Long-term efficacy and safety of perampanel as a first add-on therapy in patients with focal epilepsy: Three-year extension study

PURPOSE
We investigated the long-term efficacy and safety of perampanel as a first add-on therapy in patients with focal epilepsy.


METHODS
This retrospective study represented the 3-year extension phase of a multicenter, open-label, phase 4, prospective study of perampanel as a first add-on therapy in patients with focal epilepsy. Seizure and safety outcomes were assessed annually from the start of the extension study, and the retention rate was calculated from the start of perampanel exposure in the original study.


RESULTS
The 50% responder and seizure freedom rates were 84.8% and 58.7%, respectively, during the third year and 71.7% and 32.6%, respectively, during the entire 3-year period of the extension study. The 1-, 2-, and 3-year retention rates were 62.5%, 53.1%, and 52.1%, respectively. Efficacies were higher in patients that were aged >55 years, male, and receiving ≤4 mg of perampanel. Perampanel was generally well tolerated; 47.3% of patients experienced at least one adverse event during the 3 years of extension (46 adverse events (AEs) in 35 patients). The most common AEs were dizziness (33.8%), somnolence (5.4%), anger (4.1%), and irritability (4.1%). AEs were resolved with perampanel dose reduction or discontinuation in 10 (13.5%) and 12 (16.2%) patients, respectively.


CONCLUSION
Long-term treatment with perampanel as a first add-on therapy did not raise new safety signals in patients with focal epilepsy. Especially at low perampanel doses (≤4 mg/day), sustained improvement in seizure control was achieved, which could potentially avoid adverse drug reactions.


Introduction
Epilepsy characterized by recurrent seizures is one of the most common chronic brain diseases affecting both sexes and all ages. In a recent meta-analysis, the point prevalence of active epilepsy was 6.38 per 1000 people, indicating that there are more than 50 million patients with epilepsy worldwide [1]. Recurrent seizures in patients with epilepsy have substantial impacts on cognitive, psychological, and social functioning and may even cause sudden death [2,3]. Antiepileptic drugs (AEDs) represent the most important approach for treating epilepsy. Monotherapy with a single AED can produce seizure freedom for at least one year in 61% of previously untreated patients [4]. Usually, when seizure recurrence persists despite two or three trials with monotherapy, combining two or more AEDs with different mechanisms of action is required to control seizures [5,6].
Perampanel, a selective and non-competitive a-amino-3-hydr oxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor antagonist, is a third-generation AED [7]. It inhibits AMPA receptors, which can lead to broad-spectrum anti-seizure effects [7]. The efficacy of perampanel treatment for focal seizures with or without focal to bilateral tonic-clonic seizures (FBTCS) has been documented [8][9][10]. In 2015, the South Korean Ministry of Food and Drug Safety approved perampanel as an add-on therapy for focalonset seizures including FBTCS and generalized tonic-clonic seizures (GTCS) in patients with idiopathic generalized epilepsy. Since May of 2020, perampanel is approved in South Korea as monother- apy and adjunctive therapy for focal seizures including FBTCS in patients aged >4 years and as adjunctive treatment for GTCS in patients aged >7 years.
Early clinical trials have evaluated perampanel as an adjunctive therapy for patients with drug-resistant epilepsy who were taking two or more AEDs [8][9][10]. Recently, perampanel as an early add-on therapy has been investigated in patients with epilepsy who do not respond to AED monotherapy [11][12][13][14][15][16]. These studies generally showed much higher rates of seizure reduction and seizure freedom than early phase III studies [11][12][13][14]. For example, in an open-label prospective study of perampanel as a first add-on therapy, the 50% responder rate was as high as 80.0% during a 24-week maintenance period [11]. This is in sharp contrast to the 50% responder rates (28.5%, 35.3%, and 35.0% with 4 mg/d, 8 mg/d, and 12 mg/d of perampanel, respectively) of a pooled analysis of three phase III studies of adjunctive perampanel for the treatment of refractory focal seizures [8]. Furthermore, the efficacy of perampanel is higher when perampanel is administered as an add-on therapy to the first AED monotherapy rather than the second or third AED monotherapy [15,16]. These findings suggest that perampanel has more favorable efficacy in patients with fewer refractory seizures [17].
Because the history of perampanel development is relatively short, there are few long-term studies on the efficacy and tolerability of perampanel, except open-label extension (OLEx) studies of phase II or III clinical trials [18][19][20][21]. Especially for chronic diseases such as epilepsy, long-term studies are essential for monitoring the effectiveness and long-term safety of drugs that are newly approved largely based on short-term randomized controlled trials. Long-term drug retention represents a useful measure of the effectiveness and tolerability combined over time [22]. Thus, we examined the long-term efficacy and tolerability of perampanel as a first add-on therapy in patients with focal-onset seizures including FBTCS.

Patients
This retrospective study presents the 3-year extension phase of a multicenter, open-label, phase 4, prospective study that evaluated the efficacy and safety of perampanel as a first add-on therapy for the treatment of focal epilepsy [11]. The initial phase 4 study included male and female patients aged 12 years with a diagnosis of focal epilepsy who required add-on therapy after failure to control seizures with AED monotherapy. The main inclusion criterion for patients was at least two focal-onset seizures with or without FBTCS during the 8-week baseline period during which patients were maintained on a constant dose of one AED. Patients who completed the 12-week titration phase and 24-week maintenance phase of the initial phase 4 study were eligible for inclusion in this study. The main exclusion criteria were the presence of only non-motor focal aware seizures, GTCS, absence seizures, and/or myoclonic seizures of idiopathic generalized epilepsies or a history of Lennox-Gastaut syndrome or status epilepticus. Data were collected from clinical records.
The study protocol was reviewed and approved by the Institutional Review Boards for ethical approval of all participating centers. The study was conducted in accordance with the principles of the World Medical Association Declaration of Helsinki (2008). The need for informed consent was waived owing to the retrospective nature of the study.

Perampanel treatment
Patients orally received perampanel tablets once daily before bedtime. The dose of perampanel upon entry into the present study (i.e., the final dose at the end of the maintenance period of the initial study) was individualized for each patient with the aim of achieving the best seizure control with acceptable tolerability. Perampanel could be increased by 2 mg/d over 2-week intervals to a maximum of 12 mg/d. The dose of perampanel could be reduced or withdrawn based on the patient's and physician's judgment. Concomitant treatment with other AEDs was allowed, as clinically indicated and based on each investigator's clinical judgment.

Study assessments
The primary efficacy outcome measures were the 50% and 100% (seizure freedom) responder rates for total seizures during the 3year extension of the initial study. A 50% responder was defined as a patient who had a 50% reduction in total seizure frequency from the baseline of the original study. Seizure freedom was defined as no seizure recurrence. The primary efficacy endpoints were assessed annually from the start of the extension study. Focal aware non-motor seizures were not assessed. The secondary efficacy endpoint comprised perampanel retention rates. Retention was defined as the time from the start of perampanel exposure in the original study to withdrawal of perampanel due to any reason.
The safety outcome was measured using the proportions of patients with adverse events (AEs) and discontinuation of perampanel from the start of the extension study. Adverse events were obtained from clinical records and classified as any AEs, AEs necessitating dose reduction, and AEs leading to discontinuation of perampanel.

Statistical analysis
The primary efficacy outcomes for each treatment year of the 3year extension period were evaluated using the full analysis set defined as all patients who were treated with at least 6 months of each treatment year and had efficacy assessment data available. The retention rate was assessed using the safety analysis set defined as all patients who were treated with at least one dose of perampanel from the start of perampanel exposure in the original study. Safety/tolerability assessments were performed using the safety analysis set defined as all patients who were treated with at least one dose of perampanel during the 3-year extension study.
For retention rates, percentages of patients who continued with perampanel treatment at 1 year, 2 years, and 3 years were reported. A Kaplan-Meier survival curve was generated to analyze the retention of perampanel during the follow-up period. Using either the Pearson chi-squared test or Fisher's exact test, post hoc analyses were performed to determine associations of clinical factors with efficacies and AEs when stratified by sex; age (>55 vs. 55 years); epilepsy etiology (known vs. unknown); the presence of FBTCS at baseline; perampanel dose (4 vs. 6 mg/d); enzyme-inducing activity of concomitant AEDs; or use of levetiracetam, carbamazepine, or oxcarbazepine (used vs. not used). All statistical tests were two-tailed, and p < 0.05 was considered statistically significant. Statistical analyses were performed using the IBM SPSS Statistics for Windows, version 21.0 (IBM Corp., Armonk, NY, USA).

Patient characteristics
Of the 80 patients who completed the initial phase 4 prospective study, 6 were excluded due to administrative reasons. Thus, 74 patients were enrolled in this retrospective extension study and were included in the safety analysis set (Fig. 1). A further 19, 7, and 2 patients were excluded because their primary efficacy outcomes were not evaluated during the first, second, and third years, respectively, of the extension study. Therefore, 55, 48, and 46 patients were included in the full analysis set of the first, second, and third years, respectively.
The retention rate was calculated from the start of perampanel exposure in the original study. Of the 106 patients enrolled in the initial study, 4 patients did not receive any perampanel, and 6 patients were not included in the extension study (Fig. 1). Therefore, 96 patients were included for calculation of the retention rates. The 1-, 2-, and 3-year retention rates were 62.5%, 53.1%, and 52.1%, respectively ( Table 2). The most common reason for discontinuation of perampanel were AEs (24.0%, n = 23), followed by follow-up loss (10.4%, n = 10) and lack of efficacy (6.3%, n = 6). Most discontinuations occurred during the first year of perampanel exposure (Fig. 3).
Results from the post hoc analyses are shown in Table 3. Efficacies, especially 3-year seizure freedom rates and/or 50% responder rates, were higher in patients that were aged >55 years, male, and receiving the final dose of perampanel at 4 mg/day. Efficacy outcomes did not differ regarding FBTCS at baseline, epilepsy etiology (known vs. unknown), or types of concomitant AEDs. The 3-year retention rate was significantly higher in patients taking enzymeinducing AED (EIAED) (p = 0.034), especially oxcarbazepine (p = 0.027). Levetiracetam was a concomitant AED in 10 of 23 patients who discontinued perampanel due to AEs.

Discussion
In the 3-year extension study of the initial phase 4 prospective study [11], the long-term efficacy and tolerability of perampanel were retrospectively investigated as a first add-on therapy in patients with focal epilepsy who did not respond to AED monotherapy. We found that the 50% responder and seizure freedom rates were 76.4% and 43.6%, respectively, during the first year from the start of the extension study and 84.8% and 58.7%, respectively, during the third year of extension. These efficacies were similar to those of the short-term studies of perampanel as a first add-on therapy. For example, our initial phase 4 prospective study showed that the 50% responder and seizure freedom rates were 80.0% and 47.1%, respectively, during a 24-week maintenance period after the 12-week titration phase [11]. Another one-year study of perampanel as a first add-on therapy also showed that 50% responder and seizure freedom rates were 82.3% and 40.7%, respectively, in patients with focal epilepsy [12]. Taken together, these findings suggest sustained long-term efficacy of perampanel in patients with focal epilepsy over 3-4 years of treatment. Although the surviving cohorts have a selection bias, as only the patients doing well remained on treatment, less than 7% of patients in the present study discontinued perampanel treatment because of perceived inadequate therapeutic effects. To date, there have been no comparable long-term studies of perampanel as a first add-on therapy in patients with focal epilepsy. The sustained long-term efficacy in the present study is in agreement with an OLEx study (Study 307) from phase III trials for patients with focal epilepsy that demonstrated that the reduction in seizure frequency and 50% responder rate remained stable over 4 years in each yearly completer cohort [20]. Our efficacy outcomes were more favorable than those of OLEx studies of phase II or III randomized trials in which patients with more drug-resistant seizures were enrolled [18,20,21]. Specifically, in Study 307, the 50% responder and seizure freedom rates were 72.1% and 40.0%, respectively, after 3 years in a cohort treated for at least 3 years and 78.6% and 53.6%, respectively, after 4 years in a cohort treated for at least 4 years [20].
During the 3-year extension period in which patients received the final dose of perampanel at 4 mg/day, 87.5% of patients experienced 50% seizure reduction, and 45.8% of patients experienced seizure freedom. These efficacies were higher than those of patients receiving perampanel at 6 mg/day. Our findings are consistent with a recent Japanese study in which >50% of patients obtained seizure freedom with a minimum perampanel dose of 2 mg [16]. These findings suggest that some patients may achieve seizure control with relatively low perampanel doses, whereas others may require up-titration to higher doses. We also found that patients aged >55 years were more likely to achieve seizure freedom within 3 years. Similarly, a multicenter, retrospective, 1year, observational study showed a better clinical response to perampanel in patients with focal epilepsy aged 65 years than in those aged <65 years [17]. Furthermore, our study revealed no effects of EIAEDs on efficacy, most likely because most patients (89.2%) received 6 mg of the final dose of perampanel. Data from the phase III trials showed that perampanel doses of 8 and 12 mg resulted in significantly greater treatment responses in patients receiving non-EIAEDs [23]. Pooled observational data on the routine use of perampanel at 45 European centers revealed that the use of EIAEDs was weakly associated with lower chances of seizure freedom after 1 year [24].   Male patients had higher seizure freedom rates and 50% responder rates than female patients during a whole 3-year extension period, but not during the third year of extension. These findings were likely to result from different retention rates between sexes. Specifically, male patients had the lower retention rate (42.5%) than female patients (57.1%), although it was not statistically significant. Such a lower retention rate of perampanel would link to the higher rates of efficacies in male patients in the present study. There have been no studies supporting difference in efficacy of perampanel between sexes in the literature. A Spanish study found no sex difference in 12-month seizure freedom rate in patients with focal epilepsy who were treated with adjunctive perampanel for one year [17]. In addition, a Portuguese study did not find sex dif-  ference in 1-year retention rate of perampanel as an add-on treatment for patients with focal epilepsy [25].
In the present study, the 1-, 2-, and 3-year retention rates assessed from the start of perampanel exposure were 62.5%, 53.1%, and 52.1%, respectively. These retention rates were similar to those of the SANAD study, an unblinded, randomized, controlled trial of AED monotherapy in newly diagnosed patients with focal epilepsy [26]. In the SANAD trial, 3-year retention rates were 59%, 53%, 52%, and 45% for the first monotherapy with lamotrigine, carbamazepine, oxcarbazepine, and topiramate, respectively. Furthermore, our results were also similar to those of the OLEx studies of phase II or III trials [18,20,21,27,28], which enrolled patients with drug-resistant epilepsy. Specifically, in the OLEx Study 307 of perampanel, the 3-year retention rate was 46.2% in patients with focal-onset seizures (with or without FBTCS) [20]. Similarly, the OLEx studies of lacosamide showed that the retention rates were 50.8%-52.9% after 3 years in patients with focal-onset seizures [27,28]. A retrospective hospital-based study of new AEDs also revealed that drug retention rates were not significantly different in patients who were taking AEDs as monotherapy compared to adjunctive therapy [29]. The primary reason for drug discontinuation differs depending on the type of AEDs. In the present study, the most common reason for discontinuation of perampanel were AEs (48.9% of 47 discontinued cases), followed by follow-up loss (21.3%) and lack of efficacy (12.8%). Approximately two-thirds of perampanel discontinuation occurred within a year of exposure. In a retrospective hospital-based retention study, levetiracetam was discontinued primarily due to behavioral side effects (40.4% of discontinued cases), whereas oxcarbazepine was discontinued primarily due to drug ineffectiveness (45.2% of discontinued cases) [29].
We found that the 3-year retention rate was significantly higher (64.1%) in patients taking EIAEDs, especially oxcarbazepine, than in those taking non-EIAEDs as concomitant AEDs. Our findings support data from the phase III trials, which showed that discontinuation rates due to AEs were lower in patients receiving EIAEDs and perampanel doses of >4 mg/day [23]. A retrospective observational study also showed that patients on an EIAED were less likely  to experience an AE within 12 months of using perampanel and to discontinue perampanel due to an AE [17]. Also in the present study, patients taking EIAEDs, especially carbamazepine, were less likely to have any AEs. Nevertheless, this finding was not statistically significant, probably because of the small sample size. We found no new safety signals comparing to previous phase III trials and their OLEx studies [8][9][10][18][19][20][21]. Perampanel was generally well tolerated, and no serious AEs were observed over the 3year extension period in the present study. The most common AEs were dizziness, somnolence, anger, and irritability. In total, 12 patients discontinued perampanel due to AEs: seven (9.5%) patients due to dizziness and three (4.1%) patients due to anger/ir-ritability. In patients with focal epilepsy treated with adjunctive perampanel for up to 4 years in an OLEx study of phase III trials [20], dizziness (4.1%), irritability (1.2%), and fatigue (1.1%) were identified as the most common AEs leading to discontinuation of perampanel. Aggression (0.6%) and convulsion (0.4%) were the most common serious AEs leading to discontinuation of perampanel [20]. We found that patients aged >55 years were more likely to have AEs leading to dose reduction or discontinuation. Female patients were more likely to have AEs than male patients; however, the proportion of patients with AEs leading to dose reduction or discontinuation did not differ between sexes. Similarly, when analyzed by sex, phase III studies showed that common AEs, such Data are presented as number of patients (%). as dizziness and headache, occurred more frequently in female subjects; however, tolerability was similar between sexes [30]. Regarding patients not receiving EIAEDs, a pharmacokinetic analysis revealed that perampanel oral clearance was 17% lower in female patients than in male patients [30]. Some caution should be exercised when interpreting our findings. First, retrospective multicenter studies with long-term follow-ups have an uncontrolled nature and thus may result in the use of unstandardized methods for collecting information on clinical efficacies or AEs across centers. Second, the number of patients included in this study was relatively small, which may have influenced our results. Furthermore, multiple regression could not be applied in the statistical analysis. Third, it is possible that the results may be influenced by changes in background AEDs during the 3-year extension study. However, additional add-on therapy did not improve seizure outcome in the present study (Fig. 2), signifying the refractoriness of these patients. Specifically, the 50% responder rate in the third year tended to be higher in patients who remained on the same AEDs throughout the study than in those who received additional AEDs during follow-up (93.1% vs. 70.6%, p = 0.083). Fourth, focal aware sensory seizures were not considered when performing the analysis of seizure efficacy outcomes. It is well known that patients with focal epilepsies, even without focal seizures with impaired awareness, can experience some subtle subjective auras such as predominant viscerosensory auras in mesial temporal lobe epilepsy [31]. Therefore, it is reasonable that seizure freedom rates could be overestimated. Finally, the etiology of epilepsy was unknown in more than half of patients, and it was another limitation of the study given the real-world studies focusing on a syndromespecific efficacy of perampanel [32,33]. Despite these limitations, the strength of the current study is its long follow-up duration for patients with focal epilepsy in a real-world setting.
In conclusion, this retrospective observational study demonstrates sustained long-term efficacy and safety of perampanel as a first add-on therapy in patients with focal-onset seizures (with or without FBTCS) and aged 19 years. Seizures can be effectively controlled with low doses of perampanel (4 mg/day), which could potentially avoid adverse drug reactions.

Conflicts of interest
Ji Woong Lee is an employee of Eisai Korea Inc. All other authors declare no conflicts of interest in relation to this study.