Efficacy and safety of perampanel in patients with seizures associated with Lennox-Gastaut syndrome: A randomized trial

Abstract

Objectives: The Phase 3 Study 338 (NCT02834793) assessed long-term clinical outcomes of adjunctive perampanel in patients ≥2 years of age with uncontrolled seizures associated with Lennox-Gastaut syndrome (LGS).

Methods: Eligible patients were diagnosed with LGS and receiving one to four concomitant antiseizure medications with an average of two or more drop seizures/week during baseline. The study comprised an 18-week double-blind, randomized, placebo-controlled Core Study and ≥52-week open-label Extension. The primary endpoint was median percent change in drop seizure frequency per 28 days during the Core Study. Key secondary endpoints included responder rates, seizure-freedom rates, and safety outcomes. Post hoc analyses were performed encompassing a broader range of drop seizures or all countable motor seizures.

Results: Seventy patients were randomized into the Core Study (perampanel, n = 34; placebo, n = 36), and 58 entered the Extension. In the Core Study, numerically greater median percent reductions in drop seizure frequency were observed with perampanel (23.1%) vs placebo (4.5%) using prespecified assessments (p = .107), whereas significantly greater reductions were detected using the broader definition (48.6% vs -.7%, respectively, p = .001) or all countable motor seizures (44.0% vs -.6%, respectively, p = .017). The 50% responder rate for drop seizures was higher with perampanel vs placebo using modern definitions. Reductions in seizure frequency with perampanel were maintained over 52 weeks. Treatment-emergent adverse events occurred in 85.3% of perampanel-treated patients (somnolence [23.5%] was the most frequent) and 72.2% of placebo-treated patients.

Significance: This study had a reduced sample size and was underpowered. Although the difference in reductions in drop seizure frequency between treatments was not statistically significant by prespecified assessments, adjunctive perampanel demonstrated sustained efficacy in reducing drop seizures associated with LGS for ≤71 weeks using modern definitions. No new safety signals emerged. These observations suggest the long-term efficacy and safety of perampanel in the LGS population.

Keywords: adults; antiseizure medication; atonic; children; drop seizures; tonic.

FIGURE 1

Patient disposition. AE, adverse event. ^aOne patient passed screening but was not randomized to study treatment due to study termination by the sponsor.

FIGURE 2

(A) Median percent reduction in seizure frequency per 28 days, (B) 50% responder rate, (C) 75% responder rate, and (D) seizure‐freedom rate based on prespecified efficacy assessments throughout Study 338 (FAS; LOCF). Seizure outcomes were analyzed based on changes in seizure frequency per 28 days from baseline in patients who received ≥1 perampanel dose during the Core Study or Extension A (i.e., the LOCF method). FAS, Full Analysis Set; FOSs, focal‐onset seizures; GTCSs, generalized tonic–clonic seizures; LOCF, last observation carried forward. ^aDrop seizures were prespecified as atonic, tonic, or myoclonic seizures that led to or could have led to a fall if the patient had not been supported in the study protocol. ^bNon‐drop seizures were defined as myoclonic without fall, FOSs, GTCSs, absence, atypical absence, or clonic. ^cSeizure‐free is a patient who experienced no seizures during the analysis period and completed the period of interest.

FIGURE 3

Median percent change from pre‐randomization in drop seizure frequency per 28 days by time interval during the Core Study (FAS). FAS, Full Analysis Set. ^aDrop seizures were prespecified as atonic, tonic, or myoclonic seizures that led to or could have led to a fall if the patient had not been supported in the study protocol.

FIGURE 4

Efficacy outcomes by drop seizures or all countable motor seizures during the Core Study and Extension Phase A. (A) Median percent reduction from baseline in seizure frequency per 28 days, (B) 50% responder rates, (C) 75% responder rates, and (D) seizure‐freedom rates (FAS). FAS, Full Analysis Set; FOSs, focal‐onset seizures. ^aDrop seizures were prespecified as atonic, tonic, or myoclonic seizures that led to or could have led to a fall if the patient had not been supported in the study protocol. In accordance with the study design, statistical analyses were not conducted for secondary endpoints because the primary endpoint was not met based on the hierarchical testing. ^bThe broader definition of drop seizures encompassed atonic, tonic, tonic–clonic seizures that led or could have led to a fall. ^cAll countable motor seizures were defined as atonic, tonic, tonic–clonic, clonic, or FOSs. ^dSeizure‐free is a patient who experienced no seizures during the analysis period and completed the period of interest.