New-Generation Antiseizure Medications: Who Takes the Lead in the Real World?

Abstract

Comparative Effectiveness of Brivaracetam, Cenobamate, Lacosamide, and Perampanel in Focal Epilepsy

Cerulli Irelli E, Roberti R, Sole Borioni M, Anzellotti F, Belcastro V, Beretta S, Boero G, Bonanni P, Chiesa V, D’Aniello A, Dainese F, Deleo F, DeMaria G, DiGennaro G, DiGennaro G, Didato G, Dono F, Falcicchio G, Ferlazzo E, Fortunato F, LaNeve A, Mecarelli O, Montalenti E, Morano A, Nilo A, Operto F, Paladin F, Pascarella A, Pauletto G, Pietrafusa N, Pignatta P, Pulitano P, Renna R, Rosati E, Sammarra I, DiBonaventura C, Russo E, Lattanzi S; for the Comparative REal-World Evidence (CREW) Study Group. JAMA Neurol. 2026 Apr 1;83(4):320–328. PMID: 41661607 PMCID: PMC12887838 DOI: 10.1001/jamaneurol.2025.5625

Importance: Treatment decisions in drug-resistant focal epilepsy remain largely empirical, as direct comparative evidence among newer antiseizure medications (ASMs) is limited. Real-world data can complement randomized clinical trials by providing insights into long-term effectiveness and safety across diverse populations. Objective: To compare effectiveness and safety of brivaracetam, cenobamate, lacosamide, and perampanel as adjunctive therapies in adults with drug-resistant focal epilepsy. Design, setting, and participants: This was a multicenter pooled analysis of 4 previously conducted retrospective real-world medical record-review studies (January 2017 to January 2024). Included were adult patients (aged ≥16 years) with drug-resistant focal epilepsy, as defined by the International League Against Epilepsy. Participants were recruited from 71 epilepsy centers. Exposures: Add-on treatment with brivaracetam, cenobamate, lacosamide, or perampanel. Main outcomes and measures: The primary outcome was the responder rate at 6 months, defined as greater than or equal to 50% seizure frequency reduction from baseline. Secondary outcomes included 12-month responder rate, seizure freedom (≥3 months at 6 months and ≥6 months at 12 months), and 12-month ASM retention. Safety was assessed by incidence of adverse effects. Generalized linear mixed models adjusted for demographic and clinical covariates were used to compare treatment outcomes, with cenobamate as reference ASM. Results: Of 2386 ASM prescriptions screened, 1993 prescriptions from 1949 patients (1036 of 1947 female [53.2%]; sex information was missing in 0.1% of prescriptions) with a median (IQR) age of 42 (29-55) years at ASM prescription, met inclusion criteria and were included in the pooled analysis. Brivaracetam accounted for 953 prescriptions (47.8%), followed by perampanel (607 [30.5%]), lacosamide (241 [12.1%]), and cenobamate (192 [9.6%]). After adjustment, cenobamate demonstrated significantly higher odds of 50% or greater response at 6 months compared with brivaracetam (odds ratio [OR]: 0.18; 95% CI: 0.12-0.28; P < .001), perampanel (OR: 0.26; 95% CI: 0.16-0.42; P < .001), and lacosamide (OR: 0.29; 95% CI: 0.17-0.49; P < .001). Results were consistent for secondary effectiveness outcomes at 12 months, with cenobamate outperforming other ASMs in terms of 50% or greater response and seizure freedom. Cenobamate was associated with the highest rate of adverse effects during follow-up (111 [57.8%]), and lacosamide was associated with the lowest (35 [14.8%]). Cenobamate was associated with a higher likelihood of treatment retention at 12 months compared with brivaracetam (OR: 0.43; 95% CI: 0.26-0.69; P < .001) and perampanel (OR: 0.56; 95% CI: 0.32-0.99; P = .047), with no significant difference versus lacosamide (OR: 0.81; 95% CI: 0.41-1.59; P = .53). Conclusions and relevance: These study findings suggest superior effectiveness of cenobamate over brivaracetam, lacosamide, and perampanel in adults with drug-resistant focal epilepsy in a large real-world setting.

Commentary

Focal epilepsy is a prevalent disorder that remains difficult to treat. While about two-thirds of patients with newly diagnosed epilepsy achieve seizure control with medications, the remaining one-third develop drug-resistant epilepsy and may require additional antiseizure medications (ASMs) or epilepsy surgery. ¹ Selecting adjunctive medications is challenging because efficacy must be balanced with tolerability and cost, and adverse effects are common during treatment. As a result, many patients are hesitant to try newer medications after previous treatment failures or intolerable side effects. Clinicians face the difficult task of selecting the optimal agent, often in the absence of strong comparative effectiveness data and with limited evidence regarding long-term tolerability.

In the last decade, several new-generation ASMs have been introduced and widely adopted by the epilepsy community. Brivaracetam was approved in Europe in 2016 after having been designated an orphan drug for more than a decade (Table 1). It selectively binds to synaptic vesicle protein 2A (SV2A). Compared with levetiracetam, brivaracetam has a much higher affinity for SV2A, more rapid brain penetration, and fewer psychiatric adverse effects.^2,3 Introduced in 2019, cenobamate has a dual mechanism of action: it enhances voltage-gated sodium channel inactivation and modulates the function of gamma-aminobutyric acid type A (GABA_A) by augmenting its currents (Table 1). ⁴ This dual mechanism of action likely contributes to the remarkable efficacy of cenobamate in drug-resistant focal epilepsy. ⁴ Lacosamide received regulatory approval in 2008 for the treatment of partial-onset seizures in adults. It has a distinct mechanism of action, selectively enhancing the slow inactivation of voltage-gated sodium channels and thereby stabilizing the channels without altering their other physiological properties (Table 1).^5,6 Perampanel is a selective α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-type glutamate receptor antagonist (Table 1). It has very limited activity at other glutamate receptor subtypes, which is thought to help minimize off-target effects. ⁷ It was first approved in the USA and Europe in 2012 as an adjunctive therapy for focal seizures and has been used in adults and children. ⁷ Although clinical trials have provided data on the effectiveness of these agents (Table 1), head-to-head comparisons of their effectiveness outside formal trials remain scarce.

Table 1.

Pharmacological and Clinical Properties of Four New-Generation Antiseizure Medications.

Anti-Seizure Medication	Regulatory Approval Mechanism of action	Efficacy in Clinical Trials	Most Common Adverse Effects
Brivaracetam	2016 Selective SV2A ligand	50% responder rates: 22–56% Seizure-free rates: 1–5%	Somnolence, dizziness, fatigue, irritability
Cenobamate	2019 (1) Selective inhibition of persistent sodium currents (INaP) (2) Positive allosteric modulation of GABAAR	50% responder rates: 40–82% Seizure-free rates: 3–14%	Somnolence, dizziness, diplopia, fatigue, balance difficulties
Lacosamide	2008 Selective enhancement of slow inactivation voltage-gated sodium channels	50% responder rates: 30–53% Seizure-free rates: 4–9%	Dizziness, diplopia, headache, nausea, vomiting
Perampanel	2012 Selective noncompetitive AMPAR antagonist	50% responder rates: 29–36% Seizure-free rates: 3–7%	Dizziness, somnolence, fatigue, aggression, irritability, homicidal ideations

Abbreviations: SV2A: synaptic vesicle protein 2A; GABA_AR: gamma aminobutyric acid receptor type A; AMPAR: α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor.

In this multicenter retrospective study, Cerulli and coauthors analyzed pooled data on the use of brivaracetam, cenobamate, lacosamide, and perampanel from more than 70 epilepsy centers over a 7-year period. ⁸ They provided real-world estimates of treatment outcomes by comparing the effectiveness and side effects of these four medications in over 1900 patients, using cenobamate as the reference category. In the combined dataset, nearly half of the patients received brivaracetam (47.8%), while the other three agents were prescribed less commonly (perampanel, 30.5%; lacosamide, 12.1%; and cenobamate, 9.6%). Notably, cenobamate was administered to the most severely affected patients, who had the highest baseline seizure burden and greatest ASM utilization. The authors found that, after the generalized linear mixed model (GLMM) adjustment, cenobamate was associated with the highest responder rates and a greater likelihood of remaining seizure-free at the 6-month follow-up compared with the other studied agents. The effectiveness of cenobamate was also demonstrated over a longer follow-up period (i.e., 12 months), when it was similarly associated with a higher responder rate and a higher probability of maintaining seizure freedom.

The GLMM model applied to the analysis of adverse effects revealed that patients taking cenobamate were more likely to continue the medication at 12 months compared with those taking brivaracetam and perampanel, although a similar retention was also noted with lacosamide. Remarkably, cenobamate caused significantly more side effects, including dizziness and somnolence, compared with the other medications, but these effects nonetheless appeared to be tolerable, as they did not affect retention. Interestingly, the unadjusted responder rates for cenobamate were comparable to those of the other agents. This divergence supports the value of accounting for confounding factors when using observational data. The findings of remarkable effectiveness of cenobamate were consistent with previous findings from meta-analyses of randomized controlled trials, which showed greater effectiveness of cenobamate compared with other agents.^4,9 Among the other three agents, brivaracetam was the least effective, although patients who received it had more severe forms of epilepsy compared with those receiving lacosamide and perampanel. Lacosamide was associated with fewer side effects, while brivaracetam and perampanel were positioned intermediate in terms of tolerability and had more behavioral side effects than lacosamide or cenobamate.

The strengths of the study include its large-scale, real-world dataset and rigorous methodology. The data harmonization was a particularly valuable aspect of the approach given that the investigators relied on data obtained during medical chart review. Harmonization was achieved by adopting standardized categorical seizure-frequency variables and aligning other parameters that were reported inconsistently in the reference datasets. Another strength was the statistical approach using a GLMM, which allows the combination of fixed effects reflecting overall trends and random effects related to grouping or nesting. Limitations include the retrospective design and the inability to fully control for the accuracy of seizure and adverse-event reporting. Furthermore, cenobamate may have been prescribed to patients who were more severely affected, although this would, if anything, strengthen the evidence for its effectiveness. Overall, the findings of this study provide a valuable resource for clinicians in their search for more effective medications for difficult-to-treat focal epilepsy.