Abstract
Objective
To assess the 12‐month clinical impact of cenobamate (CNB) in adults with drug‐resistant focal epilepsy in clinical practice, exploring whether outcomes differed according to prior antiseizure medication (ASM) exposure.
Methods
This single‐center, retrospective, observational, real‐world study included 91 adults stratified by the number of prior ASMs: ≤ 6 prior ASMs (n = 44) and > 6 prior ASMs (n = 47). Outcomes assessed at 3, 6, and 12 months included monthly seizure frequency, seizure freedom, retention rate, CNB dose, concomitant ASM burden, and adverse drug reactions (ADRs).
Results
In the overall cohort, mean monthly seizure frequency decreased from 28.77 at baseline to 22.73 (3 months), 19.37 (6 months), and 14.58 (12 months), representing mean reductions of 21%, 33%, and 49%, respectively. Seizure freedom was achieved by 12% (11/91) of patients at 3 months, 13% (11/87) at 6 months, and 19% (16/82) at 12 months. Treatment retention rate was 100% at 3 months, 95.6% at 6 months, and 90.1% at 12 months. Both subgroups showed clinical benefit, but outcomes were more favorable in patients with ≤ 6 prior ASMs, who had a lower baseline seizure burden and higher 12‐month seizure freedom than those with > 6 prior ASMs (29% vs. 10%). At 12 months, most patients were receiving CNB doses of 200 mg or higher, and 79% were maintained on ≤ 2 concomitant ASMs. Overall, 38.5% (35/91) of patients reported at least one ADR (most frequently somnolence, vertigo, and asthenia).
Significance
In this real‐world cohort of adults with drug‐resistant focal epilepsy, CNB was associated with sustained seizure reduction, high 12‐month retention rate, and acceptable tolerability, while most patients were maintained on a low concomitant ASM burden during follow‐up. Improvement was observed even in highly treatment‐experienced patients, although outcomes were more favorable in those with lower prior ASM burden, supporting the hypothesis that earlier CNB positioning may provide greater efficacy.
Keywords: cenobamate, low‐dose efficacy, polytherapy de‐escalation, refractory epilepsy, retention, ultra‐refractory epilepsy
Plain Language Summary
This 12‐month study followed 91 adults whose focal seizures were not controlled by multiple medications. Adding cenobamate reduced average monthly seizures by about half, and 19% of patients became seizure‐free after 1 year. Nine out of ten patients continued treatment throughout follow‐up. Cenobamate was also often associated with simplification of background therapy, with most patients taking no more than two other antiseizure medications at 12 months. Benefits were seen even in very difficult‐to‐treat cases, although outcomes were more favorable when cenobamate was introduced earlier in the treatment course.
Key points.
Cenobamate was associated with 90.1% retention at 12 months and a 49% mean reduction in monthly seizure frequency in 91 adults with drug‐resistant focal epilepsy.
Seizure freedom was achieved in 19% of patients, with clinically meaningful benefit also observed in highly treatment‐experienced cases.
Low concomitant ASM burden was maintained during follow‐up, with 79% of patients receiving no more than two concomitant ASMs at 12 months.
The safety profile was consistent with previous clinical trials and real‐world studies, with no cases of DRESS syndrome.
More favorable outcomes were observed in patients with lower prior ASM burden, supporting the hypothesis that earlier CNB positioning may provide greater clinical benefit.
1. INTRODUCTION
Despite the availability of numerous anti‐seizure medications (ASMs), approximately one‐third of people with epilepsy continue to experience refractory seizures, highlighting a significant unmet medical need. 1 , 2 This persistent drug resistance necessitates the exploration of novel therapeutic agents with distinct mechanisms of action. Cenobamate (CNB), a recently approved ASM for focal seizures in adults, represents a significant advancement in this landscape. While its precise mechanism is not fully elucidated, CNB is believed to selectively inhibit persistent sodium currents and positively modulate the GABAA receptor, thereby reducing neuronal excitability. 3 This unique dual mechanism offers a promising avenue for patients who have failed to achieve adequate seizure control with conventional ASMs. While randomized controlled trials are the gold standard for establishing the efficacy of a new treatment under ideal conditions, real‐world studies are crucial for understanding drug effectiveness and safety in the general patient population and within the typical conditions of clinical practice. 4 Such studies provide a more comprehensive and pragmatic picture of a drug's utility, which can better inform clinical decision‐making.
In recent years, real‐world evidence on CNB has expanded substantially, including multicenter cohorts, early access programs, and observational studies with medium to long‐term follow‐up. 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 These studies have consistently confirmed meaningful seizure reduction, seizure freedom in a subset of patients, acceptable tolerability, and favorable retention rates in routine clinical practice. Accordingly, the main clinical question is no longer whether CNB has real‐world activity, but rather how its introduction may influence the therapeutic trajectory of patients with drug‐resistant epilepsy. In this context, two issues are especially relevant. First, CNB may affect not only seizure outcomes, but also the overall treatment burden, including the possibility of simplifying concomitant ASM regimens. This is particularly important in patients exposed to chronic polytherapy, in whom treatment success is often measured not only by seizure control, but also by improved manageability and tolerability of long‐term therapy. Second, it remains clinically relevant to understand whether outcomes differ according to prior treatment burden, since the degree of refractoriness and the timing of CNB introduction may influence the magnitude of benefit observed in practice. The present monocentric retrospective study was therefore designed to evaluate the 12‐month clinical impact of CNB in a real‐world cohort of adults with drug‐resistant focal epilepsy, with particular attention to three dimensions: sustained seizure outcomes, treatment retention, and changes in concomitant ASM burden. In addition, we compared outcomes between patients with ≤ 6 and > 6 previously tried ASMs, as a pragmatic marker of treatment‐line positioning and degree of refractoriness. By doing so, we aimed to provide clinically useful information not only on whether CNB works in routine care, but also on how it may fit within ongoing treatment strategies in refractory and ultra‐refractory focal epilepsy.
2. MATERIALS AND METHODS
2.1. Study design and participants
This was a single‐center retrospective observational cohort study conducted in routine clinical practice at the Neurology Unit–Epilepsy Center of Fondazione IRCCS Casa Sollievo della Sofferenza, San Giovanni Rotondo (FG), Italy. The study analyzed the real‐world clinical impact of CNB in a cohort of 91 individuals with focal drug‐resistant epilepsy, with particular attention to seizure outcomes, treatment retention, tolerability, and changes in concomitant ASM burden. In accordance with the International League Against Epilepsy (ILAE) definition of drug‐resistant epilepsy, 23 all participants in this study met the criteria for drug resistance. Patients were further stratified into two subgroups based on their history of ASM use 24 prior to CNB initiation:
The “drug‐resistant” group (n = 44) included patients who had tried 6 or fewer previous ASMs.
The “ultra‐refractory” group (n = 47) comprised patients who had tried more than 6 previous ASMs.
All participants had previously tried a significant number of ASMs, with a median of 7 ASMs (range 1–21), indicating a highly refractory epilepsy population.
2.2. Data collection
Baseline demographic and clinical data were collected for all participants prior to CNB initiation. This included age, sex, age at epilepsy diagnosis, and total number of ASMs previously tried. Additionally, the number of concomitant ASMs in use before CNB initiation was recorded. Baseline seizure frequency for efficacy analyses was defined as the total number of focal seizures and focal‐to‐bilateral tonic–clonic seizures recorded during the month preceding CNB initiation. The total number of seizures in the three months preceding CNB initiation was also recorded for descriptive purposes. Follow‐up data were collected at 3, 6, and 12 months post‐CNB initiation. Safety and tolerability data, including adverse drug reactions (ADRs), were assessed at month 2 as well as at months 3, 6, and 12. Key parameters recorded during follow‐up included:
The number and percentage of patients at different CNB doses (categorized as < 50, 100, 150, 200, and ≥ 250 mg).
The number and percentage of patients with varying numbers of concomitant ASMs (0, 1, 2, 3, 4, ≥ 5).
The number of patients who discontinued treatment at each follow‐up time point and the specific reason for discontinuation (e.g., lack of efficacy, ADRs, patient choice, death).
The occurrence, type, and severity of ADRs reported, and their perceived relationship to CNB therapy.
Any modifications (e.g., dosage change, withdrawal) to concomitant ASMs.
The study flow, subgroup allocation, and patient retention across follow‐up are summarized in Figure 1.
FIGURE 1.
Study flow and patient disposition according to prior antiseizure medication exposure. Patients were stratified into 2 groups (≤ 6 and > 6 previous ASMs). The figure reports the number of patients on treatment, discontinuations, and death at 3, 6, and 12 months after cenobamate initiation.
2.3. Outcome assessment
Seizure frequency reduction was the primary efficacy outcome. Baseline seizure frequency was defined as the number of seizures recorded during the month preceding CNB initiation, and follow‐up seizure frequency was assessed at 3, 6, and 12 months after the start of therapy. The percentage reduction in seizure frequency for each patient was calculated using the formula:
Seizure‐free rates (100% reduction) and responder rates (≥ 50%, ≥ 75%, and ≥ 90% seizure reduction) were also calculated at each follow‐up time point. In addition to seizure outcomes, changes in concomitant ASM burden during follow‐up were evaluated descriptively as an indicator of treatment simplification.
2.4. Statistical analysis
Continuous variables were presented as median (IQR, interquartile range) or mean (SD, standard deviation) and categorical variables as counts and percentages.
For baseline comparisons between the “drug‐resistant” and “ultra‐refractory” subgroups:
Continuous variables were compared using the Mann–Whitney U test.
Categorical variables were compared using the Chi‐squared test or Fisher's exact test.
For efficacy outcomes during the follow‐up period:
The change in median monthly seizure frequency from baseline to each follow‐up interval was analyzed using the Wilcoxon Signed‐Rank Test for paired data.
Between‐group differences in percentage seizure reduction between the “drug‐resistant” and “ultra‐refractory” subgroups were assessed using the two‐sided Mann–Whitney U test. Missing or non‐calculable values, mainly due to baseline seizure frequency equal to zero, were excluded from the corresponding analysis.
Changes in the number of concomitant ASMs across 3, 6, and 12 months were assessed using the Friedman test in patients with complete data at all three timepoints.
Retention rates were calculated using Kaplan–Meier survival analysis.
All tests were two‐tailed, and a p‐value of < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS.
2.5. Ethics approval
The study protocol was approved by the local Ethics Committee (protocol no. 7910/CE/2025). All procedures were conducted in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
3. RESULTS
3.1. Baseline features and subgroup analysis
The study population comprised 91 patients with focal drug‐resistant epilepsy initiating CNB therapy. The median age was 37 years (IQR: 27–50.3), with 55% of patients being female. The median age at epilepsy diagnosis was 11 years (IQR: 5–22). Overall, patients exhibited high resistance, having failed a mean of 7.42 ASMs (range: 1–21 drugs), and were on a mean of 2.27 concomitant ASMs (range: 1–5) at baseline. Pre‐CNB seizure burden was substantial, with a mean of 28.77 seizures in the month before CNB initiation (85.87 seizures in the 3 months before CNB). Stratification into ≤ 6 previous ASM (N = 44) and > 6 previous ASM (N = 47) subgroups revealed key clinical differences. The > 6 ASM group confirmed absolute pharmacoresistance (mean failed ASMs: 10.64 vs. 3.98), displayed a significantly earlier median age at diagnosis (8 years vs. 16.5 years), and presented with a fourfold higher mean baseline monthly seizure frequency (35.19 vs. 21.91 seizures/month), collectively indicating a greater overall disease severity (Table 1).
TABLE 1.
Baseline characteristics of the study population (N = 91).
| Total (n = 91) | ≤ 6 previous ASMs (n = 44) | > 6 previous ASMs (n = 47) | |
|---|---|---|---|
| Age | |||
| median (IQR) | 37 (27–50.3) | 37.5 (25–50.25) | 37 (30.5–50.5) |
| Sex | |||
| Male, n (%) | 41 (45%) | 20 (45%) | 21 (45%) |
| Female, n (%) | 50 (55%) | 24 (55%) | 26 (55%) |
|
Age at diagnosis of epilepsy median (IQR) |
11 (5–22) | 16.5 (9–30) | 8 (1–11.5) |
| Drugs tried | |||
| mean (SD) | 7.42 (±4.31) | 3.98 (±1.49) | 10.64 (±3.52) |
| min–max | (1–21) | (1–6) | (7–21) |
| N. of drugs before introduction of CNB | |||
| mean (SD) | 2.27 (±0.96) | 2.23 (±0.86) | 2.32 (±1.04) |
| min–max | (1–5) | (1–5) | (1–5) |
| N. of seizures in the month before CNB | |||
| mean (SD) | 28.77 (±40.07) | 21.91 (±42.93) | 35.19 (±36.50) |
| median (IQR) | 10 (4–40) | 7.5 (3.75–19.5) | 30 (7.5–60) |
| min–max | (0–252) | (0–252) | (2–180) |
| N. of seizures in the 3 months before CNB | |||
| mean (SD) | 85.87 (±115.60) | 66.34 (±128.26) | 104.15 (±100.30) |
| median (IQR) | 30 (12.5–120) | 21.5 (11.5–58.5) | 90 (20.5–180) |
| min–max | (2–756) | (2–756) | (6–420) |
Note: Data are presented for the overall cohort and for the two subgroups stratified by the number of previously tried ASMs (≤ 6 vs. > 6). Continuous variables are expressed as median (IQR) or mean (SD); categorical variables as counts and percentages.
3.2. Follow‐up outcomes
Regarding the retention rate, CNB treatment retention was 95.6% at 6 months (87/91 patients) and 90.1% at 12 months (82/91 patients), with 9 discontinuations by 12 months (8 treatment discontinuations and 1 death). The death occurred in a 57‐year‐old patient with multiple severe comorbidities due to intercurrent bronchopneumonia and was assessed as unrelated to CNB treatment (Table 2). The proportion of patients on higher CNB doses increased over time. At 3 months, 46% were on 200 mg; this increased to 66% at 6 months and 60% at 12 months. Doses ≥ 250 mg were initiated in 3% of patients at 6 months and 12% at 12 months (Table 3). The median CNB dose in the overall cohort was 150 mg/day (IQR 100–200) at 3 months, 200 mg/day (IQR 150–200) at 6 months, and 200 mg/day (IQR 150–200) at 12 months. In the ≤ 6 previous ASM subgroup, the median CNB dose was 150 mg/day (IQR 100–200) at 3 months and 200 mg/day at both 6 months (IQR 150–200) and 12 months (IQR 150–200). In the > 6 previous ASM subgroup, the median CNB dose was 150 mg/day (IQR 100–200) at 3 months, 200 mg/day (IQR 187.5–200) at 6 months, and 200 mg/day (IQR 150–200) at 12 months. Mean monthly seizure frequency decreased from 28.77 seizures/month at baseline to 22.73 at 3 months, 19.37 at 6 months, and 14.58 at 12 months, corresponding to mean reductions of 21%, 33%, and 49%, respectively (Table 2). Seizure freedom was achieved by 11/91 patients (12%) at 3 months, 11/87 patients (13%) at 6 months, and 16/82 patients (19%) at 12 months in the observed‐case/on‐treatment analysis (Table 2). As a conservative sensitivity description using the full cohort as denominator, seizure freedom rates were 11/91 patients (12.1%) at 3 months, 11/91 patients (12.1%) at 6 months, and 16/91 patients (17.6%) at 12 months. The percentage of ≥50%, 75% and 90% responders increased from 3 months to 12 months, as shown in Figure 2.
TABLE 2.
Seizure outcomes and treatment retention during follow‐up in the overall cohort and stratified by prior ASM exposure (≤ 6 vs. > 6 previous ASMs).
| Timepoint | Group | Seizure outcomes | Treatment retention | ||||
|---|---|---|---|---|---|---|---|
| Seizure‐free, n (%) | Mean seizure frequency | Reduction (%) | On treatment with CNB | Discontinued | Retention (%) | ||
| Baseline | Total | 0 (0%) | 28.77 | – | – | – | – |
| ≤ 6 previous ASMs | 0 (0%) | 21.91 | – | – | – | – | |
| > 6 previous ASMs | 0 (0%) | 35.19 | – | – | – | – | |
| 3 month | Total | 11 (12%) | 22.73 | 21% | 91 | 0 | 100% |
| ≤ 6 previous ASMs | 8 (18%) | 15.73 | 28% | 44 | 0 | 100% | |
| > 6 previous ASMs | 3 (6%) | 29.30 | 17% | 47 | 0 | 100% | |
| 6 month | Total | 11 (13%) | 19.37 | 33% | 87 | 4 | 95.6% |
| ≤ 6 previous ASMs | 8 (19%) | 10.83 | 51% | 43 | 1 | 97.7% | |
| > 6 previous ASMs | 3 (7%) | 27.33 | 22% | 44 | 3 | 93.6% | |
| 12 month | Total | 16 (19%) | 14.58 | 49% | 82 | 8 | 90.1% |
| ≤ 6 previous ASMs | 12 (29%) | 9.12 | 58% | 42 | 2 | 95.4% | |
| > 6 previous ASMs | 4 (10%) | 20.17 | 43% | 40 | 6 | 85.1% | |
Note: Seizure outcomes are reported as seizure‐free patients, mean seizure frequency, and percentage reduction from baseline. Efficacy outcomes were analyzed using an observed‐case/on‐treatment approach: At each follow‐up timepoint, seizure‐free percentages were calculated using the number of patients still receiving CNB as denominator. Treatment retention is reported as the number of patients on treatment with CNB, discontinuations, and retention rate at each timepoint. Reduction (%) was calculated relative to baseline seizure frequency. Retention (%) was calculated as the number of patients on treatment with CNB divided by the total number of patients at baseline × 100. Discontinued includes all causes of treatment discontinuation (lack of efficacy, adverse events, patient choice, and death [n = 1, in the > 6 ASMs group]).
Abbreviations: ASMs, antiseizure medications; CNB, cenobamate.
TABLE 3.
Number and percentage of patients at different CNB doses at 3, 6, and 12 months for the overall cohort and for the two subgroups stratified by prior ASM exposure (≤ 6 vs. > 6 previous ASMs).
| Number of patients at different CNB doses, n (%) | ||||||
|---|---|---|---|---|---|---|
| Dose CNB (mg) | ||||||
| Timepoint | Patients | ≤ 50 | 100 | 150 | 200 | ≥ 250 |
| 3 month | Total | 4 (10%) | 38 (42%) | 7 (8%) | 42 (46%) | – |
| ≤ 6 previous ASMs | – | 19 (43%) | 4 (9%) | 21 (48%) | – | |
| > 6 previous ASMs | 4 (9%) | 19 (40%) | 3 (6%) | 21 (45%) | – | |
| 6 month | Total | 3 (3%) | 16 (18%) | 8 (9%) | 57 (66%) | 3 (3%) |
| ≤ 6 previous ASMs | – | 10 (23%) | 6 (14%) | 27 (61%) | – | |
| > 6 previous ASMs | 3 (6%) | 6 (15%) | 2 (4%) | 30 (64%) | 3 (6%) | |
| 12 month | Total | 2 (2%) | 12 (15%) | 9 (11%) | 49 (60%) | 10 (12%) |
| ≤ 6 previous ASMs | – | 7 (16%) | 5 (11%) | 26 (59%) | 4 (4%) | |
| > 6 previous ASMs | 2 (4%) | 5 (11%) | 4 (9%) | 23 (58%) | 6 (13%) | |
Note: Percentages are calculated based on the number of patients on treatment at each timepoint. Dash (–) indicates no patients at that dose.
Abbreviations: ASMs, antiseizure medications; CNB, cenobamate.
FIGURE 2.
Percentage of responders (≥ 50%, ≥ 75%, ≥ 90% seizure reduction) at 3, 6, and 12 months in the overall cohort (N = 91).
Concomitant ASM burden remained low during follow‐up. At 12 months, most patients were maintained on one or two concomitant ASMs (33% on 1 ASM and 46% on 2 ASMs), indicating that CNB treatment was frequently managed within a simplified background regimen (Table 4).
TABLE 4.
Number and percentage of patients with concomitant ASMs at 3, 6, and 12 months for the overall cohort and for the two subgroups stratified by prior ASM exposure (≤ 6 vs. > 6 previous ASMs).
| Number of patients with concomitant ASMs, n (%) | |||||||
|---|---|---|---|---|---|---|---|
| N. Concomitant ASMs | |||||||
| Timepoint | Patients | 0 | 1 | 2 | 3 | 4 | 5 |
| 3 month | Total | – | 27 (30%) | 43 (47%) | 14 (15%) | 5 (5%) | 2 (2%) |
| ≤ 6 previous ASMs | – | 14 (32%) | 24 (55%) | 4 (9%) | 2 (5%) | – | |
| > 6 previous ASMs | – | 13 (28%) | 19 (48%) | 10 (21%) | 4 (9%) | 2 (4%) | |
| 6 month | Total | – | 26 (30%) | 45 (52%) | 10 (11%) | 5 (6%) | 1 (1%) |
| ≤ 6 previous ASMs | – | 15 (34%) | 23 (52%) | 4 (9%) | 2 (5%) | – | |
| > 6 previous ASMs | – | 12 (26%) | 22 (47%) | 6 (13%) | 3 (6%) | 1 (2%) | |
| 12 month | Total | 1 (1%) | 27 (33%) | 38 (46%) | 14 (17%) | 2 (2%) | 1 (1%) |
| ≤ 6 previous ASMs | – | 16 (36%) | 19 (43%) | 6 (14%) | 1 (2%) | – | |
| > 6 previous ASMs | 1 (2%) | 10 (21%) | 19 (40%) | 8 (17%) | 1 (2%) | 1 (2%) | |
Note: Percentages are calculated based on the number of patients on treatment at each timepoint. Dash (–) indicates no patients in that category.
Abbreviation: ASMs, antiseizure medications.
We further explored longitudinal changes in the number of concomitant ASMs during follow‐up. Among the 91 included patients, 82 had complete data at 3, 6, and 12 months and were included in a complete‐case longitudinal analysis. The median number of concomitant ASMs remained stable at 2.0 [IQR 1.0–2.0] at all timepoints; however, the Friedman test showed a significant overall change over time (p = 0.0276). From 3 to 12 months, 12/82 patients (14.6%) reduced the number of concomitant ASMs, 67/82 (81.7%) remained unchanged, and 3/82 (3.7%) increased it. Thus, although the overall median remained unchanged, a subset of patients showed treatment simplification during follow‐up. CNB was generally well‐tolerated. Overall, 35/91 patients (38.5%) experienced at least one ADR during follow‐up (months 2, 3, 6, and 12), for a total of 59 ADRs. The most frequent ADRs, as reported in the Supporting Information (Table S1), were somnolence (23.1%), vertigo (18.7%) and asthenia (7.7%). No cases of DRESS syndrome were observed during the 12‐month follow‐up. Overall, 8/91 (8.8%) patients discontinued CNB during the 12‐month follow‐up. In the ≤ 6 previous ASM group, 2/44 (4.5%) patients discontinued treatment: one due to a lack of efficacy associated with adverse events (somnolence and asthenia), and the second due to somnolence. In the > 6 previous ASM group, 6/47 (12.8%) patients discontinued treatment. Among them, one discontinuation was due to increased seizure frequency. The remaining discontinuations were attributed to a lack of efficacy associated with adverse events, including somnolence, vertigo/dizziness, postural instability, and headache. The patient who died during follow‐up had previously reported mild somnolence.
3.3. Follow‐up outcomes in drug‐resistant and ultra‐refractory groups
Both groups (drug‐resistant, N = 44; ultra‐refractory, N = 47) showed good retention, with slightly more discontinuations in the ultra‐refractory group at 12 months (7 vs. 2, including one death in the > 6 ASM subgroup) (Table 2). Regarding CNB dosing, similar trends in dose escalation were observed in both groups (Figure 3, Table 3). In the drug‐resistant group, CNB yielded gradual improvements, with mean monthly seizure frequency reductions at 12 months of 34% with 100 mg, 36% with 150 mg, 72% with 200 mg, and 47% with ≥ 250 mg. In the ultra‐refractory group, mean seizure‐frequency reductions at 12 months were 17% at 100 mg, 55% at 150 mg, 63% at 200 mg, and 22% at ≥ 250 mg. Dose‐stratified seizure‐frequency changes are summarized in Table S3. Overall, both groups experienced a reduction in mean seizure frequency from baseline to 12 months. However, the ultra‐refractory group consistently had a higher absolute mean seizure frequency throughout the treatment period compared to the drug‐resistant group, reflecting their higher baseline severity (Table S2). To further assess whether seizure reduction differed according to prior ASM exposure, we compared individual percentage seizure reduction between the drug‐resistant and ultra‐refractory groups. Median seizure reduction was numerically higher in the drug‐resistant group than in the ultra‐refractory group at all timepoints, but between‐group differences were not statistically significant. At 3 months, median seizure reduction was 42.9% in the drug‐resistant group versus 8.3% in the ultra‐refractory group (p = 0.2460). At 6 months, median seizure reduction was 50.0% versus 14.9%, respectively (p = 0.3005). At 12 months, median seizure reduction was 60.0% versus 20.0%, respectively (p = 0.2366). When response was analyzed according to CNB dose, lower response rates were observed in patients receiving ≥ 250 mg/day in both the refractory and ultra‐refractory groups. To better contextualize this finding, dose‐stratified response data for patients treated with > 200 mg/day have now been reported in detail. Given the limited size of these subgroups and the observational nature of treatment escalation, these comparisons should be interpreted descriptively.
FIGURE 3.
Mean monthly seizure frequency at baseline and at 3, 6, and 12 months in the two subgroups stratified by prior ASM exposure (≤ 6 vs. > 6 previous ASMs).
4. DISCUSSION
This real‐world, monocentric retrospective cohort study provides additional evidence supporting the long‐term effectiveness and tolerability of CNB in adults with drug‐resistant focal epilepsy. Over 12 months of follow‐up, CNB was associated with sustained seizure reduction, a high retention rate, and acceptable tolerability. However, the main clinical relevance of our findings lies not only in confirming effectiveness in routine practice, but also in showing that CNB treatment was commonly managed within a relatively low concomitant ASM burden over time, suggesting a potential impact on the broader therapeutic trajectory of patients with long‐standing refractory epilepsy. The sustained nature of the antiseizure effect is one of the most relevant findings. Mean monthly seizure frequency decreased progressively from baseline to 12 months, with an overall reduction of 49%, while seizure freedom was achieved in 19% of patients at 1 year. These results support the persistence of benefit beyond the early titration phase and are consistent with the growing body of real‐world data showing maintained effectiveness of CNB in clinical practice. 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 25 , 26 At the same time, our results should be interpreted within the context of an increasingly mature literature. Several larger multicenter studies and expanded access analyses have already demonstrated robust real‐world effectiveness and favorable retention of CNB. Therefore, the value of the present study does not lie in sample size or in being among the first 12‐month reports, but rather in its clinically oriented framing of outcomes according to treatment burden and therapeutic simplification.
The 12‐month retention rate of 90.1% is another clinically relevant result. In a cohort with a median of 7 previously failed ASMs, retention at this level suggests that CNB provided an overall balance of efficacy and tolerability considered favorable in routine care. This finding is consistent with previous real‐world studies, which have reported 12‐month retention rates ranging from approximately 70% to more than 90% in similarly refractory populations. 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 Likewise, the seizure‐free rate observed in our cohort is in line with published long‐term real‐world data in highly treatment‐experienced patients. Moreover, retention complements seizure‐frequency outcomes and provides a pragmatic indicator of the real usefulness of CNB in patients with established drug resistance.
Recent real‐world data, including the interim 24‐week results of the BLESS study, 20 have confirmed the effectiveness and safety of adjunctive cenobamate in focal‐onset drug‐resistant epilepsy. Our findings are in line with these data, as we also observed meaningful seizure reduction and good tolerability in routine clinical practice. However, our study extends this evidence by providing a 12‐month follow‐up and by specifically examining outcomes according to prior ASM burden. Notably, although our cohort included a highly treatment‐experienced population, including patients with marked pharmacoresistance, CNB retained clinically relevant benefit and high retention over time. This longitudinal perspective complements the shorter‐term observations from BLESS and supports the durability of CNB effectiveness in a real‐world setting.
A central aspect of the present analysis is that, during CNB follow‐up, most patients were maintained on a limited number of concomitant ASMs. By 12 months, 79% of the cohort remained on no more than two background ASMs. This finding is clinically important because the management of drug‐resistant epilepsy is often constrained by cumulative polytherapy, pharmacokinetic interactions, and reduced tolerability. In this setting, a treatment associated with improved seizure control within a relatively simplified background regimen may represent a meaningful shift in therapeutic strategy rather than merely the addition of another agent. This aspect is particularly relevant in routine practice, where reducing treatment burden may improve long‐term adherence, tolerability, and overall manageability of care.
Subgroup analysis according to prior ASM burden further supports this interpretation. Both patients with ≤ 6 prior ASMs and those with > 6 prior ASMs derived benefit from CNB, confirming clinical activity even in highly treatment‐experienced and markedly pharmacoresistant cases. However, outcomes were more favorable in patients with a lower prior treatment burden, who showed lower baseline seizure frequency, greater seizure reduction, and higher seizure‐free rates at 12 months than the > 6 ASM subgroup. These findings are clinically plausible and likely reflect differences in underlying refractoriness and disease severity. Importantly, they should not be interpreted as evidence of causality, given the retrospective design and the baseline imbalance between groups. Nevertheless, they are consistent with the hypothesis that earlier positioning of CNB in the treatment pathway, before extreme pharmacoresistance becomes established, may provide greater clinical benefit. The dose‐stratified findings should also be interpreted with caution. In both subgroups, clinically meaningful seizure reductions were observed across a range of CNB maintenance doses, and some benefit was seen even at 100 mg/day in selected patients. This may suggest that dose optimization in routine practice should remain individualized rather than exclusively driven by a fixed maintenance target. However, because dose categories were not randomized, subgroup sizes were small, and baseline seizure burden likely differed across dose strata, these observations cannot be interpreted as evidence of dose equivalence or of a direct dose–response comparison. Rather, they support the practical concept that clinically relevant benefit may emerge at different doses depending on patient profile and treatment context. Moreover, the apparently lower response observed at CNB doses ≥ 250 mg/day should be interpreted cautiously. In this real‐world cohort, higher doses were more likely to be used in patients with greater seizure severity and higher treatment refractoriness, introducing confounding by indication. In addition, the number of patients exposed to these dose levels was limited, which reduces the stability of subgroup comparisons. CNB dosing in clinical practice is also individualized according to efficacy and tolerability; therefore, patients achieving meaningful seizure reduction may remain on lower maintenance doses, whereas escalation to higher doses is more often pursued in those with persistent seizures. For these reasons, our data should not be interpreted as showing an inverse dose–response relationship.
Finally, the tolerability profile observed in our study was consistent with previous clinical trials and observational reports. Adverse drug reactions were reported in 38.5% of patients, with somnolence, vertigo, and asthenia being the most common. No cases of DRESS syndrome were observed. Importantly, the high retention rate and the limited number of discontinuations due to tolerability issues suggest that adverse effects were manageable in most cases. As reported in other real‐world studies, active adjustment of concomitant ASMs during CNB titration likely contributed to treatment continuation and may partly explain the treatment simplification observed during follow‐up.
4.1. Limitations
The present study has several limitations. Its retrospective design introduces the possibility of selection and information bias and precludes causal inference. The monocentric setting limits generalizability, and the absence of a control group prevents exclusion of potential confounding factors, including concomitant ASM changes and spontaneous seizure fluctuation over time. In addition, although the cohort size is acceptable for a single‐center real‐world study, it remains limited for detailed subgroup and dose‐stratified analyses. EEG data were not collected prospectively according to a standardized follow‐up schedule, precluding a systematic pre/post‐treatment electroencephalographic evaluation. Likewise, syndrome‐level electroclinical classification was not consistently available, preventing syndrome‐specific analyses. Moreover, patient‐reported outcomes, cognition, and quality‐of‐life measures were not systematically assessed, although these dimensions are highly relevant when treatment simplification is considered a meaningful clinical objective. Finally, the relatively low incidence of adverse effects observed in our cohort should be interpreted with caution. Although all safety data were rigorously recorded, the collection method based on unstructured interviews during follow‐up visits might have underestimated minor or subjective symptoms.
Despite these limitations, our findings support CNB as an effective and generally well‐tolerated long‐term treatment option in adults with drug‐resistant focal epilepsy, including highly refractory cases. More importantly, they suggest that the clinical role of CNB may extend beyond seizure reduction alone, as its introduction may also facilitate simplification of concomitant ASM regimens and thereby modify the therapeutic trajectory of chronic refractory epilepsy. The more favorable outcomes observed in patients with lower prior ASM burden further support the hypothesis that earlier positioning of CNB in the treatment pathway may be associated with greater benefit. This hypothesis should now be addressed in larger prospective multicenter studies specifically designed to evaluate treatment sequencing, de‐escalation strategies, and patient‐centered long‐term outcomes.
5. CONCLUSION
In this real‐world cohort of adults with drug‐resistant focal epilepsy, CNB was associated with sustained seizure reduction, high 12‐month retention, and acceptable tolerability. Beyond these established outcomes, CNB treatment was frequently maintained within a low concomitant ASM burden, suggesting a clinically relevant effect on the overall therapeutic trajectory. Benefit was observed even in highly treatment‐experienced patients, although outcomes were more favorable in those with lower prior ASM burden. These findings support CNB as a relevant long‐term therapeutic option in refractory and ultra‐refractory focal epilepsy and further support the rationale for evaluating earlier CNB positioning in future prospective studies.
AUTHOR CONTRIBUTIONS
Francesca Cutropia: Investigation, writing—original draft. Maria Teresa Di Claudio, Marta Robustella, Umberto Costantino, Carmela Pia Ferro, Giuseppe Miscio, Elena Conte, Nicola Amoroso, Orazio Nicolotti, Cosimo Damiano Altomare: Investigation. Massimo Carella: Supervision. Paola Imbrici, Antonella Liantonio: Investigation, review and editing. Giuseppe d'Orsi: Conceptualization, investigation, review and editing.
CONFLICT OF INTEREST STATEMENT
None of the authors has any conflict of interest to disclose.
ETHICS STATEMENT
We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
ETHICS APPROVAL
The study was approved by the local Ethics Committee (Protocol No. 7910/CE/2025).
CONSENT
Informed consent for publication was obtained from patients or their legally authorized representatives.
Supporting information
Table S1. Summary of adverse drug reactions (ADRs) reported up to 12 months after cenobamate initiation (assessed at months 2, 3, 6, and 12). Data are presented as n (%). Patients may have experienced more than one ADR. ADRs are classified according to MedDRA System Organ Class. No cases of DRESS syndrome were observed. ASMs = antiseizure medications.
The case of alopecia was assessed as unrelated to the drug; in this specific patient, the event was attributed to pre‐existing androgenetic alopecia rather than the pharmacological intervention.
Table S2. Mean monthly seizure frequency at 3, 6, and 12 months across various cenobamate (CNB) doses, stratified by prior ASM exposure (≤ 6 vs. > 6 previous ASMs). Dash (−) indicates no patients at that dose/timepoint combination. ASMs = antiseizure medications; CNB = cenobamate.
Table S3. Mean monthly seizure frequency at baseline and at 3, 6, and 12‐month follow‐up, stratified by CNB dose and prior ASM exposure (≤ 6 vs. > 6 previous ASMs). Percentage reduction was calculated as [(baseline − follow‐up) / baseline] × 100. Negative values indicate an increase in seizure frequency. Dash (−) indicates no patients at that dose/timepoint combination. ASMs = antiseizure medications; CNB = cenobamate.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- 1. Schmitz B, Montouris G, Schäuble B, Caleo S. Assessing the unmet treatment need in partial‐onset epilepsy: looking beyond seizure control. Epilepsia. 2010;51(11):2231–2240. 10.1111/j.1528-1167.2010.02738.x [DOI] [PubMed] [Google Scholar]
- 2. Ioannou P, Foster DL, Sander JW, Dupont S, Gil‐Nagel A, O'Flaherty ED, et al. The burden of epilepsy and unmet need in people with focal seizures. Brain Behav. 2022;12(9):e2589. 10.1002/brb3.2589 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Sankar R, Ferrari L, Kamin M. Exploration of the mechanism of action of cenobamate. Seizure. 2025;134:79–85. 10.1016/j.seizure.2024.11.007 [DOI] [PubMed] [Google Scholar]
- 4. Sherman RE, Anderson SA, Dal Pan GJ, Gray GW, Gross T, Hunter NL, et al. Real‐world evidence—what is it and what can it tell us? N Engl J Med. 2016;375(23):2293–2297. 10.1056/NEJMsb1609216 [DOI] [PubMed] [Google Scholar]
- 5. Czapińska‐Ciepiela E, Kondak A, Kowalczyk M, Czapiński P. Early antiseizure response to cenobamate (200 mg/day) in focal drug‐resistant epilepsy: a retrospective single‐center analysis. Epilepsy Behav Rep. 2025;32:100839. 10.1016/j.ebr.2024.100839 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Plaquevent A, Goff FL, Chastan N. A French real‐world experience with cenobamate in patients with drug‐resistant focal epilepsy: a retrospective observational study. Epilepsy Behav Rep. 2025;31:100782. 10.1016/j.ebr.2024.100782 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Bosak M, Podraza H, Włoch‐Kopeć D, Rysz A, Wężyk K, Grabska‐Radzikowska K, et al. Efficacy and safety of cenobamate: a multicenter, retrospective evaluation of real‐world clinical practice. Seizure. 2025;130:25–31. 10.1016/j.seizure.2025.05.002 [DOI] [PubMed] [Google Scholar]
- 8. Lauxmann S, Heuer D, Heckelmann J, Fischer FP, Schreiber M, Schriewer E, et al. Cenobamate: real‐world data from a retrospective multicenter study. J Neurol. 2024;271(10):6596–6604. 10.1007/s00415-024-12587-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Rodríguez‐Uranga JJ, Sánchez‐Caro JM, Hariramani RR. Treatment simplification to optimize cenobamate effectiveness and tolerability: a real‐world retrospective study in Spain. Epilepsia Open. 2024;9(4):1345–1356. 10.1002/epi4.12948 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Beltrán‐Corbellini Á, Romeral‐Jiménez M, Mayo P, Sánchez‐Miranda Román I, Iruzubieta P, Chico‐García JL, et al. Cenobamate in patients with highly refractory focal epilepsy: a retrospective real‐world study. Seizure. 2023;111:71–77. 10.1016/j.seizure.2023.07.015 [DOI] [PubMed] [Google Scholar]
- 11. Villanueva V, Santos‐Carrasco D, Cabezudo‐García P, Gómez‐Ibáñez A, Garcés M, Serrano‐Castro P, et al. Real‐world safety and effectiveness of cenobamate in patients with focal onset seizures: outcomes from an expanded access program. Epilepsia Open. 2023;8(3):918–929. 10.1002/epi4.12755 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Badr M, Helmstaedter C, Moskau‐Hartmann S, Pukropski J, Witt JA, Rüber T, et al. Cenobamate in real‐world scenario: results on efficacy, side effects, and retention rate in a single center retrospective study. Brain Behav. 2025;15(5):e70567. 10.1002/brb3.70567 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Rheims S, Steinhoff BJ, Hirsch E, Rosenow F, Biraben A, Thomas R, et al. Real‐world effectiveness and tolerability of cenobamate in drug‐resistant epilepsy: a retrospective analysis of the patients included into the early access programs (EAP) in Germany, France, and United Kingdom. Epilepsia Open. 2025;10(3):736–748. 10.1002/epi4.12987 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Strzelczyk A, von Podewils F, Hamer HM, Rosenow F, Knake S, Langner‐Lemercier S, et al. Post‐marketing experience with cenobamate in the treatment of focal epilepsies: a multicentre cohort study. CNS Drugs. 2025;39(3):321–331. 10.1007/s40263-025-01158-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Steinhoff BJ, Georgiou D, Intravooth T. The cenobamate KORK study — a prospective monocenter observational study investigating cenobamate as an adjunctive therapy in refractory epilepsy, with comparisons to historical cohorts treated with add‐on lacosamide, perampanel, and brivaracetam. Epilepsia Open. 2024;9(4):1502–1514. 10.1002/epi4.12963 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Lasek‐Bal A, Kściuk B, Zieliński T, Krzak‐Kubica A, Kowalski J, Żorniak‐Milach B, et al. Long‐term cenobamate retention, efficacy, and safety: outcomes from expanded access Programme. Neurol Neurochir Pol. 2023;57(6):492–496. 10.5603/pjnns.96547 [DOI] [PubMed] [Google Scholar]
- 17. Stern S, Weingarten M, Mandapati S, Ferrari L, Wade CT. Real‐world analysis of retention on cenobamate in patients with epilepsy in the United States. Epilepsy Res. 2023;197:107207. 10.1016/j.eplepsyres.2023.107207 [DOI] [PubMed] [Google Scholar]
- 18. Peña‐Ceballos J, Moloney PB, Munteanu T, Doyle M, Colleran N, Liggan B, et al. Adjunctive cenobamate in highly active and ultra‐refractory focal epilepsy: a “real‐world” retrospective study. Epilepsia. 2023;64(5):1225–1235. 10.1111/epi.17549 [DOI] [PubMed] [Google Scholar]
- 19. Klein P, Aboumatar S, Brandt C, Dong F, Krauss GL, Mizne S, et al. Long‐term efficacy and safety from an open‐label extension of adjunctive cenobamate in patients with uncontrolled focal seizures. Neurology. 2022;99(10):e989–e998. 10.1212/WNL.0000000000200839 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lattanzi S, Dono F, d'Orsi G, D'Aniello A, Panebianco M, Bonanni P, et al. Effectiveness and safety of adjunctive cenobamate in people with focal‐onset epilepsy: interim results after 24‐week observational period from the BLESS study. Epilepsia. 2025;66:2239–2252. 10.1111/epi.18357 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Buhleier E, Schubert‐Bast S, Knake S, von Podewils F, Hamer HM, Melzer N, et al. A multicenter cohort study on the efficacy, retention, and tolerability of cenobamate in patients with developmental and epileptic encephalopathies. Epilepsia. 2025;66(5):1519–1528. 10.1111/epi.18308 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Brandt C, Sánchez‐Álvarez JC, Steinhoff BJ, Milanov I, Serratosa JM. Efficacy and safety of adjunctive cenobamate: post‐hoc analysis of study C017 in patients grouped by mechanism of action of concomitant antiseizure medications. Seizure. 2022;96:86–93. 10.1016/j.seizure.2022.01.016 [DOI] [PubMed] [Google Scholar]
- 23. Kwan P, Arzimanoglou A, Berg AT, Brodie MJ, Allen Hauser W, Mathern G, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc task force of the ILAE commission on therapeutic strategies. Epilepsia. 2010;51(6):1069–1077. 10.1111/j.1528-1167.2009.02397.x [DOI] [PubMed] [Google Scholar]
- 24. Vinklárek J, Búřilová P, Raouf Zadeh M, Zatloukalová E, Kočvarová J, Strýček O, et al. Ultra‐refractory epilepsy: the newly described entity. Epilepsia Open. 2025;10(1):1–10. 10.1002/epi4.12889 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Roberti R, Assenza G, Bisulli F, Boero G, Canafoglia L, Chiesa V, et al. Adjunctive cenobamate in people with focal onset seizures: insights from the Italian expanded access program. Epilepsia. 2024;65:2909–2922. 10.1111/epi.18091 [DOI] [PubMed] [Google Scholar]
- 26. Villanueva V, Serratosa JM, Fernández‐Cabrera A, Toledo M, González‐Giráldez B, Estevez JC, et al. Freedon study: real‐life outcomes of cenobamate in different lines of treatment. Epilepsia. 2026;00:1–13. 10.1002/epi.70106 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Summary of adverse drug reactions (ADRs) reported up to 12 months after cenobamate initiation (assessed at months 2, 3, 6, and 12). Data are presented as n (%). Patients may have experienced more than one ADR. ADRs are classified according to MedDRA System Organ Class. No cases of DRESS syndrome were observed. ASMs = antiseizure medications.
The case of alopecia was assessed as unrelated to the drug; in this specific patient, the event was attributed to pre‐existing androgenetic alopecia rather than the pharmacological intervention.
Table S2. Mean monthly seizure frequency at 3, 6, and 12 months across various cenobamate (CNB) doses, stratified by prior ASM exposure (≤ 6 vs. > 6 previous ASMs). Dash (−) indicates no patients at that dose/timepoint combination. ASMs = antiseizure medications; CNB = cenobamate.
Table S3. Mean monthly seizure frequency at baseline and at 3, 6, and 12‐month follow‐up, stratified by CNB dose and prior ASM exposure (≤ 6 vs. > 6 previous ASMs). Percentage reduction was calculated as [(baseline − follow‐up) / baseline] × 100. Negative values indicate an increase in seizure frequency. Dash (−) indicates no patients at that dose/timepoint combination. ASMs = antiseizure medications; CNB = cenobamate.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.