In the last decades, numerous anti-seizure medications (ASMs) with different mechanisms of action have been marketed representing new therapeutic alternatives for clinicians and for patients with epilepsy (Perucca et al., 2020). Despite the early use of new ASMs, up to 30% of patients will have inadequate seizure control (Kwan and Sander, 2004; Mohanraj and Brodie, 2005) with the chances to reach seizure freedom being lower with each treatment failed (Chen et al., 2018). The severity of epilepsy results in an urgent need for developing new and more effective pharmacologic treatments (Perucca et al., 2020). Cenobamate (CNB) is the latest ASM approved as adjunctive treatment for focal-onset seizures (FOS) in adult patients who have not been adequately controlled despite a history of treatment with at least two ASMs (Roberti et al., 2021). CNB has multiple mechanisms of action: it mainly inhibits the persistent component of the sodium current trough the inactivation of the neuronal voltage-gated sodium channels (Nakamura et al., 2019), and acts as a positive allosteric modulator of the γ-aminobutyric acid (GABA_A) ion channel binding at a non-benzodiazepine site (Sharma et al., 2020). Adjunctive CNB in adult patients with uncontrolled FOS has been associated with a greater reduction in seizure frequency than placebo in two pivotal multicenter, randomized, double-blind, placebo-controlled trials enrolling 659 patients (Lattanzi et al., 2020; Specchio et al., 2021). Interestingly, CNB treatment has been associated with seizure freedom rates which are considerably higher than those reported for other ASMs (Villani et al., 2022). A phase III, open-label, safety study (Sperling et al., 2021) further documented the safety of CNB when the drug is started at a lower dose (12.5 mg/day) and up-titrated bi-weekly. To date, only few real-world studies evaluated the efficacy and safety of CNB. Recently published series provided evidence about the efficacy and safety of CNB in both pediatric and adult patients, including patients with developmental and epileptic encephalopathies (DEEs) (Elliott et al., 2022; Makridis et al., 2022; Varughese et al., 2022; Beltrán-Corbellini et al., 2023; Friedo et al., 2023; Peña-Ceballos et al., 2023; Villanueva et al., 2023).

While data of safety and effectiveness are available, data on effects of CNB on neuropsychological functions are still lacking, moreover drug to drug interactions and blood levels of CNB are also missing. In the present study, we report the effectiveness, tolerability, neuropsychological outcomes, and occurrence of drug-drug interactions through the analysis of the CNB blood levels in a cohort of drug-resistant adult patients treated with CNB as adjunctive treatment.

Fifty-four patients (31 F) with a median age of 27.9 years (IQR = 21.7-33.4) were enrolled. The mean follow-up duration was 10.7 months [range ±standard deviation (SD) 6.0-15.0 ± 2.1]. All patients presented FOS, 18 (33.3%) of them suffering from focal to bilateral tonic-clonic seizures. Patients were previously treated with a median of 9 ASMs (IQR = 7–9). When CNB was started, patients were taking a median of three ASMs (IQR = 1–4). The most common concomitant ASMs at the enrollment were carbamazepine (n = 24, 44.4%), clobazam (n = 22, 40.7%), and lacosamide (n = 17, 31.5%). Sixteen patients (29.6%) had vagal nerve stimulation therapy and 19 out of 54 patients (35.2%) had failed epilepsy surgery. Details of demographic and clinical characteristics of the study cohort are shown in Table 1.

Titration period up to 100 mg/day lasted for a mean period of 58.4 (range ±SD 30-151 ± 18.9) days; 107.7 (70-278 ± 40.3) days was the mean time to reach 200 mg/day and 167.9 (14-434 ± 109.7) days were needed to achieve the maximum dose of CNB (204 mg, 50-400 ± 90). At LFV, the mean dose of CNB was 201.8 mg/day (25-400 ± 96.9). In 9 patients (16.7%) CNB was withdrawn due to AEs (n = 5; 9.2%), inefficacy (n = 2; 3.7%) or seizure worsening (n = 2; 3.7%).

In this retrospective, real-world study of CNB treatment in patients with drug resistant focal epilepsies, add-on CNB provided a clinically significant reduction in seizure frequency in most of the patients. The results we found in the current study are consistent with those observed in previously published randomized controlled trials (RCTs) (Chung et al., 2020; Krauss et al., 2020) and real-life experiences (Elliott et al., 2022; Makridis et al., 2022; Varughese et al., 2022; Beltrán-Corbellini et al., 2023; Friedo et al., 2023; Peña-Ceballos et al., 2023; Villanueva et al., 2023).

Of note, this study allowed to obtain more information about the use of CNB in the context of the real-world practice, adopting a more flexible drug dosing regimen and providing data over a longer follow-up period if compared to earlier RCTs.

The median percentage reduction of seizure frequency of 69.5% and the responder rate of 59.2% were slightly higher compared to those reported in both pivotal studies: in C013 and in C017 trials, the reductions in median seizure frequency were 55.6% and 56% and the responder rates were 50.4% and 56% at CNB dose of 200 mg/day, respectively (Chung et al., 2020; Krauss et al., 2020; Lattanzi et al., 2020). The proportion of seizure-free patients in the present study was 20.4% at a mean CNB dose of 200 mg/day. This figure is lower if compared with the rate found in the C013 study (28% at maintenance dose of CNB 200 mg/day) (Chung et al., 2020), and higher than the rate reported in the C017 study (11% at maintenance dose of CNB 200 mg/day) (Krauss et al., 2020). Considering the different study designs, results in our cohort of patients seem to be slightly better if compared with the previous RCTs. This can be explained by the possibility of adjusting and personalizing the dose of CNB and concomitant ASMs, which may have given higher responder rates.

The results of this study were similar to results from other real-world studies of CNB, which also included patients with highly refractory epilepsy.

In a recent large series of patients with highly drug-resistant epilepsy, including 170 patients, the rate of seizure freedom was 13.3%; ≥90%, ≥75%, and ≥50% responder rates were 27.9%, 45.5%, and 63%, respectively. There was a significant reduction in the number of seizures per month (mean: 44.6%; median: 66.7%) between baseline and the last visit (p < 0.001) (Villanueva et al., 2023). Compared with this study (Villanueva et al., 2023), we observed a higher percentage of seizure-free patients (13.3% vs. 20.4%). This could be due to differences in the cohort composition, since our patients were slightly younger, and a lower number of previous treatments, suggesting that CNB could be associated with a better response when used earlier. Of note, we also found that 8.9% of patients were nearly to seizure freedom (patients with no more than one seizure during six consecutive months of the study period).

From a series of 45 adolescent and adult patients who had received a mean of 12 prior ASMs, Elliot et al. reported seizure freedom in 16% of patients and a ≥50% response in 60% (Elliot et al., 2022). A pediatric real-world study found a seizure freedom rate of 31% and a >50% response rate of 37.5% (Makridis et al., 2022). In another pediatric real-world study, 19% of patients achieved seizure freedom, 52% had a ≥75% response rate, and 63% had a ≥50% response rate (Varughese et al., 2022). Additionally, our results were in line with previous findings from an EAP in Ireland, which 57 patients with ultraresistant epilepsy and showed seizure reduction in 75%–99% of the cohort in 42.1% of patients (Peña-Ceballos et al., 2022).

More recently, in the Beltran-Corbellini et al. study, in 51 patients with highly refractory focal epilepsy, retention rate at the last follow-up was 80.4% and the 50% responder rate in focal seizures was 56.5% and in focal to bilateral tonic-clonic seizures was 63.6% (Beltran-Corbellini et al., 2023).

Finally, a drug load reduction was possible in 12 patients with developmental and epileptic encephalopathies, thanks to CNB effectiveness, maintaining seizures reduction (Friedo et al., 2023).

Considering the responder rates, we found some differences in patients with and without concomitant SCBs: the concomitant use of SCBs increased the percentage of seizure free patients from 10% to 23% and the percentage of responders from 40% to 64%. Concomitant use of SCBs and CNB, should be carefully evaluated for tolerability issues. Although a similar increase of responders was seen in patients with and without concomitant CLB (67% versus 56%), the seizure freedom rate was higher among patients not taking concomitant CLB (26.5% versus 6.7%); we do not have a clear-cut explanation for this latter finding, which should be further investigated in future studies.

The longitudinal evaluation of efficacy revealed an increase in the responder rates from the first to the third trimester, going from 53% to 67.3% (≥50% responders) and from 8.2% to 18.4% (seizure freedom). An improvement of the drug efficacy is seen over time, possibly related also to the increase in drug dosage, and clinicians should wait before considering CNB withdraw if the tolerability is kept, even if we have to acknowledge that—as it always happens in long term observation—the overall number of patients during time decrease making the results more attractive.

Reducing the load of concomitant ASMs is highly warranted in patients with drug resistant epilepsy: in the study cohort, most patients were taking a median of 3 ASMs and still experiencing a median of 11.7 seizures per month. In this regard, the 60% of the participants were able to reduce or withdraw one or more concomitant ASMs during the follow-up.

The incidence of AEs was higher if compared with literature data, even though a clear association with a specific CNB dose was not identified. Most AEs occurred at a relatively low dose of CNB (100 mg/day) and without the dose-related association registered in previous clinical trials (Chung et al., 2020; Krauss et al., 2020). A possible explanation of the poorer tolerability may be the use of multiple concomitant ASMs in our cohort of patients. Central nervous system-related side effects were the most reported as already found with the use of other currently available ASMs (Zaccara et al., 2013). Ataxia and diplopia were more frequent when a concomitant SCB was taken, likely as the result of the pharmacodynamic interaction that may occur among agents acting as SCBs.

In many cases, a reduction in the number and/or dose of concomitant ASMs was required to resolve AEs. SCBs and CLB were the most frequently modified concomitant drug regimens. A proactive lowering of ASMs with known pharmacodynamic (carbamazepine and lacosamide) and pharmacokinetic interactions (phenobarbital, phenytoin, CLB) with CNB may be indicated to prevent AEs and allow an optimal CNB dosing (Smith et al., 2022).

A linear correlation was shown between the daily dose and the plasma concentration of CNB in the dose range 100–400 mg/day. Dose-proportional increases in both maximum concentration and plasma exposure are confirmed by single and multiple ascending dose studies (Vernillet et al., 2020), but non-linear pharmacokinetics is expected at higher doses (more than 300 mg). In our cohort, the number of patients treated with more than 300 mg of CNB was too small to assess any disproportionality in dose responses.

Plasma levels measurements confirm that CNB acts both as “victim” and as “perpetrator” of drug-drug interactions. Indeed, in patients treated with at least one inducer, the mean C/D ratio of CNB was significantly lower if compared with patients not taking an inducer. On the other hand, the reduction of CBZ plasma concentrations induced by CNB concomitant administration (Roberti et al., 2021) has been confirmed in our cohort. Also, the pharmacokinetic interactions involving CLB, and its metabolite have been confirmed, considering that we observed a 2-fold increase in N-CLB levels at CNB 200 mg/day. Such increase of N-CLB was associated with somnolence in our patients. Notably, a variable (between 145% and 1852%) increase in N-CLB serum concentration has been reported in a recent study (Elakkary et al., 2023). The authors suggest a potential role of the pharmacogenetics of the hepatic enzyme CYP2C19 to explain this high rate of variability and report fatigue as a direct consequence of N-CLB increase (Elakkary et al., 2023). In the same study, a synergistic interaction between CLB and CNB has been hypothesized, because both ASMs potentiate GABA transmission acting at different sites of the GABA_A receptor (Elakkary et al., 2023).

Most notably, there was a certain degree of variability within the same dose among patients, therefore, TDM is likely recommended to better adjust therapy. This is also in agreement with our observation of the reduction of CNB concentration when inducer ASMs are concomitantly used, and the bidirectional influence observed.

The lack of a significant correlation among CNB doses and plasma concentrations of other concomitant ASMs may be due to the overall limited number of observations.

General Anxiety Disorder-7 test showed a significant improvement in symptom severity of anxiety score in our cohort. This result could be related to the reduction of seizures burden or to the CNB pharmacodynamic mechanism as positive allosteric modulator of the GABA_A ion channel (Löscher et al., 2021). We failed to find any correlations between reduction of anxiety, the status of responder, and CNB blood concentrations. More studies should be performed to document and confirm if the reduction of anxiety is related to seizure’s reduction or to the GABA_A modulation. We observed an improvement in Raven’s Progressive Matrices 38 IQ, TMT-A and ABAS-II probably due to the reduction of the concomitant drug load.

If confirmed by future studies, the efficacy shown by CNB in reducing seizures in patients with FOS may support the indication to use CNB sooner in the treatment algorithm of focal epilepsy. We reported that lower doses of CNB may be as effective as those used in previous RCTs. Moreover, the use of add-on CNB allowed to reduce the overall load of concomitant ASMs.

Therapeutic drug monitoring could be a useful tool to improve CNB dosing and prevent possible AEs. CNB appears to reduce symptom severity of anxiety.

CNB was safe and provided sustained, clinically meaningful seizure reduction in this real-world study. More studies may better establish the long-term safety and efficacy of CNB and identify possible predictors of seizure control.