% for valproate, and 62.5% for topiramate (p < 0.05). No discontinuations for adverse events occurred in the cenobamate group [20]. In the CENKORK study by Steinhoff et al. (LoE 2b), cenobamate retention was 88.7% at 6 months and 80% at 12 months in a refractory population. At 6 months, retention significantly exceeded that of historical cohorts treated with lacosamide, perampanel, or brivaracetam (p < 0.001), though comparisons were non-concurrent [24].

A large claims-based retrospective analysis by Stern et al. (LoE 2b) reported 12-month retention of 50.3% for cenobamate versus 40.5% for pooled branded third-generation ASMs (p = 0.045). Median time to discontinuation exceeded 12 months for cenobamate versus 7.7 months for the comparator group. Retention advantages reached statistical significance versus lacosamide and perampanel, while numerical differences versus brivaracetam and eslicarbazepine acetate were observed but not reported as significant [35]. Long-term open-label extension data [69, 70] provided retention estimates for cenobamate up to 2–8 years (73–80% at 1 year; 59–72% at 2 years). Nevertheless, comparator data were not available within the same study framework; therefore, any comparison would be indirect and across different trials, which are characterized by differences in populations, designs, and follow-up procedures . Retention rates should be interpreted cautiously, as they may reflect a composite of efficacy, tolerability, patient expectations, and clinician preference, and are subject to channeling and survivorship bias in observational datasets.

Area 3 – Pharmacological Interactions

PICO 3.1 – Tolerability of Specific ASM Combinations

A total of120 unique citations were screened, and 28 full-text articles were assessed for eligibility, with no additional studies retrieved by hand-searching [4, 5, 9, 13, 14, 25, 32, 39, 41–44, 47, 48, 50, 52, 53, 64, 69–78]. Evidence was derived predominantly from real-world observational cohorts and post hoc analyses of randomized or open-label studies (LoE 2–4), with several expert-consensus documents (LoE 5). No head-to-head RCTs of cenobamate with specific concomitant ASM classes were identified.

Across post hoc analyses of RCTs and long-term extension studies, tolerability was dominated by dose-related CNS adverse events—somnolence, dizziness, ataxia, diplopia—with higher incidence among patients receiving multiple concomitant ASMs [24, 42, 64]. In the randomized post hoc analysis by Brandt et al. (LoE 2), concomitant sodium-channel blockers were associated with higher rates of dizziness and more AE-related discontinuations at the 400-mg cenobamate daily dose, whereas overall AE profiles remained similar between SCB and non-SCB backgrounds [64]. In the RCT pooled analysis by Rosenfeld et al. (LoE 2), patients on more than two concomitant ASMs had the highest rates of CNS AEs and AE-related withdrawal, independent of specific class [42]. Post hoc analyses of open-label cohorts showed that early reductions of phenytoin, phenobarbital, clobazam, valproate, and lacosamide were frequently required to mitigate sedation, dizziness, or ataxia, and were associated with higher long-term continuation of cenobamate [70].

Large real-world cohorts provided convergent results. In Badr et al. (LoE 4), psychiatric comorbidity predicted AEs, whereas no concomitant ASM class—including SCBs, GABAergics, SV2A modulators, or perampanel—was independently associated with increased AE incidence or reduced retention [48]. However, a dedicated TDM-based real-world study by Bender et al. focusing on brivaracetam suggested a measurable pharmacokinetic interaction. In a retrospective cohort of patients on stable brivaracetam, the introduction of cenobamate was associated with a statistically significant mean increase (~ 27%) in brivaracetam serum concentrations, which was only infrequently accompanied by adverse effects, but prompted brivaracetam dose reduction in a minority of cases due to transient fatigue, consistent with a CYP2C19-mediated mechanism (LoE 4) [66].

In Beltrán-Corbellini et al. (LoE 4), SCB reductions were common during titration and at follow-up (61.7% at 3 months; 52% at last visit), but no specific class consistently increased AE burden [76]. In Bosak et al. (n = 475; LoE 4), clobazam was associated with increased odds of AEs (OR 2.46), although confounding by indication and polytherapy intensity cannot be excluded, while carbamazepine was associated with lower AE risk; SCBs were linked to lower somnolence but higher dizziness in univariate analyses [32]. In the Danish cohort studied by Gjerulfsen et al. (LoE 4), lacosamide was most frequently reduced and lamotrigine most often discontinued; AE incidence was lower when cenobamate was introduced as a third ASM rather than in heavier polytherapy [52]. In Rodríguez-Uranga et al. (LoE 4), early CNS adverse effects were mitigated by substantial reductions in SCBs and clobazam, with AE prevalence decreasing from 71.9% to 52.3% between 3 and 12 months [43].

Further real-world studies corroborated combination-dependent tolerability effects. In Connor and Williamson et al. (LoE 4), dizziness, drowsiness, and ataxia often resolved after reducing lacosamide or clobazam; all five patients treated with clobazam required dose reduction [53]. In Peña-Ceballos et al. (LoE 4), AEs were frequently associated with higher cenobamate daily doses (≥ 250 mg) combined with multiple CNS-active ASMs, improving after reductions of clobazam, eslicarbazepine acetate, and perampanel [25]. In the time-to-event analysis by Roberti et al. (LoE 4), somnolence, dizziness, and balance disturbances were more frequent in SCB-containing regimens, with exploratory hazards suggesting later and fewer AEs in patients on GABAergic-only co-therapy [73]. Pooled long-term data from the cenobamate development program (LoE 2b) showed no major differences in retention across commonly co-administered ASMs, although class-specific AE comparisons were not available [69].

Additional mechanistic or special-population data reinforce these patterns. In O’Dwyer et al. (LoE 2), older adults experienced higher rates of somnolence, dizziness, and falls, particularly at higher cenobamate doses and in the presence of CYP2C19-substrate ASMs; the authors recommended early reduction of clobazam, phenytoin, and phenobarbital [50]. Case-level evidence by Zillgitt et al. (LoE 5) illustrated profound sedation from the cenobamate–clobazam combination in a CYP2C19 poor metabolizer, driven by persistent elevations of N-desmethylclobazam [39].

Lastly, expert consensus statements (LoE 5) uniformly recommended proactive reduction of clobazam, phenytoin, phenobarbital, and high-dose SCBs during cenobamate titration, with reactive reductions for other ASMs based on emerging CNS-related AEs [41, 47, 74].

PICO 3.2 – Synergism with Low-Dose Clobazam

A total of 21 unique citations were screened and a0 full-text articles were assessed for eligibility, with 1 additional study retrieved by hand-searching [5, 14, 39–42, 44, 47, 73, 79, 80]. Across all sources, no randomized or prospective comparative study evaluating cenobamate plus low-dose clobazam versus clobazam alone was identified. Evidence was derived from real-world cohorts with exploratory subgroup analyses (LoE 2b–4), pharmacokinetic investigations (LoE 4), expert-consensus recommendations (LoE 5), and isolated case reports (LoE 5). Comparative efficacy between cenobamate-alone and cenobamate-plus-clobazam cannot be determined based on available data.

Expert consensus publications (LoE 5) consistently described the pharmacokinetic effect whereby cenobamate inhibits CYP2C19 and increases N-desmethylclobazam concentrations, often by 2–6-fold. These elevations are associated with somnolence, dizziness, and psychomotor slowing, typically occurring early during titration. Panels uniformly recommended proactive clobazam dose reduction at low cenobamate doses (50–100 mg/day) and maintaining low-dose clobazam (5–10 mg/day) when clinically useful. These practice statements referenced a possible efficacy benefit of the combination but provided no controlled data [41, 47, 74].

Real-world cohorts provided indirect clinical signals. In Pietrafusa et al. (LoE 4), clobazam was used in ~ 41% of patients, with marked increases in N-desmethylclobazam levels correlating with cenobamate dose (r = 0.74). Clinical improvement was observed across the cohort (59.2% responders, 20.2% seizure-free), but no cenobamate-alone comparator was available [14]. In the Italian Expanded Access Program (EAP) by Roberti et al. (LoE 4), patients receiving clobazam showed a non-significant trend toward higher responder rates, while tolerability remained acceptable after clobazam dose reduction. Most common AEs were somnolence, dizziness, and balance disturbances; clobazam and sodium-channel blockers accounted for the largest background dose reductions [44]. The time-to-event analysis from the same cohort (LoE 2b) showed fewer and later AEs in regimens containing only GABAergic co-therapy, though clobazam-specific effects were not isolated [73].

Additional observational datasets supported these patterns. In Loushy et al. (LoE 4), clobazam use was independently associated with 100% seizure-reduction outcomes in multivariable analysis, although no plasma levels were recorded and no direct comparator was provided [5]. In Rosenfeld et al. (LoE 2; post hoc RCT analysis), cenobamate efficacy was consistent across co-therapy strata, including clobazam, and tolerability worsened with heavier polytherapy [42]. A single case report by Zillgitt et al. (LoE 5) documented profound sedation due to elevated N-desmethylclobazam in a CYP2C19 poor metabolizer receiving cenobamate and clobazam [39].

PICO 3.3 – Interactions with Non-ASM Medications

A total of 15 unique citations were screened, and 2 full-text articles were assessed for eligibility, with 3 additional studies identified through hand-searching [8, 81–83]. No randomized or prospective comparative studies assessing cenobamate with versus without non-ASM co-medications were found. Evidence consisted of one phase-I drug–drug interaction study in healthy volunteers (LoE 2), two small retrospective clinical cohorts (LoE 4), and one narrative review summarizing PK findings (LoE 5). Clinical outcome data related to non-ASM interactions remain limited.

In the phase-I interaction study by Greene et al. (LoE 2), co-administration of cenobamate 200 mg/day produced a mixed enzymatic profile: induction of CYP3A4/5 reduced midazolam exposure by 72%, induction of CYP2B6 reduced bupropion exposure by 39%, and inhibition of CYP2C19 increased omeprazole exposure by 107%, with minimal effect on warfarin (CYP2C9 substrate). Adverse events were mostly mild, mainly represented by somnolence [81]. A retrospective cohort in patients with tuberous sclerosis complex treated with mTOR inhibitors by Becker et al (LoE 4) showed decreases in everolimus serum levels after cenobamate initiation (median 5.1→3.4 ng/mL, p = 0.012), consistent with CYP3A induction. Sirolimus levels were unchanged. Seizure outcomes remained stable (45% reduction; 10% seizure-free), and AE reporting was limited [82].

A single-center descriptive analysis by Jimenez et al. (LoE 4) evaluated multiple non-ASM classes in adults treated with cenobamate. Signals included two pregnancies among hormonal contraceptive users, absence of INR changes among anticoagulant users, modest lipid variability, and increased HbA1c in insulin-treated diabetics. Psychiatric hospitalizations occurred in four patients receiving psychotropic medications. The authors noted limitations due to sample size and retrospective design [83]. Lastly, a narrative review by Roberti et al. (LoE 5) summarized mechanistic evidence from healthy-volunteer studies, confirming cenobamate’s CYP2C19 inhibition and CYP3A4/2B6 induction. Reported interaction patterns included increased exposure of proton-pump inhibitors and decreased exposure of estrogen–progestin contraceptives and CYP3A substrates. Additive CNS depression with sedatives was noted. The effects on warfarin exposure were minimal [8]. These findings underscore the importance of systematic medication review when initiating cenobamate, particularly in patients receiving CYP3A or CYP2C19 substrates with narrow therapeutic windows.

Area 4 – Beyond Seizures

PICO 4.1 – Cognition

A total of 7 unique records were screened and all were assessed in full text [13, 14, 16, 59, 84–86], of which 6 met the PICO criteria and provided longitudinal cognitive evaluations in adults with focal epilepsy treated with adjunctive cenobamate; 1 pediatric retrospective study was excluded due to population mismatch and absence of validated cognitive measures.

In the prospective cohort by Catalán-Aguilar et al. (LoE 3), repeated neuropsychological testing at baseline, 3 months, and 6 months of cenobamate demonstrated cognitive stability across multiple domains, including attention, executive function, memory, naming, and processing speed. Reliable Change Index analyses confirmed the absence of clinically meaningful declines at individual-patient level. Mood and quality-of-life measures (BDI-II, STAI-T, QOLIE-31) remained stable [16].

In the real-world cohort by Pietrafusa et al. (LoE 4), a subset of adults (n = 32) underwent neuropsychological assessment at baseline and after 6 months of cenobamate. No deterioration was observed on composite cognitive scales. Processing speed improved significantly on TMT-A (68.3 s → 41.7 s; p = 0.01), with a trend toward improvement on Raven’s matrices (p = 0.05). QOLIE-31 cognitive subscores were unchanged, while anxiety scores improved (GAD-7: 6.8 → 4.5; p = 0.03). Authors noted that background ASM reduction during cenobamate therapy may have contributed to these patterns [14].

In Schuetz et al. (LoE 4), cognitive performance assessed with EpiTrack at baseline and after reaching the first target dose (typically ~ 3 months; median cenobamate dose 125 mg/day) was stable in 72% of patients, improved in 16%, and deteriorated in 12%, with a significant group-level improvement (p < 0.01) and no association with cenobamate dose or changes in drug load [85]. In Serrano-Castro et al. (LoE 4), a retrospective Spanish EAP cohort (n = 20) assessed at baseline and 6 months showed significant improvements in verbal and visuospatial episodic memory measures, alongside worsening in attention (p = 0.030); concomitant ASM number and drug-load metrics decreased significantly, and covariance analyses suggested cognitive changes were more closely related to reduced pharmacological burden than to seizure control [13]. In Witt et al. (LoE 4), available at the time of writing only as a preprint (medRxiv) and therefore not yet peer-reviewed, a retrospective longitudinal cohort (n = 84) found a dose-related signal: cenobamate doses ≥ 300 mg/day were associated with a significant decline in executive functions (EpiTrack; p = 0.014; correlated with dose), and, in a subgroup with extended memory testing (n = 22), verbal retention declined at group level independent of dose (p = 0.011), supporting the need for dose optimization and cognitive monitoring, particularly at higher doses [86]. In the exploratory real-world study by Ramírez-García et al. (LoE 4), neuropsychological testing at baseline, 6 months, and 18 months showed no cognitive deterioration and improvements in processing speed and episodic memory domains, in parallel with seizure reduction and concomitant ASM simplification [59]. Overall, available data suggest cognitive neutrality at standard doses in most patients, with a possible dose-related executive signal at higher doses (≥ 300 mg/day). However, interpretation is complicated by concomitant ASM reductions and seizure improvement, which may independently influence neuropsychological outcomes.

PICO 4.2 – Psychiatric Sphere and Comorbidities

A total of 24 unique citations were screened and 5 full-text articles were assessed [5, 14, 16, 62, 87]. No RCTs or active-comparator studies formally evaluating psychiatric outcomes were found. Evidence was derived from 1 prospective cohort with serial validated scales (LoE 3), 3 observational real-world studies (LoE 4) and 1 meta-analysis (LoE 1).

In the prospective study by Catalán-Aguilar et al. (LoE 3), repeated measures of depression and anxiety (BDI-II, STAI-T) at baseline, 3 months, and 6 months of cenobamate showed no significant changes, with all group results and Reliable Change Index analyses indicating stability [16]. In Pietrafusa et al. (LoE 4), depressive symptoms remained unchanged, while anxiety scores improved significantly at 6 months (GAD-7: 6.8→4.5; p = 0.03). No worsening was detected on QOLIE-31 mood domains [14]. A systematic review and meta-analysis by Klein et al. (LoE 1) reported no increased risk of suicidality with newer ASMs, including cenobamate, compared with placebo across RCT datasets [62].

In the large retrospective cohort by Loushy et al. (LoE 4), psychiatric AEs were uncommon: one case of severe depression was reported, with no systematic mood deterioration across the cohort [5]. An expert report by Krauss et al. (LoE 5) summarized psychiatric safety data across RCTs, extensions, and prospective observational evidence, noting low rates of psychiatric AEs without identifiable worsening trends [87]. While current evidence does not indicate a major psychiatric safety signal, the absence of dedicated prospective psychiatric trials limits definitive conclusions.

PICO 4.3 – Quality of Life

A total of 27 unique citations were screened and 5 full-text articles were assessed [16, 23, 76, 88, 89]. No randomized or active-comparator studies were identified. Evidence was derived from 1 prospective cohort with validated QOLIE testing (LoE 3), 2 retrospective real-world studies using clinical global impression (CGI) scales (LoE 4), 1 mixed-design case report plus systematic review in non-focal syndromes (LoE 5), and 1 long-term open-label cohort from cenobamate extension studies (LoE 4).

In the prospective study by Catalán-Aguilar et al. (LoE 3), QOLIE-31 composite and domain scores showed no significant group-level change at 3 or 6 months of cenobamate. Reliable Change Index analyses indicated that most individuals remained stable, with a minority showing improvement or worsening without a consistent pattern [16]. In Beltrán-Corbellini et al. (LoE 4), global patient-perceived status improved on CGI-I: 66% were rated as “clinically significantly improved” at 3 months and 49% at last follow-up. No QOLIE measures were collected [76]. In the German multicenter cohort by Strzelczyk et al. (LoE 4), CGI-C ratings indicated global improvement in 53.4% of patients, minimal improvement in 13.2%, stability in 24.4%, and worsening in 8.9% at final follow-up. Standardized QOLIE scales were not administered [23]. In the mixed-design study by Vlakou et al. (LoE 5), qualitative caregiver-reported improvements in alertness and daily activities were noted in a single adult case, although no validated QoL measures were used. The accompanying systematic review included non-focal populations and did not provide standardized QoL outcomes [88].

In the long-term single-center cohort by Elizebath et al. (LoE 4), among 49 adults with drug-resistant focal epilepsy treated with adjunctive cenobamate and followed for up to 8 years (median ~ 5.6 years), QOLIE-31 scores at latest follow-up were higher among patients achieving ≥ 90% seizure reduction, suggesting quality-of-life improvement in those with substantial clinical response; no baseline-to-follow-up QOLIE comparison was available, and most patients (76%) remained on cenobamate through the long-term observation period [89]. Taken together, quality-of-life data remain limited and heterogeneous, with improvements appearing to parallel seizure response rather than reflecting an independent treatment-specific effect.

PICO 4.4 – Post-surgical/Neuromodulation Failure

A total of 28 unique citations were screened and 7 full-text articles were assessed, with no additional studies retrieved [3, 25, 90–94]. No randomized or controlled comparative studies were identified. Evidence was derived from post hoc analyses of long-term open-label trials (LoE 2), multicenter and single-center real-world cohorts (LoE 4), one case report (LoE 5), and one expert consensus document (LoE 5). All available studies used within-patient comparisons before and after cenobamate initiation.

In the post hoc analysis by Abou-Khalil et al. (LoE 2), 85 adults with prior epilepsy surgery achieved substantial seizure reduction with adjunctive cenobamate: 30.6% achieved ≥ 12-month seizure freedom at any interval, 23.5% were seizure-free at last visit, and, among continuers, 29.2% maintained ≥ 12 months seizure freedom at data cutoff. Retention was 76.5% [94]. In the companion post hoc analysis by Aboumatar et al. (LoE 2), drug-load reduction and cenobamate effectiveness were evaluated over 24 months. In patients with prior surgery, total ASM load decreased by ~ 35% while ≥ 50% response was maintained in 86–91% across 12–24 months [93]. In the multicenter real-world study by Hogeveen et al. (LoE 4), adults with prior or current VNS (n = 54) treated with adjunctive cenobamate achieved ≥ 50% responder rates of 44.4%, seizure freedom in 11.1%, and a median seizure reduction of 32%. Retention was 87%, with AE-related discontinuation in 13%. Better outcomes were observed in patients using VNS with auto-stimulation enabled [91].

In Peña-Ceballos et al. (LoE 4), a highly refractory cohort (n = 57) included 56% participants with prior epilepsy surgery and 59.6% with VNS (87.8% with at least one prior surgical or neuromodulation intervention). Overall, ≥ 50% responder rate was 70.2%, with seizure freedom in 5.3%. Subgroup responder rates were similar in VNS (66.7%) and patients with prior surgery (81.3%) [25]. In the multicenter EAP reported by Rheims et al. (LoE 4), 41.9% of adults had prior epilepsy surgery. At 3-month of cenobamate maintenance, ≥ 50% responder rate was 49.3%; at 6-month maintenance, seizure freedom reached 24.2%. Subgroup-specific results for patients who have had surgery were not separately reported [3]. A single-patient case report by Atanasio et al. (LoE 5) described seizure freedom after VNS implantation in a patient previously on cenobamate; attribution of effect to cenobamate alone was not possible due to the study design [92]. The expert consensus document by Laxer et al. (LoE 5) recommended cenobamate as a treatment option during presurgical evaluation and after unsuccessful surgery or neuromodulation, referencing published post hoc and real-world data but providing no new outcomes [90].

Discussion

On the basis of the available evidence and the thematic considerations discussed above, a set of practice-oriented considerations was formulated to support clinical interpretation of the findings. These reflect areas of relative consistency across studies while acknowledging the limitations and uncertainties inherent to the current evidence base (Table 2).

Early Use and Comparative Efficacy

Cenobamate consistently emerged as a highly effective option for adults with focal epilepsy, including when introduced early after failure of two prior ASMs. Across network meta-analyses, cenobamate was repeatedly ranked among the agents with the highest probability of achieving both ≥ 50% response and seizure freedom, although indirect pairwise comparisons remain underpowered and should be interpreted cautiously. Real-world early-line cohorts further suggested that introducing cenobamate after two or three prior ASMs may be associated with a higher likelihood of seizure freedom compared with early use of lacosamide, levetiracetam, valproate, or topiramate, despite the inherent limitations of observational data [7, 24, 26]. Because these signals derive exclusively from non-randomized comparisons subject to residual confounding and channeling bias, they should be interpreted as associations rather than as evidence of a causal benefit of early use, and require confirmation in prospective controlled studies.

Overall Tolerability and Dose Considerations

Across placebo-controlled RCTs and network meta-analyses, cenobamate at the 200-mg/day maintenance dose demonstrated a risk of treatment-emergent AEs broadly comparable to other third-generation ASMs at their respective maintenance doses. Brivaracetam and lacosamide generally ranked slightly better for tolerability, whereas cenobamate showed an acceptability profile similar to perampanel and eslicarbazepine acetate. Higher cenobamate doses (400 mg/day) were associated with increased CNS AEs and discontinuations, consistent with dose–response patterns observed in individual trials. No randomized comparison has evaluated early add-on cenobamate versus alternative ASMs for tolerability, but real-world early-line matched studies suggested that cenobamate may not be associated with an increased AE burden when titrated appropriately. DRESS remains a notable historical safety concern for cenobamate. However, with the current slow titration protocol, no new cases have been reported in controlled trials, and post-marketing evidence has not suggested a disproportionate risk compared with other ASMs. Although the rarity of DRESS and the lack of dedicated comparative studies prevent definitive exclusion of risk, accumulated clinical experience indicates that DRESS is uncommon and appears manageable under recommended titration schedules. Continued vigilance remains appropriate, but the available evidence supports a substantially mitigated risk in contemporary practice [24, 26].

Retention as an Integrated Measure of Effectiveness

Retention findings integrated efficacy, tolerability, and patient preference. Across comparative cohorts, cenobamate generally showed higher or comparable 12-month retention compared with lacosamide, levetiracetam, valproate, and topiramate, and exceeded pooled third-generation branded ASMs in claims-based analyses. In highly refractory populations, retention around 80% at 12 months and sustained continuation in long-term extensions could suggest a favorable long-term benefit–risk balance for many patients. However, retention is influenced by prescribing practices, healthcare systems, and patient expectations; thus, it should be interpreted alongside detailed safety and quality-of-life data.

Polytherapy Reduction and Combination-Dependent Tolerability

Across 12–24 months, overall ASM load often decreased by 20–35%, particularly for benzodiazepines and sodium-channel blockers, without apparent loss of seizure control and with improved tolerability, suggesting that cenobamate may facilitate stepwise de-escalation of polytherapy. At the same time, tolerability appeared strongly dependent on the background regimen. Combinations involving clobazam, phenytoin, phenobarbital, and stacked high-dose sodium-channel blockers may be associated with higher rates of somnolence, dizziness, ataxia, and balance disturbances, particularly during titration. Early dose reduction of these agents—especially clobazam and sodium-channel blockers—may mitigate CNS AEs and support long-term continuation [10].

Cenobamate–Clobazam: Effectiveness Versus Tolerability

Several real-world cohorts suggested that co-administration of clobazam may be associated with higher responder rates or greater seizure reduction compared with regimens not including clobazam, although these findings remained exploratory and uncontrolled. At the same time, the combination can increase the likelihood of dose-dependent CNS adverse effects due to cenobamate-mediated elevation of N-desmethylclobazam and may require proactive dose adjustments. Whether any particular sequencing of these two agents confers a clinical advantage cannot be determined from the available evidence and remains a hypothesis to be tested in prospective studies; it should not be regarded as a treatment recommendation.

Pharmacokinetic Interactions and Clinical Management

From a pharmacokinetic perspective, cenobamate has been described as acting primarily as a “perpetrator” of drug–drug interactions rather than being substantially affected by co-medications. Available data indicate that inhibition of CYP2C19 may increase exposure to substrates such as clobazam and certain proton-pump inhibitors, while induction of CYP3A4 and CYP2B6 may reduce plasma levels of drugs including midazolam, bupropion, estrogen–progestin contraceptives, and some mTOR inhibitors. Observational therapeutic drug monitoring data suggest that these enzyme-mediated effects may translate into clinically meaningful consequences, such as sedation or reduced efficacy of non-ASM treatments. Anticipatory dose adjustments, therapeutic drug monitoring where available, and targeted patient counseling—particularly regarding hormonal contraceptives—may support safer use in patients receiving complex polytherapy. Evidence supporting clinically relevant alterations in cenobamate exposure due to concomitant medications remains limited.

Cognition, Psychiatric Outcomes, and Quality of Life

Across observational studies, adjunctive cenobamate was generally not associated with cognitive deterioration, with formal neuropsychological assessments showing overall stability and occasional improvements in processing speed in the context of achieved seizure control and reduced polytherapy. Psychiatric symptoms generally remained stable or improved, with no clear increase in adverse psychiatric events. Quality-of-life outcomes were largely stable over short-term follow-up, while long-term observational data suggested subjective improvements in functioning and well-being in patients achieving seizure reduction and ASM simplification.

Notably, one study reported an opposite signal, with cognitive worsening in some patients, but possibly related to higher cenobamate doses and/or limited reduction in concomitant drug load [86]

Post-surgical/Neuromodulation Failure

Post hoc and real-world data indicated that cenobamate may provide seizure reduction in highly refractory populations, with responder rates often exceeding 40–50% and meaningful seizure freedom achieved in a minority. These observations support its consideration during presurgical evaluation or after unsuccessful interventions, while acknowledging the uncontrolled nature of available studies. In patients with pre-existing psychiatric or neurodevelopmental comorbidities, most cohorts reported stability or improvement, with worsening largely confined to individuals with intellectual disability and behavioral dysregulation, underscoring the need for careful monitoring.

Limitations and Future Directions

This synthesis is based predominantly on indirect comparisons, post hoc analyses, and observational cohorts. Network meta-analyses relied on assumptions of trial similarity and consistency that may not fully hold across heterogeneous RCTs, while real-world studies are subject to confounding by indication, variable titration schedules, and non-standardized outcome assessment. Data on long-term safety, psychiatric outcomes, cognition, and non-ASM interactions remain relatively sparse, and many signals—both favorable and unfavorable—are derived from small samples or exploratory analyses. Future research should ideally include comparative pragmatic trials or well-designed prospective cohorts that directly evaluate cenobamate versus specific third-generation ASMs in defined treatment lines, with standardized assessment of seizure outcomes, AEs, cognitive and psychiatric status, and patient-reported measures. Dedicated studies in older adults, patients with comorbidities, and those receiving complex comedication regimens would be particularly informative.

Conclusion

Cenobamate may represent a valuable therapeutic option for adults with uncontrolled focal epilepsy, including when introduced early as a third add-on therapy. Indirect comparisons and real-world cohorts suggested that earlier use may be associated with a greater probability of seizure freedom and higher retention than other ASMs, although these observations remained hampered by non-randomized designs. Tolerability appeared broadly comparable to that of other third-generation ASMs at standard maintenance doses, with combination-dependent AEs that may be mitigated through proactive adjustment of concomitant therapies. Cenobamate may also facilitate meaningful reductions in overall drug load, potentially contributing to improved long-term tolerability and patient-reported outcomes. Pharmacokinetic interactions involving CYP2C19 inhibition and CYP3A4/2B6 induction require attention and individualized management, especially in patients receiving ASMs such as clobazam, phenobarbital, and phenytoin, hormonal contraceptives, direct oral anticoagulants, or other sensitive substrates. Current data suggest that cenobamate does not generally worsen cognitive or psychiatric profiles, and may support functional improvement in subgroups, although systematic evidence remains limited. Its potential utility in patients who have failed surgery or neuromodulation requires further confirmation.

Overall, while the evidence base is still evolving and primarily observational, cenobamate may offer a favorable balance of effectiveness and tolerability when appropriately titrated and integrated into optimized polytherapy regimens. Further comparative and long-term studies are needed to refine its place within the treatment pathway for focal epilepsy. Current evidence suggests that cenobamate can be integrated into complex treatment regimens, including in patients with prior surgery or neuromodulation.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contribution

Simona Lattanzi, Giancarlo Di Gennaro, Carlo Andrea Galimberti, Alfonso Iudice, Carlo Di Bonaventura contributed to the conception and design of the work and to the definition of the clinical questions; participated in the expert appraisal and interpretation of the evidence; critically revised the manuscript for important intellectual content; approved the final version submitted; and agree to be accountable for all aspects of the work. All authors meet the authorship criteria of the International Committee of Medical Journal Editors (ICMJE).

Funding

This publication was supported by an unrestricted grant from Angelini Pharma SpA. Angelini Pharma SpA funded the development of this article and the journal’s Rapid Service Fee. The sponsor had no role in the selection, analysis, or interpretation of the evidence, or in the decision to submit the article for publication.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declarations

Conflict of Interest

The authors declare the following conflicts of interest. Giancarlo Di Gennaro has received speaker’s honoraria from UCB Pharma, Eisai, Lusofarmaco, LivaNova, and Angelini Pharma, and has served on advisory boards for Angelini Pharma, LivaNova, and UCB Pharma. Carlo Di Bonaventura has received speaker’s or consultancy fees from Eisai, UCB Pharma, Jazz Pharmaceuticals, Angelini Pharma, and Lusofarmaco, and has participated in advisory boards for Eisai, UCB Pharma, Jazz Pharmaceuticals, BIAL, and Angelini Pharma. Carlo Andrea Galimberti and Alfonso Iudice declare no conflicts of interest related to this work. Simona Lattanzi has received speaker's or consultancy fees from Angelini Pharma, Eisai, GW Pharmaceuticals, Hikma Pharmaceuticals, NewBridge Pharmaceuticals, and UCB Pharma and has served on advisory boards for Angelini Pharma, Arvelle Therapeutics, BIAL, Eisai, GW Pharmaceuticals, Hikma Pharmaceuticals, Rapport Therapeutics, and UCB Pharma outside the submitted work. Simona Lattanzi is an Editorial Board member of Neurology and Therapy. Simona Lattanzi was not involved in the selection of peer reviewers for the manuscript nor any of the subsequent editorial decisions.

Ethical Approval

This article is based exclusively on previously published studies and does not report any new study involving human participants or animals performed by any of the authors. Ethics committee approval and informed consent were therefore not required.