Cenobamate in the management of focal-onset epilepsy in adults – practical considerations for daily practice

The current dosing regimen was introduced after three cases of DRESS syndrome, including one fatality, were identified among the first 953 participants exposed to cenobamate during early clinical development. DRESS occurred in two healthy volunteers and one patient with epilepsy. DRESS cases were associated with the initiation of CNB treatment at a high starting dose (50, 100, and 200 mg, respectively) and/or dose titration every week or faster (Sperling et al., 2020). The schedule in which treatment is commenced at a low dose (12.5 mg) and increased incrementally every 2 weeks appears to have mitigated the risk of DRESS. In another study assessing the drug’s safety, no other cases of this syndrome were reported (Sperling et al., 2020). As with other ASMs, the ‘start-low, go slow’ titration strategy reduces the risk of hypersensitivity reactions (Zaccara et al., 2007; Hirsch et al., 2006).

While the mechanism underlying DRESS syndrome remains poorly understood, its pathophysiology involves three essential factors: genetic predisposition linked to some specific Human Leukocyte Antigen (HLA) alleles; change in the metabolic pathways of drugs, mainly aromatic antiseizure medications; and reactivation of the human herpesvirus 6 (HHV-6) leading to T-cell mediated inflammatory response, causing tissue damage (Wang and Mei, 2017; Calle et al., 2023).

The response that occurred during the initial use of CNB is not unique. At least 44 drugs have been implicated in DRESS, including antiepileptics, antibiotics, antituberculosis agents, and non-steroidal anti-inflammatory drugs. This particularly applies to the ASMs that contain an aromatic ring, i.e. carbamazepine, lamotrigine, phenobarbital, phenytoin, and oxcarbazepine (Calle et al., 2023 ).

It needs to be highlighted that DRESS is a rare syndrome, with an incidence of 1 : 10,000 to 1 : 1,000 in patients treated with ASMs (Bocquet et al., 1996).

DRESS syndrome is characterized predominantly by the presence of fever above 38°C, maculopapular rash, enlarged lymph nodes, internal organ involvement, and peripheral blood smear changes (eosinophilia, neutrophilia, neutropenia, thrombocytopenia, hemolytic anemia) (Bocquet et al., 1996).

It is essential to explain to patients the reasons for slow CNB dose titration, pay attention to the risk of hypersensitivity-type skin reactions, and monitor patients for adverse reactions that usually occur at the beginning of treatment (between 2 weeks and 3 months).

Given the high mortality rate among patients with DRESS syndrome, which is approximately 10%, depending on the involvement of internal organs, an accurate and quick diagnosis is essential (Wolfson et al., 2019). It is crucial to recognize the early symptoms and discontinue the medication suspected of causing them as promptly as possible. The definitive confirmation of the diagnosis and the initiation of active therapeutic measures need to be made in a specialized dermatological center. However, cooperation with a neurologist or epileptologist is important to protect the patient from the risk of exacerbation of epileptic seizures or the development of status epilepticus in association with ASM discontinuation.

In a long-term study evaluating the safety of CNB (Sperling et al., 2020), aside from skin symptoms associated with DRESS, adverse reactions involving the skin and subcutaneous tissue were relatively rare. The most common were rash, pruritus, dermatitis, and alopecia.

Both RCTs (Krauss et al., 2020; Chung et al., 2020) revealed a dose-dependent shortening of the QT interval on ECG. The effect was most pronounced with the 500 mg dose exceeding the recommended dose. Consequently, CNB is contraindicated in patients with familial short QT syndrome. Also, caution should be exercised when CNB is used concomitantly with other drugs known to shorten the QT interval, such as rufinamide or primidone, as there may be synergistic effects (Zaccara, Lattanzi, 2019). However, no guidelines recommend performing an ECG as a routine procedure before initiating CNB therapy.

The management of women of childbearing potential who are not planning a pregnancy and/or use effective methods of contraception is similar to that of other women. The efficacy of hormonal contraception, such as oral contraceptives, is not guaranteed. CNB acts as a CYP3A4 inducer, which means it can potentially decrease estrogen and progestin levels by over 50% (Zhang et al., 2018). CYP3A4 induction is dose-dependent and mainly occurs with higher doses of CNB (Greene et al., 2019). Therefore, women of reproductive potential are advised to use additional or alternative non-hormonal measures of birth control during CNB treatment. Data on pregnancy in women treated with CNB are very scarce (Ontozry. Summary of product characteristics). As with all new ASMs, based on the currently available evidence, pregnancy is contraindicated in women treated with CNB. Animal studies suggest a potential adverse impact of CNB on the developing fetus (Ontozry. Summary of product characteristics). However, research in humans is needed. If an unplanned pregnancy occurs, the decision to continue or modify the treatment regimen must take into account the yet undetermined risk of congenital malformations and/or developmental disorders in the child, as well as the patient’s seizure control status and other medical factors. The patient and her family must receive all information relevant to them to make an informed decision.

It is unknown whether CNB is excreted in human milk. Furthermore, there is no data on the effects of CNB on breastfed infants or lactation.

Elderly patients are at a greater risk of adverse effects, particularly those associated with the CNS when compared to younger individuals (Motika and Spencer, 2016). Data on the efficacy and safety profile of CNB remain limited. A post hoc analysis of a phase 3 open-label trial in patients aged 65–70, who comprised slightly over 3% of the study population, found good efficacy in this age group. However, these patients were more than twice as likely to discontinue treatment because of adverse effects compared to the other study subjects (O’Dwyer et al., 2023). Adverse reactions reported in ≥20% of elderly patients included dizziness, somnolence, falls, fatigue, balance disorders, and upper respiratory tract infections. In elderly patients starting treatment with CNB, reducing the doses of concomitant ASMs may be necessary, potentially to a greater extent than in younger patients.

As with other patients, treatment should be initiated at 12.5 mg and gradually titrated. The rate of titration to the target dose, as well as the target dose itself, depends on efficacy and tolerability.

Prescription of appropriate therapy should be based on the understanding that roughly 50% of adults with epilepsy have comorbidities (Gaitatzis et al., 2012) and typically take multiple other medications. When introducing a new ASM, it is essential to have an in-depth knowledge of its metabolism. Similarly to other special patient populations, clinical data on the use of CNB in patients with psychiatric and somatic comorbidities are very limited.

In a small cohort of 28 patients with developmental disabilities and uncontrolled epilepsy residing in a nursing home, a significant improvement was noted after 12 months of CNB treatment in the majority of subjects. In some cases, it was possible to either reduce the dose of other ASMs or discontinue them altogether (Connor, 2017).

CNB is largely metabolized in the liver via glucuronidation and oxidation and then excreted mainly via the kidneys (Ontozry. Summary of product characteristics). Since pharmacokinetic studies have found elevated plasma levels of CNB in patients with hepatic and/or renal impairment, the CNB dose should be adjusted in these patient groups (Ontozry. Summary of product characteristics). Reduction of the target dose may be considered in patients with renal impairment if their creatinine clearance (CrCl) is <90 mL/min (Ontozry. Summary of product characteristics). CNB is not recommended in patients with CrCl below 15 mL/min and in patients with end-stage renal disease undergoing dialysis treatment.

CNB should not be used in patients with severe hepatic impairment (Ontozry. Summary of product characteristics), while the maximum recommended dose in patients with mild and moderate hepatic impairment is 200 mg/day.

The most prevalent psychiatric condition among individuals with epilepsy is depression. It exhibits a close bidirectional relationship with both biological and psychosocial factors (Mula, 2019). The presence of mood disorders is a strong predictor of suicidal behavior.

Most ASMs are labelled with an FDA-issued warning stating that ASMs as a drug class and across all indications, including but not limited to epilepsy, may be associated with an increase in suicidal ideation or behavior (Bell et al., 2009) based on a meta-analysis of RCTs conducted before 2008. Since then, all approved ASMs have carried this warning, even though they were never evaluated in similar studies. A recent 2021 meta-analysis including CNB (Klein et al., 2021) is critical of this cautionary advice. In a long-term open-label trial assessing the safety of CNB, three cases of attempted suicide and one case of severe suicidal ideation were reported (Sperling et al., 2020). While no definitive link has been established between CNB and suicide, the FDA recommends, as with other ASMs, monitoring patients for signs of suicidal ideation and behaviors. This is very important because patients with epilepsy face an almost threefold more significant risk of suicide compared to the general population (Christensen et al., 2007). Clinicians should be especially vigilant in assessing the mental well-being of patients, even in the absence of solid evidence confirming the causal role of ASMs (Mula et al., 2013; Kanner, 2022).

Overall, in a long-term open-label trial of CNB, psychiatric adverse reactions were rarely reported and included anxiety, irritability, and depression (Sperling et al., 2020).

A 36-year-old patient with drug-refractory focal-onset epilepsy of inflammatory etiology, persisting since the age of 5, participated in 2014 in a randomized, double-blind, placebo-controlled trial evaluating the efficacy of cenobamate. In the course of the disease, the patient experienced frequent focal-onset cluster seizures evolving to bilateral tonic-clonic seizures. The monthly frequency of seizures was approximately 6 to 9. At baseline, the patient was treated with oxcarbazepine (OXC) at a dose of 1,500 mg/day and valproate at 1800 mg/day.

Based on the study design, the patients randomized to the study drug received a starting dose of 50 mg, escalated at weekly intervals. After 2 weeks of treatment, the drug was well-tolerated. During this period, the patient had five typical seizures. After another 2 weeks, the patient reported two episodes of diplopia, each lasting approximately half an hour. No seizures occurred during this period. A few days into week 5 of treatment, diplopia accompanied by reduced alertness and balance disorders occurred at a daily frequency, with the onset approximately one hour after taking the medicine and lasting up to three hours. An attempt to stagger the administration of the study drug and the patient’s preexisting medications was unsuccessful. Because of intolerable side effects, the drug was discontinued over a few weeks, and the patient withdrew from the study. In the 3rd week of discontinuation, the patient’s seizures returned.

Considering that the only seizure-free period in the course of the patient’s treatment was the time when he was receiving CNB in 2022 after the drug was approved, but before it was included in the reimbursement scheme, another attempt at CNB treatment was made. In November 2022, the patient had between 3 and 6 seizures a month. During that period, he received OXC at a dose of 900 mg/day, LCM at 300 mg/day, and brivaracetam (BRV) at 100 mg/day.

Given the high number of ASMs (three) and adverse effects related to drug interactions observed during the clinical trial, a decision was made to discontinue one of the drugs. OXC, which was probably responsible for the drug interaction, was previously used at a considerably higher dose (1,800 mg vs 900 mg/day). Several prior attempts to withdraw OXC were unsuccessful and resulted in cluster seizures followed by delusional episodes. The last drug introduced was BRV, which contributed to a reduction in seizure frequency, though the effect was insufficient. On the other hand, LCM appeared not to exhibit therapeutic efficacy. Given the potential interaction, gradual complete discontinuation of LCM before introducing CNB was advised. CNB treatment was initiated according to the recommended schedule, starting at 12.5 mg/day and escalating the dose by increments every 2 weeks up to the target 200 mg/day dose.

The drug was well tolerated. The patient has now been seizure-free for 4 months.

The patient’s case serves as an illustration of how therapeutic management impacts the success of treatment. Most importantly, the significance of the starting dose and titration rate is evident. Patients participating in the clinical trial were not allowed to adjust the dose of concomitant ASMs during rapid titration of cenobamate, which was likely to have caused the adverse effects (Krauss et al., 2020). Given the potential pharmacodynamic interactions resulting from combining drugs with a similar mechanism of action via sodium channels, particularly at high doses, LCM was proactively discontinued in this patient, which was additionally validated by the lack of significant benefits. The decision was also motivated by the fact that the patient took as many as three ASMs. Withdrawing ASMs that are unnecessary and potentially harmful is an important therapeutic goal. Nonetheless, in some patients, especially those with a history of seizure exacerbation during ASM withdrawal, CNB therapy may be initiated first, and after titration to higher potentially effective doses, the ASM scheduled for discontinuation can be gradually withdrawn.

In the reported patient, the further course of treatment will be determined by the duration of seizure freedom. Some drug-resistant patients experience a phenomenon referred to as the “honeymoon effect”, observed for all ASMs (Peña-Ceballos et al, 2023; Löscher and Schmidt, 2006), which may suggest the evolution of drug tolerance. Therefore, it is reasonable to anticipate that with CNB, the seizure cessation effect might also be temporary in a specific population of patients. However, the retention rates in long-term studies are promising (French et al., 2021; Sander et al., 2022; Sperling et al., 2021). The reported patient’s dose may be increased above 200 mg if needed. Then, the patient will require monitoring for potential dose-dependent adverse reactions associated with CNB itself, given that CNB used at doses exceeding 300 mg exhibits non-linear pharmacokinetics (Ontozry. Summary of product characteristics; Vernillet et al., 2020). In addition, the patient may require readjustment of the doses of concomitant medications, though OXC and BRV are not currently used at high doses.