Perampanel for adjunctive treatment of partial-onset seizures: A pooled dose–response analysis of phase III studies

Authors


Summary

Objective

To better understand the relationship between efficacy and perampanel dose, integrated actual (last) dose data from three phase III trials and an extension study (blinded Conversion Period; open-label Maintenance Period) were analyzed.

Methods

Seizure frequency data were analyzed in patients who were randomized to and completed the 13-week Maintenance Period of the phase III studies on perampanel 8 mg, and who received an actual (last) dose of 12 mg during (1) the extension 16-week blinded Conversion Period or (2) weeks 1–13 of the extension Maintenance Period. Due to a treatment-by-region interaction (p = 0.042), analyses excluded patients from the Latin America region (n = 162/1,480; 10.9% of the treated cohort).

Results

Of 372 patients randomized to 8 mg in the phase III studies, 273 completed the Maintenance Period at 8 mg and 267 entered the extension study. In patients who then had an actual (last) dose of 12 mg during the extension blinded Conversion Period (n = 217), median percent change in seizure frequency per 28 days improved from −32.4% (8 mg, phase III Maintenance Period) to −44.2% (12 mg, extension blinded Conversion Period); 50% responder rates increased slightly from 37.3% to 42.9%. In patients who completed the phase III studies on 8 mg and had an actual (last) dose of 12 mg during weeks 1–13 of the extension Maintenance Period (n = 181), median percent change in seizure frequency per 28 days improved from −34.1% (phase III Maintenance Period) to −46.0% (weeks 1–13 extension Maintenance Period); 50% responder rates were 39.2% and 46.4%. Seizure control remained substantially unchanged in patients who completed the phase III studies at 12 mg and continued on that dose during the extension.

Significance

Increasing perampanel dose from 8 to 12 mg can produce additional benefits in seizure control in at least some patients who tolerate the higher dose.

The α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) subtype of ionotropic glutamate receptors has been shown to have an important role in seizure spread.[1, 2] In preclinical models, AMPA receptor antagonists have demonstrated antiseizure activity.[2] Based on this, perampanel, an orally active, selective, noncompetitive AMPA receptor antagonist, was designed and developed as an adjunctive therapy for the treatment of partial-onset seizures.

Two phase II studies in patients with refractory partial-onset seizures demonstrated that perampanel, at 2–12 mg/day, had an acceptable tolerability profile.[3] These studies provided proof of concept for the efficacy of perampanel as adjunctive therapy in this population. Pooled phase II pharmacokinetic/pharmacodynamic (PK/PD) analyses demonstrated a concentration-dependent efficacy response relative to placebo.[4] Together, these results identified the effective dose range for perampanel (4–12 mg) and guided the design of the phase III development program.

Three multinational, multicenter, randomized, double-blind, placebo-controlled phase III trials (studies 304/NCT00699972, 305/NCT00699582, and 306/NCT00700310) enrolled patients with uncontrolled partial-onset seizures despite ongoing treatment with between one and three antiepileptic drugs (AEDs).[5-7] Patients who completed the phase III studies could enroll in a long-term extension study (study 307/NCT00735397). Approximately 35% of patients in the phase III trials were receiving three concomitant AEDs. In these studies, once-daily perampanel (4–12 mg) significantly reduced seizure frequency and improved responder rates.[5-7] A dose–response for efficacy was established in study 306, which evaluated perampanel doses of 2–8 mg.[7] However, perampanel 12 mg did not consistently show an additional efficacy benefit over 8 mg in randomized dose analyses in the two studies that assessed 12 mg.[5, 6] In a pooled phase III efficacy analysis, responder rates and changes in seizure frequency were similar with perampanel 8 and 12 mg.[8] Similarly, improvements in secondarily generalized (SG) seizure frequency or SG responder rates were not consistently higher with 12 mg versus 8 mg.[9]

The apparent plateau in efficacy at higher doses was inconsistent with pooled phase III PK/PD modeling (n = 1,109), which demonstrated that seizure frequency decreased log-linearly as perampanel average exposure at steady state increased over the 2–12 mg dose range (Fig. 1).[10] Furthermore, the probability of an individual being a responder increased significantly with increasing perampanel exposure. This log-linear relationship between exposure and efficacy was independent of demographic factors and concomitant AEDs, suggesting all patients, regardless of background AED, achieve comparable efficacy benefit with increased perampanel exposure.[10]

Figure 1.

Perampanel Cavss across the phase III studies in relation to the predicted pharmacokinetics/efficacy model. Doses shown are based on the actual (last) dose achieved by patients at the end of the Maintenance Period. The percent change in seizure frequency corresponding to each actual median dose represents the model-predicted value (reproduced from Gidal et al.[10] with permission; Epilepsia © ILAE). Cavss, average exposure at steady state.

The phase III studies employed a fixed/flexible dosing regimen during the Titration Period in order to retain as many patients as possible in the study; however, a number of patients failed to reach or maintain their target dose because of intolerability. Therefore, assessments based on randomized dose could have underestimated the efficacy at higher doses, because several of these randomized patients were actually receiving lower doses.

The objective of the present work was to explore the relationship between perampanel dose and response in those patients who could tolerate the highest dose (12 mg), by analyzing integrated pooled actual (last) dose data from patients who completed the Maintenance Period of the phase III studies on perampanel 8 or 12 mg and then received 12 mg in a phase III extension study. Additional analyses assessed efficacy based on the actual (last) dose received in patients who completed studies 304 and 305 (the only phase III studies that assessed 12 mg).

Methods

Standard protocol approvals, registration, and patient consents

Study 304 was conducted between April 2008 and November 2010, at 68 centers in the United States, Canada, and Latin America.[5] Study 305 was conducted between May 2008 and January 2011, at 78 centers in Africa, Asia, Australia, Europe, and the United States.[6] Study 306 was conducted between August 2008 and May 2010, at 116 centers in Asia, Australia, and Europe.[7]

All trials were performed in accordance with the Declaration of Helsinki, ICH-E6 Guideline CPMP/ICH/135/95, European Directive 2001/83/EC, and the U.S. Code of Federal Regulations Part 21. The trial protocols, amendments, and informed consents were reviewed by national regulatory authorities in each country and independent ethics committees or institutional review boards for each site. Before trial participation, all patients gave written informed consent.

Patients

Detailed inclusion/exclusion criteria have been reported elsewhere.[5-7] At Baseline, patients were receiving stable doses of 1–3 approved AEDs.

Trial design

Phase III trials

The phase III trials were randomized, double-blind, placebo-controlled trials.[5-7] Following a 6-week Baseline Period, patients were randomized to 19 weeks (6-week Titration; 13-week Maintenance) once-daily treatment with the following: placebo, perampanel 8 or 12 mg (1:1:1; studies 304 and 305); or placebo, perampanel 2, 4 or 8 mg (1:1:1:1; study 306). Patients completing the Maintenance Period could enter a long-term extension study (study 307).[11] Patients who discontinued treatment during Maintenance or chose not to enter the extension trial entered a 4-week Follow-up Period.

During the Titration Period, perampanel doses were increased weekly in increments of 2 mg/day until the target doses were reached. Patients experiencing difficult-to-tolerate adverse events (AEs) could remain at a given dose or have their dose reduced, according to the investigator's judgment; doses could be increased when tolerability improved. During the Maintenance Period, patients continued treatment with the perampanel dose achieved during Titration. Dose changes in concomitant AEDs were not permitted during the phase III studies.

Extension study

Study 307 consisted of a 16-week blinded Conversion Period, a planned 256-week open-label Maintenance Period, and a 4-week Follow-up phase.[11] The data cutoff date for this analysis was 1 October 2011, which enabled all patients to have the opportunity to complete the 16-week Conversion Period and weeks 1–13 of the Maintenance Period. During the blinded Conversion Period, patients who had completed the Maintenance Period of the phase III trials on placebo or perampanel doses <12 mg/day had their perampanel dose blindly titrated upwards to 12 mg in 2 mg increments every 2 weeks; dose adjustment was based on tolerability. Patients who completed the phase III studies on 12 mg maintained this dose throughout the extension study Conversion and Maintenance Periods. Throughout the extension study Maintenance Period, patients remained on the dose achieved at the end of the Conversion Period, unless further adjustment was necessary for tolerability and/or efficacy reasons. Changes in concomitant AEDs (dose adjustment, discontinuation, or change in AED) were permitted during the extension study Maintenance Period at the investigator's discretion.

Endpoints

For registration in all non–European Union (EU) countries, percent change in seizure frequency per 28 days in the Double-blind phase compared with Baseline was the primary efficacy endpoint in the phase III studies. This was a secondary endpoint for registration in the EU. Within the EU, 50% responder rate (percentage of patients who experienced a ≥50% reduction in seizure frequency in the Maintenance Period relative to Baseline) was the primary efficacy endpoint. Seizure freedom rates during the whole Maintenance Period categorized percent change in seizure frequency per 28 days (75–100% response rate), percent change in SG seizure frequency in the Double-blind phase relative to Baseline, and 50% responder rates for SG seizures were exploratory efficacy endpoints. Safety assessments included recording of AEs.

Analyses

In study 304 an unusually high placebo response was observed among patients in the Latin America region,[5] and responder rates in the perampanel 8 and 12 mg groups were not significantly different versus placebo.[5] A rank analysis of covariance using data from all regions and treatment groups common to all studies (placebo and perampanel 8 mg) showed a significant (p = 0.042) treatment-by-region interaction. A larger decrease in percent change in seizure frequency per 28 days was observed in the perampanel 8 mg group versus placebo in all regions except Latin America.[12] Hence, patients enrolled in Latin America were excluded from the pooled efficacy analysis. The basis for the regional difference in placebo rate is unexplained.[5]

Efficacy analyses based on actual (last) dose increases from 8 mg in the phase III studies to 12 mg in the extension study

Percent change in seizure frequency and responder rate were analyzed for the two different subsets of patients who were randomized to perampanel 8 mg in the phase III studies, completed the Maintenance Period of the phase III studies on perampanel 8 mg and who: (1) had an actual (last) dose of 12 mg during the Conversion Period of the extension study (Fig. 2) or (2) had an actual (last) dose of 12 mg during weeks 1–13 of the extension study Maintenance Period (Fig. 2). This interval equals the duration of the Maintenance Period in the phase III studies, and was chosen to minimize the impact of comparing different exposure lengths.

Figure 2.

Patient flow for dose–response analysis based on integrated data from patients who completed the phase III study Maintenance Period on perampanel 8 or 12 mg and who had an actual (last) dose of 12 mg during the extension study blinded Conversion Period or weeks 1–13 of the Maintenance Period.

Similar analyses were conducted for patients who were randomized to and completed the phase III studies on 12 mg and who had an actual (last) dose of 12 mg during (1) the extension study blinded Conversion Period or (2) weeks 1–13 of the extension study Maintenance Period.

Efficacy analyses based on actual (last) dose in patients completing the Maintenance Period of studies 304 and 305

These analyses were performed in patients who completed the Maintenance Period of the phase III studies on an actual (last) dose of perampanel 8 or 12 mg, regardless of randomized dose. Percent change in seizure frequency, 50% responder rate, 75% response rate, and seizure freedom rates were analyzed. Percent change in seizure frequency and 50% responder rate for SG seizures were also assessed.

Statistical analysis

Although the efficacy endpoints assessed were prespecified in the phase III studies, the analyses presented here using actual dose (last dose in the analysis periods) for integrated phase III data were all ad hoc analyses. Therefore, all efficacy endpoints for these analyses were summarized using descriptive statistics. No inferential (statistical) analyses were performed.

Results

Patient disposition

Overall, 1,480 patients (study 304, n = 388; study 305, n = 386; and study 306, n = 706) were enrolled and treated in the three phase III trials (Fig. 2). Of these, 442, 180, 172, 431, and 255 patients were randomized to placebo or perampanel 2, 4, 8, and 12 mg, respectively. In total, 1,264 patients (85.4%) completed the phase III trials. The proportion of patients in each randomized dose group who discontinued the trials prematurely was: 11.3%, placebo; 14.4%, perampanel 2 mg; 8.1%, 4 mg; 14.8%, 8 mg; and 24.3%, 12 mg.[8] Of these, 3.8%, 5.6%, 2.9%, 7.2%, and 18.4% of patients randomized to placebo, perampanel 2, 4, 8, and 12 mg, respectively, gave AEs as the primary reason for discontinuation. The pooled intent-to-treat analysis set included 1,478 patients. Study 304 enrolled 162 patients from the Latin America region; all efficacy analyses below excluded these patients.

The numbers of patients who completed the Maintenance Period of the phase III studies at each actual (last) dose (excluding Latin American patients and irrespective of the target dose) were 348, 161, 159, 46, 287, 14, and 114 for placebo, perampanel 2, 4, 6, 8, 10, and 12 mg, respectively.

Efficacy

Increasing perampanel actual (last) dose from 8 mg in the phase III studies to 12 mg in the extension study

Excluding Latin American patients, a total of 372 patients were randomized to 8 mg in the phase III studies; 273 completed the Maintenance Period at the target dose (Fig. 2). Of these, 267 patients (97.8%) continued into the extension study and 217 had an actual (last) dose of perampanel 12 mg during the 16-week blinded Conversion Period. In these patients, the median percent change in seizure frequency per 28 days versus Baseline improved when perampanel dose was titrated from 8 mg in the phase III studies’ Maintenance Period (−32.4%) to 12 mg in the extension study’s blinded Conversion Period (−44.2%) (Fig. 3A). Fifty percent responder rates increased slightly from 37.3% to 42.9% in these patients (Fig. 3B).

Figure 3.

(A) Median percent change in seizure frequency per 28 days (treatment versus Baseline) and (B) 50% responder rates (patients experiencing ≥50% reduction in seizure frequency per 28 days) for patients who were randomized to perampanel 8 or 12 mg in the phase III studies (studies 304, 305 and 306), completed the Maintenance Period of the three phase III on the target dose, and who had an actual (last) dose of perampanel 12 mg during the extension study blinded Conversion Period or weeks 1–13 of the extension study Maintenance Period (actual [last] dose analysis, excluding patients from Latin America region).

In patients who completed the phase III Maintenance Period on perampanel 12 mg, and whose actual (last) dose was 12 mg during the blinded Conversion Period (n = 103), the median percent change in seizure frequency was virtually unchanged between the phase III Maintenance Period (−34.9%) and the extension study Conversion Period (−36.8%) (Fig. 3A). Fifty percent responder rates were also comparable during the phase III studies Maintenance Period (40.8%) and the extension study Conversion Period (42.7%) in patients who remained on perampanel 12 mg.

Actual (last) dose/seizure frequency relationships were also examined for patients who continued into the first 13 weeks of the extension study Maintenance Period. A similar outcome was noted, in that patients who completed the phase III studies on perampanel 8 mg and whose actual (last) dose was 12 mg during weeks 1–13 of the extension Maintenance Period (n = 181) showed an improvement in median percent change in seizure frequency from −34.1% (8 mg phase III studies) to −46.0% (12 mg weeks 1–13 of the extension study Maintenance Period) (Fig. 3A). Fifty percent responder rates improved from 39.2% to 46.4% in these patients (Fig. 3B).

In patients who completed the phase III studies on perampanel 12 mg and had an actual (last) dose of 12 mg during weeks 1–13 of the extension Maintenance Period (n = 81), improvements in seizure frequency and 50% responder rates were maintained. Specifically, median percent changes in seizure frequency were −41.9% (phase III studies) and −45.3% (weeks 1–13 extension Maintenance Period) and 50% responder rates were 43.2% and 45.7%, respectively (Fig. 3A, B).

Actual (last) dose analysis of patients completing phase III studies 304 and 305

Including Latin America, 774 patients were randomized and treated in studies 304 and 305; 257 patients were randomized to placebo, 262 to perampanel 8 mg, and 255 to perampanel 12 mg. Excluding Latin America (and irrespective of randomized dose), there were 454 patients who completed the Maintenance Period with an actual (last) dose of perampanel 8 mg (n = 158), perampanel 12 mg (n = 114), or placebo (n = 182).

Perampanel doses of 8 and 12 mg were associated with lower seizure frequencies and higher 50% responder rates compared with placebo (Fig. 4A,B). Median percent change in seizure frequency, 50%, and 75% response rates were similar in patients who completed the studies on an actual (last) dose of perampanel 12 mg versus 8 mg (Fig. 4A, B; Table 1). Concomitant treatment with perampanel metabolism–inducing AEDs, which may reduce perampanel exposure, is one potential confounding factor that may contribute to the lack of additional benefit with perampanel 12 mg versus 8 mg in these analyses. A higher proportion of patients who completed studies 304 or 305 with an actual (last) dose of 12 mg were receiving concomitant treatment with at least one perampanel metabolism-inducing AED (defined as carbamazepine, oxcarbazepine or phenytoin[13]) versus patients who completed with an actual (last) dose of 8 mg (69.3% vs. 59.5%, respectively; Table 1).

Table 1. Proportion of patients who were receiving ≥1 perampanel metabolism–inducing AED at Baseline, proportion of patients who achieved a ≥75% reduction in seizure frequency per 28 days (Maintenance Period vs. Baseline), and seizure freedom rates for patients who completed the Maintenance Period of studies 304 and 305, which included perampanel doses of 8 and 12 mg (actual [last] dose analysis, excluding patients from Latin America region)
 Placebo n = 182Perampanel 8 mg n = 158Perampanel 12 mg n = 114
  1. AED, antiepileptic drug.

  2. a

    Defined as carbamazepine, oxcarbazepine, or phenytoin.

Patients receiving ≥1 perampanel metabolism–inducing AEDa, n (%)102 (56.0)94 (59.5)79 (69.3)
75% Responder rate (%)3.318.420.2
Seizure-free status during entire Maintenance Period, n (%)2 (1.1)4 (2.5)5 (4.4)
Figure 4.

(A) Median percent change in partial-onset seizure frequency and SG seizure frequency per 28 days (Double-blind phase vs. Baseline) and (B) 50% responder rates for all partial-onset seizures and SG seizures (patients experiencing ≥50% reduction in seizure frequency per 28 days during the Maintenance Period vs. Baseline) for patients who completed the Maintenance Period of studies 304 or 305 on perampanel 8 mg or perampanel 12 mg (actual [last] dose analysis, excluding patients from Latin America region). SG, secondarily generalized.

Seizure freedom rates over the entire Maintenance Period were 2.5% for patients completing the Maintenance Period of studies 304 or 305 with an actual (last) dose of perampanel 8 mg and 4.4% in patients completing with an actual (last) dose of 12 mg (Table 1).

There was a slightly greater incremental benefit for perampanel 12 mg versus 8 mg in median percent change in the frequency of SG seizures (−46.0% with 8 mg versus −54.0% with 12 mg; Fig. 4A). A similar difference was found for 50% responder rates for SG seizures, which were slightly greater in patients who completed studies 304 and 305 on perampanel 12 mg versus 8 mg (46.8% with 8 mg vs. 53.5% with 12 mg; Fig. 4B).

Safety

The incidence of treatment-emergent AEs (TEAEs) throughout the phase III Double-blind treatment phase in the pooled safety analysis set, based on the actual dose patients were taking at onset, is shown in Table 2. Actual dose at onset refers to the dose that the patient was taking at onset of the TEAE. Patients who had the same TEAEs occurring with different doses were counted once within each dose group. The denominator for the incidence rate was the number of subjects who were exposed to the doses within each dose group.

Table 2. Proportion of patients who were receiving ≥1 perampanel metabolism-inducing AED at Baseline and incidence of TEAEs by actual dose at onset (pooled phase III safety population)
 Patients, n (%)
Placebo (n = 442)Perampanel
<4 mg (n = 1,038)4 mg (n = 846)>4–8 mg (n = 667)>8–12 mg (n = 212)Total (n = 1,038)
  1. AED, antiepileptic drug; TEAE, treatment-emergent adverse event.

  2. A TEAE is defined as an AE that either begins on or after the first dose of study drug and up to 30 days after the last dose, or increases in severity during treatment. Actual dose at onset refers to the dose the patient was taking at onset of the TEAE. Patients who had the same TEAEs that occurred with different doses were counted once within each dose group. The denominator for the incidence rate was the number of subjects who were exposed to the doses within each dose group.

  3. a

    Defined as carbamazepine, oxcarbazepine, or phenytoin.

Patients receiving ≥1 perampanel metabolism inducing AEDa166 (37.6)204 (19.7)127 (15.0)245 (36.7)101 (47.6)461 (44.4)
Any TEAE294 (66.5)354 (34.1)239 (28.3)444 (66.6)161 (75.9)799 (77.0)
TEAEs in ≥5% of perampanel-treated patients      
Dizziness40 (9.0)63 (6.1)53 (6.3)165 (24.7)52 (24.5)292 (28.1)
Somnolence32 (7.2)58 (5.6)25 (3.0)62 (9.3)13 (6.1)150 (14.5)
Headache50 (11.3)40 (3.9)25 (3.0)47 (7.0)15 (7.1)118 (11.4)
Fatigue20 (4.5)35 (3.4)16 (1.9)32 (4.8)9 (4.2)88 (8.5)
Irritability13 (2.9)20 (1.9)11 (1.3)29 (4.3)16 (7.5)73 (7.0)
Nausea20 (4.5)13 (1.3)10 (1.2)25 (3.7)10 (4.7)54 (5.2)
Fall15 (3.4)4 (0.4)9 (1.1)30 (4.5)15 (7.1)53 (5.1)

Overall, the incidence of TEAEs was 66.5% in patients who received placebo and 77.0% in perampanel-treated patients. The most common TEAEs (occurring in ≥5% of perampanel-treated patients) by actual dose at onset are shown in Table 2. Among these common AEs, the onset of dizziness, irritability, and fall increased with increasing dose but only irritability and fall had ≥2% greater frequency with perampanel 10–12 mg (n = 212) compared with perampanel 6–8 mg (n = 667) (Table 2). At Baseline, the proportion of patients receiving concomitant treatment with ≥1 AED that induced perampanel metabolism was higher in patients who had TEAEs with onset at perampanel 10–12 mg versus 6–8 mg (Table 2).

Discussion

In the phase III studies, perampanel 12 mg was consistently superior to placebo but did not consistently show an additional benefit over 8 mg in standard randomized dose analyses.[5, 6] This was in contrast to pooled phase III PK/PD analyses (including Latin American patients), which demonstrated that seizure frequency decreased log-linearly with increasing average serum perampanel concentrations over the entire 2–12 mg dosing range.[10]

Efficacy analyses in patients who completed the Maintenance Period of the phase III studies on perampanel 8 mg and had an actual (last) dose of 12 mg during either assessed periods of the extension study (16-week blinded Conversion Period or weeks 1–13 of the Maintenance Period) were consistent in showing improved benefits with the higher dose. In particular, compared with the phase III studies, the median reduction in seizure frequency improved by 36% when doses were increased from 8 mg (phase III) to 12 mg during the extension study blinded Conversion Period (−32.4% vs. −44.2%) and by 35% during weeks 1–13 of the extension study Maintenance Period (−34.1% vs. −46.0%). These improved outcomes following perampanel dose increases from 8 to 12 mg in the same patient were associated with a less pronounced increase in 50% responder rates. In contrast, patients who remained on the 12 mg dose during both the phase III studies and the extension study had generally maintained outcomes throughout the assessment periods.

Further analyses based on the actual (last) dose received by patients completing studies 304 and 305 showed no clear difference between perampanel 12 and 8 mg in terms of reduction in overall seizure frequency and response rates. The reason that these comparisons failed to show differences in overall seizure reduction and response rates comparable to those observed in the extension study are unclear. Pooled phase III PK/PD modeling has shown that perampanel exposure is lower in patients receiving concomitant perampanel metabolism–inducing AEDs (Fig. 1), but that the exposure/efficacy relationship is maintained irrespective of AED use.[10] One possible explanation may be that perampanel exposure in patients with an actual (last) dose of 12 mg in the phase III studies could be comparable to exposure with lower doses, as a larger proportion of patients with an actual (last) dose of 12 mg were receiving concomitant AEDs that induce perampanel clearance versus patients with an actual (last) dose of 8 mg. This could potentially dampen the impact of increasing perampanel dose when comparing efficacy across different patient populations.

Discontinuation rates in the pooled phase III safety analysis set were higher in patients randomized to perampanel 8 mg versus 12 mg, and more patients randomized to 12 mg discontinued due to AEs. This is consistent with 12 mg being the maximum tolerated dose. A PK/PD analysis of the phase III trials has shown that there is an increased probability of fatigue, somnolence, gait disturbances, dizziness, weight increase, irritability, dysarthria, and euphoric mood with increasing perampanel exposure.[10] With randomized dose analyses, certain AEs appeared to be considerably more frequent with perampanel 12 mg than 8 mg in the individual phase III studies (dizziness and fatigue),[5, 6] and in the pooled phase III analyses (dizziness, somnolence, fatigue, irritability, nausea and fall).[8] In contrast, when the same AEs were analyzed by actual dose at onset, a less clear-cut dose–response relationship was demonstrated. In fact, only irritability and fall, both of which occurred in <10% of perampanel-treated patients overall, had a clearly greater frequency with perampanel doses >8 mg compared with 4–8 mg. More patients who experienced the onset of AEs at perampanel 10–12 mg were receiving concomitant AEDs that induce perampanel clearance versus patients who developed AEs at 6–8 mg. The flattening of the dose–AE relationship when assessed based on actual dose at onset may reflect lower perampanel exposure in patients titrated to the higher doses.

As a consequence of a by-region subgroup analysis of the pooled phase III studies, which detected a significant treatment-by-region interaction due to a higher placebo response in the Latin America region,[12] the analyses reported here excluded patients enrolled in Latin America. The reasons for the high placebo response are unknown. There were some regional differences in demographic and baseline characteristics; however, none unique to Latin America. Furthermore, the relationship between perampanel plasma concentration and efficacy did not differ as a function of region,[10] suggesting that the degree of improvement in seizure control with perampanel is consistent across all geographic regions.

A number of interrelated factors intrinsic to study design can affect the assessment of dose–response relationships in AED trials. The perampanel phase III trials employed a fixed target-dose design with added flexibility to permit adjustments based on tolerability during the Titration Period. These trials were primarily designed to compare the efficacy of perampanel versus placebo, but not to discriminate between perampanel doses. When a significant proportion of patients do not achieve the target dose due to intolerability, an analysis based on randomized rather than actual (last) dose may appreciably underestimate the efficacy associated with the highest doses. Restricting assessment of dose–response relationships to those patients who actually reached and completed treatment on the target dose, as in the present analysis, may allow a more precise definition of efficacy at each individual dose. However, this analysis also has limitations in that the results are applicable to only those patients who tolerated the randomized dose. This analysis, or interpretation, could be confounded by (1) exclusion of patients who discontinued prematurely for other reasons and (2) differences in exposure to interacting co-medications across groups. The confounding effects of excluding patients who discontinued prematurely may apply not only to the comparison of patients completing the phase III trials on their randomized dose, but also to the assessment of patients who had their dosage increased from 8 to 12 mg during the extension study. The latter offers the advantage of within-patient assessment and comparison with a parallel group of patients who remained at an unchanged 12 mg dose during at the end of the phase III studies and during the extension study. Patients who remained on the 12 mg dose had generally unchanged outcomes in the extension study versus the phase III study Maintenance Periods, suggesting that benefits in efficacy observed in patients who titrated from perampanel 8 mg in the phase III studies to 12 mg in either assessment period of the extension study were not a consequence of increased treatment duration. However, results should be interpreted with some caution because of the lack of a control group continuing on 8 mg during the extension trial. Moreover, although in the extension study both patients and investigators remained blinded to prior treatment, outcomes could have been influenced by the knowledge that all patients were being titrated to their highest tolerable dose. Although changes in concomitant AEDs during the extension study are another potential confounding factor, it is worth noting that no changes in concomitant AEDs were permitted during the blinded Conversion Period of the extension study and there were only minimal changes during weeks 1–13 of the Maintenance Period.

Despite the limitations discussed earlier, and consistent with results from pooled phase III PK/PD analyses, within-patient efficacy analyses suggest that a perampanel dose of 12 mg, although associated with higher overall rates of discontinuation due to AEs, may produce additional therapeutic benefit in some patients who do not fully respond to doses up to 8 mg and who tolerate the higher dose.

Acknowledgments

This study was funded by Eisai, Inc. Editorial support was provided by Deborah McGregor, Ph.D., of Complete Medical Communications and was funded by Eisai Inc.

Conflict of Interest

L. Kramer is an employee of Eisai Inc. A. Satlin is an employee of Eisai Inc. G. L. Krauss is currently an investigator for Eisai, UCB Pharma, Neuronex, Sunovion, and National Institutes of Health/National Institute on Aging (NIH/NIA). He is a consultant for Eisai as a member of the Epilepsy Study Consortium. J. French serves as the President of The Epilepsy Consortium, a nonprofit organization. She receives 25% salary support from The Epilepsy Study Consortium. The consortium receives funding from Eisai Inc. for work performed (National PI, advisory board, FDA submission, consulting) by Dr French on behalf of The Epilepsy Study Consortium, as well as many other companies with AEDs in development or marketed AEDs. E. Perucca received research grants from the European Union, the Italian Medicines Agency, the Italian Ministry of Health, and the Italian Ministry for Education, University and Research. He also received speaker's or consultancy fees and/or research grants from Bial, Eisai, GlaxoSmithKline, Johnson & Johnson, Medichem, Novartis, Pfizer, Supernus, UCB Pharma, Upsher-Smith, Valeant, Vertex, and Viropharma. E. Ben-Menachem is currently an investigator for Eisai and UCB Pharma, and is a consultant for UCB Pharma, Eisai, Janssen-Cilag, Biocontrol, and Lundbeck. She has received a research grant from Västra Götalands Region and is chief editor of Acta Neurologica Scandinavica. P. Kwan's institution received research grants from the NIH, Hong Kong Research Grants Council, Innovation and Technology Fund, and Health and Health Services Research Fund. He/his institution also received speaker's or consultancy fees and/or research grants from Eisai, GlaxoSmithKline, Johnson & Johnson, Pfizer, and UCB Pharma. J. J. Shih received research grants from the NIH, U.S. National Science Foundation, and the Mayo Clinic Foundation. In the past 5 years, he has also received research funding from Eisai, UCB Pharma, and GlaxoSmithKline. A. Laurenza is an employee of Eisai Inc. H. Yang is an employee of Eisai Inc. J. Zhu is an employee of Eisai Inc. D. Squillacote is a former employee of Eisai Inc. 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.

Biography

  • Image of creator

    Lynn D. Kramer is Chief Clinical Officer of Eisai Company Ltd., and President of the Neuroscience & General Medicine PCU, Eisai Product Creation Systems, Eisai Inc.

Ancillary