A randomized, double-blind, placebo-controlled, parallel-group study of rufinamide as adjunctive therapy for refractory partial-onset seizures

Summary

Purpose:  Efficacy and safety of adjunctive rufinamide (3,200 mg/day) was assessed in adolescents and adults with inadequately controlled partial-onset seizures receiving maintenance therapy with up to three antiepileptic drugs (AEDs).

Methods:  This randomized, double-blind, placebo-controlled, parallel-group, multicenter study comprised a 56-day baseline phase (BP), 12-day titration phase, and 84-day maintenance phase (MP). The primary efficacy variable was percentage change in total partial seizure frequency per 28 days (MP vs. BP). Secondary efficacy outcome measures included ≥50% responder rate and reduction in mean total partial seizure frequency during the MP. Safety and tolerability evaluation included adverse events (AEs), physical and neurologic examinations, and laboratory values. Pharmacokinetic and pharmacodynamic assessments were conducted.

Results:  Three hundred fifty-seven patients were randomized: 176 to rufinamide and 181 to placebo. Patients had a median of 13.3 seizures per 28 days during BP; 86% were receiving ≥2 AEDs. For the intent-to-treat population, the median percentage reduction in total partial seizure frequency per 28 days was 23.25 for rufinamide versus 9.80 for placebo (p = 0.007). Rufinamide-treated patients were more than twice as likely to have had a ≥50% reduction in partial seizure frequency (32.5% vs. 14.3%; p < 0.001) and had a greater reduction in median total partial seizure rate per 28 days during the MP (13.2 vs. 5.2; p < 0.001). Treatment-emergent AEs occurring at ≥5% higher incidence in the rufinamide group compared with placebo were dizziness, fatigue, nausea, somnolence, and diplopia.

Conclusions:  Adjunctive treatment with rufinamide reduced total partial seizures in refractory patients. AEs reported were consistent with the known tolerability profile of rufinamide.

Despite an expanding number of antiepileptic drugs (AEDs) available to treat partial-onset epilepsy, about one-third of patients with epilepsy remain resistant to available treatments (Perucca et al., 2007). In addition, intolerable side effects and/or idiosyncratic reactions can lead to discontinuation of AEDs (Kwan & Brodie, 2000). Therefore, additional effective and well-tolerated AEDs are needed.

Rufinamide is an AED with a novel triazole-derivative structure. It limits sodium-dependent action potentials in neuronal models with a membrane-stabilizing effect (Hakimian et al., 2007). Rufinamide is approved for the adjunctive treatment of seizures associated with Lennox-Gastaut syndrome in patients aged ≥4 years in the United States (Banzel, 2008) and in some countries in Europe (Inovelon, 2007). Previous studies have evaluated rufinamide in doses up to 3,200 mg/day as adjunctive treatment of partial-onset seizures and have shown efficacy and good tolerability (Palhagen et al., 2001; Brodie et al., 2009; Elger et al., 2010). Of particular interest is the absence of associated serious cognitive side effects (Aldenkamp & Alpherts, 2006).

This study was conducted to evaluate and confirm the efficacy and safety of rufinamide at a dose of 1,600 mg twice daily as adjunctive treatment for refractory partial-onset seizures in a population similar to that previously studied (Brodie et al., 2009).

Methods

Study

This was a randomized, double-blind, placebo-controlled, parallel-group study in adolescent and adult patients with inadequately controlled partial-onset seizures, treated with a stable dose of up to three AEDs. Partial-onset seizures included simple or complex partial seizures, with or without secondary generalization. The study comprised three phases:

• 1. Baseline/screening phase (BP, 56 days): Patients or caregivers completed a patient seizure diary to establish the baseline types and frequency of seizures. Those who had ≥6 seizures during the 56 days, with no 21-day seizure-free periods, were eligible for randomization into the double-blind phase of the study. Simple partial seizures without motor signs were not included in determining this criterion.

• 2. Titration phase (12 days): Eligible patients were randomized to receive either rufinamide or placebo in a 1:1 ratio using a computer-generated schedule supplied by the sponsor, with randomization done in blocks of four. Treatment assignment was unknown to investigators, clinic and sponsor personnel, and patients.

• 3. The total daily dose (TDD) was taken orally, half in the morning, and half in the evening, starting at a TDD of 800 mg, which was increased every 3 days in 400 mg twice-daily increments to the target TDD of 3,200 mg. Patients unable to tolerate the target dose, regardless of whether they were taking active drug or placebo, could have their dose reduced to a TDD of 2,400 mg (during titration phase only), which they could remain on for the maintenance phase (MP) of the study. Patients attended clinic on days 1 and 12, when concomitant medications, adverse events (AEs; a seizure was recorded as an AE if it was a new type of seizure or greater in intensity or frequency than those previously experienced), and laboratory values were checked. Physical and neurologic examinations were performed. On day 1, electrocardiography (ECG) was obtained and blood was taken before medications were administered to evaluate plasma AED levels.

Maintenance phase (MP, 84 days). Patients continued to receive the dose achieved during the end of titration phase for 12 weeks. Patients attended clinics on days 40, 68, and 96 when similar evaluations were made as during the titration phase visits; and an additional 12-lead ECG was obtained on day 96. The Clinical Global Impression of Change (GIC), completed by investigators during the last patient visit, compared the patient’s clinical status then with baseline. The investigators used a seven-point Likert-type scale ranging from “very much improved” (1) to “very much worse” (7), which incorporated consideration of seizure frequency and severity, AEs, and overall functional status of the patient. The Patient GIC was self-reported on the same seven-point scale as used for the Clinical GIC; patients assessed their status at the end of the study by responding to the question: “How have you felt compared with before you entered this clinical trial?”

Patients

Eligible patients were male or female, aged 12–80 years, with partial-onset seizures with or without secondarily generalized seizures, as defined by the International League Against Epilepsy (ILAE) Classification of Epileptic Seizures (International League Against Epilepsy, 1981). Diagnosis was established by clinical history, electroencephalography, and computed tomography (CT)/magnetic resonance imaging (MRI) of the brain performed within the last 10 years. Patients’ seizures were inadequately controlled on stable doses of up to three concomitantly administered AEDs, with no evidence of AED treatment noncompliance. All medication taken regularly by patients, including AEDs, remained unchanged for at least 1 month prior to study start and throughout the study. Female patients capable of becoming pregnant were required to take steps to prevent pregnancy. Vagus nerve stimulators had to have been implanted for at least 6 months before randomization, and stimulator parameters had to have been unchanged for at least 1 month prior to screening and for the duration of the study. For the purposes of study inclusion criteria, vagus nerve stimulators were not counted as an AED. Magnet use was allowed and was documented throughout the study.

Patients were excluded from the study if they had generalized epilepsies or a history of status epilepticus or seizure clusters (i.e., individual seizures could not be counted) in the past year, or if they required felbamate, vigabatrin, or rescue benzodiazepines (more than once a month). Patients were also excluded if they had clinically significant medical or psychiatric disease, clinically significant ECG abnormality, or a diagnosis of congenital short QT syndrome, psychogenic seizures in the previous year, a history of drug abuse and/or positive finding on urinary drug screening, or a history of alcohol abuse in the past 2 years.

Ethical approval

This study was conducted in compliance with the Declaration of Helsinki (revised edition, October 1996) and good clinical practice for trials on medicinal products. Approval for the study protocol and informed consent form was obtained from the Western Institutional Review Board (a centralized IRB) for sites in the United States or from independent ethics committees for sites in Canada.

Written informed consent from each participant in the study was obtained before screening.

Efficacy evaluation

Primary efficacy variable

The primary efficacy variable was the percent change in the total partial seizure frequency per 28 days during the double-blind MP relative to the BP.

Data were analyzed for the intent-to-treat (ITT) population, which was defined as all randomized patients who had baseline patient seizure diary data and had completed at least the titration phase. A sensitivity analysis was performed using the per protocol (PP) population, which comprised all randomized patients who completed the MP, had no major protocol deviations, and had at least 80% study drug compliance.

Secondary efficacy variables

Secondary efficacy variables, all measured per 28 days during the MP relative to the BP, were the number of patients who had at least a 50% reduction in partial seizures (“responders”), the total partial seizure frequency, and the reduction in total partial seizure frequency (RRATIO), which was defined as 100 × (T−B)/(T+B), where T was the total partial seizure frequency per 28 days during the MP and B was the total partial seizure frequency per 28 days during the BP.

Exploratory efficacy variables

The exploratory variables, measured per 28 days during MP relative to BP, were the percent change in frequency of each partial seizure type (simple partial, simple partial with motor signs, simple partial without motor signs, complex partial, and secondarily generalized); the change in the number of seizure-free days; the responder rate (≥75% reduction in partial seizure frequency); Clinical and Patient GIC data with additional analysis by collapsing the seven possible scores into three categories [improved (scores of 1, 2, and 3), no change (score of 4), and worsening (scores of 5, 6, and 7]; and the difference between the numbers of patients with either a ≥25% or ≥100% increase in partial seizure frequency.

Safety evaluation

Safety variables included treatment-emergent AEs (TEAEs), laboratory evaluations, vital signs, ECGs, and physical and neurologic examinations. AEs were considered to be TEAEs if they started on or after the date of administration of the first dose of study drug or if they were present prior to the administration of the first dose of study drug and increased in severity during the study. A seizure was recorded as an AE if it was a new type of seizure or of greater intensity or frequency than those experienced by the patient prior to this study. Serious AEs (SAEs) were defined as any untoward medical occurrence that resulted in death, was life-threatening, required inpatient hospitalization or prolongation of existing hospitalization, or resulted in persistent or significant disability/incapacity. Laboratory assessments included hematology, clinical chemistry, urinalysis, and urine drug screen.

Statistical methods

The sample size estimation was based on data from a previous clinical study for adjunctive rufinamide (1,600 mg twice daily) in refractory partial seizures. A sample size of 162 patients in each group was estimated to have 80% power to detect a difference in means of 0.104 (corresponding to a ratio of 1.27 or a 27% difference) assuming that the common standard deviation is 0.333 using a two-group, two-tailed t-test with an α level of 0.05.

For the primary efficacy variable, the treatment groups were compared using a Wilcoxon rank-sum test. Covariates such as baseline seizure frequency, age, sex, and race were tested to determine their association with the percentage change in seizure frequency using an analysis of covariance (ANCOVA) model with the ranks of the percentage change as the response variable and ranked seizure frequency per 28 days during the BP as a covariate.

The responder analyses (≥50% and ≥75% reduction in partial seizures) were tested using a two-sided, chi square test with 1 d.f. For the other secondary end points, the total partial seizure frequency was calculated and transformed (log₁₀, thus circumventing results that were expected to be not normally distributed). Treatment groups were compared using an ANCOVA model with treatment as a factor and 28-day total partial seizure frequency in the BP as a covariate. For the reduction in total partial seizure frequency, the treatment groups were compared using a two-sided t-test, with a 5% significance level (α = 0.05).

The statistical tests used for the exploratory end points were the Wilcoxon rank-sum test and ANCOVA for the change in frequency of each partial seizure type; the Wilcoxon rank-sum test for the change in the number of seizure-free days per 28 days; the Mantel-Haenszel procedure for Clinical and Patient GIC data, which were analyzed using a modified RIDITS with the score option in the computation of row mean scores; and treatment-group row mean scores compared using a chi square test. An additional analysis was conducted by collapsing the seven possible scores into three categories [improved (scores of 1, 2, and 3), no change (score of 4), and worsening (scores of 5, 6, and 7)] using the same test statistic; and the difference between the numbers of patients with either a ≥25% or ≥100% increase in partial seizure frequency was compared between rufinamide and placebo group patients using the Fisher’s exact test.

Population pharmacokinetic (PK) and PK/pharmacodynamic (PD) analyses were based on multiple regression analyses using nonlinear mixed-effect models.

Safety data that were continuous variables were summarized using standard summary statistics, and those that were categorical variables were summarized by frequency and percentage.

Results

Patients

Patients were enrolled at 61 centers in the United States and at four centers in Canada between February 2006 and March 2009. In total, 357 patients were randomly assigned to receive rufinamide (n = 176) or placebo (n = 181) and entered the titration phase, and 139 and 156 patients, respectively, completed the study (Fig. 1).

Figure 1.   Disposition of patients. ^aOne patient was withdrawn from the study before receiving study drug because required laboratory assessments were not obtained.

The safety, ITT, and PP populations comprised 356, 335, and 272 randomized patients, respectively. Three hundred five patients (rufinamide, 138; placebo, 167) from the ITT population were included in the PK/PD population.

Among the ITT population, 120 patients receiving rufinamide entered the MP on a dose of 3,200 mg/day, whereas 41 did so at a dose of 2,400 mg/day. Study drug was taken for at least 12 weeks during the study by 79.0% of the rufinamide-treated patients and 85.0% of the placebo patients.

The demographic and baseline characteristics of the treatment groups were similar (Table 1). Patients in the rufinamide group had a median of 13.00 seizures per 28 days during the BP compared with 13.80 in the placebo group. The most common seizure types were complex partial seizures (86.2%) and complex partial seizures with secondary generalization (72.5%), and the distribution of patients was comparable between the two treatment groups.

Table 1.   Demographic and baseline disease characteristics (safety population)

	Rufinamide (n = 176)	Placebo (n = 180)	Total (n = 356)
^aOther includes Native American, Asian Pacific Islander, and other. ^b28-Day total partial seizure frequency during the BP. ^cSGS, secondarily generalized seizures.
Sex, male, n (%)	84 (47.7)	83 (46.1)	167 (46.9)
Age, years
Mean (SD)	36.4 (14.77)	38.1 (14.84)	37.3 (14.81)
Range	12–77	12–75	12–77
12–<18, n (%)	15 (8.5)	21 (11.7)	36 (10.1)
18–<65, n (%)	155 (88.1)	153 (85.0)	308 (86.5)
≥65, n (%)	6 (3.4)	6 (3.3)	12 (3.4)
Race, n (%)
Black	14 (8.0)	19 (10.6)	33 (9.3)
White	145 (82.4)	140 (77.8)	285 (80.1)
Hispanic	13 (7.4)	14 (7.8)	27 (7.6)
Other^a	4 (2.2)	7 (3.8)	11 (3.0)
Mean weight (SD), kg	77.4 (22.36)	79.0 (23.36)	78.2 (22.84)
Range	29.9–134.1	32.7–140.0	29.9–140.0
Median number of seizures^b	13.0	13.8	13.3
Range	2.1–864.6	2.8–284.5	2.1–864.6
SGS^c
Yes, n (%)	76 (43.2)	65 (36.1)	141 (39.6)
No, n (%)	99 (56.3)	114 (63.3)	213 (59.8)

Of the 356 patients in the safety population, 49 (13.8%) received one AED, 157 (44.1%) received two AEDs, 124 (34.8%) received three AEDs, and 26 (7.3%) received four AEDs during the BP (the patients on four AEDs were included in the ITT, but not the PP population). The type and number of AEDs used at baseline were comparable in the two treatment groups. During the BP, the AEDs taken by >20% of the patients were levetiracetam (39.0%), lamotrigine (36.0%), topiramate (21.9%), and carbamazepine (20.2%).

Primary efficacy variable

In the ITT population, rufinamide-treated patients had a median percent reduction from baseline in total partial seizure frequency per 28 days of 23.3% compared with 9.8% for placebo treatment (p = 0.007) (Fig. 2). ANCOVA showed no association of this result with baseline seizure frequency, age, sex, or race. PP analysis concurred with the magnitude and direction of the ITT result (25.4% vs. 10.5%; p = 0.005) (Fig. 2).

Figure 2.   Median reduction in total partial seizure frequency from baseline (Wilcoxon rank-sum test).

Secondary efficacy variables

Significantly more rufinamide-treated patients (32.5%) than placebo-treated patients (14.3%) had at least a 50% reduction in partial seizure frequency per 28 days during the MP relative to the BP [odds ratio (OR) 2.9; 95% confidence interval (CI) 1.7–4.9; p < 0.001] (Table 2). Similar results were seen in the PP population (33.6% vs. 13.9%; OR 3.1; 95% CI 1.7–5.7; p < 0.001). There was no effect of the covariates of race, sex, or age. The median reduction in the total partial seizure frequency rate per 28 days (MP vs. BP) was 13.2% in the rufinamide group and 5.2% in the placebo group (p < 0.001, Table 2). A statistically significant reduction favoring rufinamide was also seen for the PP population.

Table 2.   Results from secondary and exploratory efficacy variables, per 28 days during the MP relative to the BP (ITT population)

	Rufinamide (n = 160)	Placebo (n = 175)	p-value
NS, not significant. ^aChi square test. ^bt-test. ^cWilcoxon rank-sum test. ^dSeven-category analysis. ^eCochran-Mantel-Haenszel test. ^fScores of: 1+2+3 = improved; 4 = no change; 5+6+7 = worse. ^gFisher’s exact test.
≥50% Responders, n (%)	52 (32.5)	25 (14.3)	<0.001^a
≥75% Responders, n (%)	26 (16.3)	2 (1.1)	<0.001^a
Median reduction in
Total partial seizure frequency rate, %	13.2	5.2	<0.001^b
Total simple partial seizures, %	30.4	20.9	0.264^c
Total complex partial seizures, %	23.8	8.4	0.015^c
Total secondary generalized partial seizures, %	40.0	25.2	0.205^c
Mean increase in seizure-free days, days	2.4	0.9	0.015^c
Clinical Global Impression of Change score,^d n (%)			0.160^e
1 – very much improved	13 (8.1)	6 (3.4)
2 – much improved	39 (24.4)	32 (18.3)
3 – improved	38 (23.8)	36 (20.6)
4 – no change	44 (27.5)	80 (45.7)
5 – minimally worse	3 (1.9)	2 (1.1)
6 – much worse	2 (1.3)	0
7 – very much worse	0	0
Clinical Global Impression of Change (collapsed categories),^f n (%)			0.007^e
Improved	90 (56.3)	74 (42.3)
No change	44 (27.5)	80 (45.7)
Worsened	5 (3.1)	2 (1.1)
Patient Global Impression of Change score,^d n (%)			0.416^e
1 – very much improved	25 (15.6)	7 (4.0)
2 – much improved	33 (20.6)	32 (18.3)
3 – improved	33 (20.6)	37 (21.1)
4 – no change	42 (26.3)	69 (39.4)
5 – minimally worse	4 (2.5)	5 (2.9)
6 – much worse	1 (0.6)	1 (0.6)
7 – very much worse	0	1 (0.6)
Patient Global Impression of Change (collapsed categories),^f n (%)			0.008^e
Improved	91 (56.9)	76 (43.4)
No change	42 (26.3)	69 (39.4)
Worsened	5 (3.1)	7 (4.0)
Number of patients with an increase of ≥25% in partial seizure frequency	30 (18.8%)	28 (16.0%)	NS^g

Exploratory efficacy variables

The median seizure frequency of each partial seizure type (simple partial, complex partial, and secondarily generalized) was reduced (MP vs. BP), but only the reduction in frequency of complex partial seizures reached statistical significance (rufinamide 23.8%; placebo 8.4%; Table 2). Simple partial seizures with motor signs were associated with equivalent percent reductions in median seizure frequency in the rufinamide and placebo groups, whereas rufinamide-treated patients had a numerically greater percent reduction in simple partial seizures without motor signs compared with placebo treatment (45.1% vs. 33.3%; p = 0.438).

There was a statistically significant mean increase in seizure-free days in the rufinamide group compared with the placebo group, and significantly more rufinamide-treated patients showed a 75% reduction in partial seizure frequency (Table 2).

Both the Clinical and Patient GIC seven-category results were notable for a larger percentage of rufinamide-treated patients rated “very much improved,”“much improved,” or “improved,” although the overall result was not significant (Table 2). For the collapsed category analyses, which were planned a priori because of the expectation from previous studies that tail cells would be sparsely populated, there were statistically significant differences between the groups in favor of rufinamide (Clinical GIC: p = 0.007; Patient GIC: p = 0.008) (Table 2).

An increase of ≥25% and of ≥100% in partial seizure frequency (per 28 days, MP vs. BP) was recorded in 18.8% versus 16.0% and in 6.3% versus 4.6% in the rufinamide and placebo groups, respectively.

Safety and tolerability

TEAEs occurred in 141 (80.1%) rufinamide-treated patients and 128 (71.1%) placebo-treated patients. The TEAEs that occurred in >10% of the patients in the rufinamide group were dizziness (26.7%), headache (16.5%), fatigue (15.3%), nausea (13.1%), and somnolence (12.5%). The TEAEs occurring at a rate ≥5% higher in the rufinamide group than the placebo group are shown in Table 3. In addition, there were notably higher percentages of patients in the rufinamide group compared with the placebo group, respectively, with eye disorders (14.2% and 1.7%) and gastrointestinal disorders (31.3% and 17.8%). Diplopia and blurred vision comprised 80% of eye disorder TEAEs, and nausea and vomiting comprised two thirds of gastrointestinal AEs. The majority of TEAEs reported were of mild or moderate severity; there were a greater number of severe TEAEs in the placebo group (n = 14) than in the rufinamide group (n = 10).

Table 3.   Treatment-emergent AEs occurring in ≥5% of rufinamide-treated patients and at a rate ≥5% higher than in placebo (safety population)

	Rufinamide (n = 176), %	Placebo (n = 180), %
Dizziness	26.7	8.3
Fatigue	15.3	10.0
Nausea	13.1	5.0
Somnolence	12.5	7.2
Diplopia	8.0	1.1

TEAEs occurred during the MP in 53.7% and 53.8% of rufinamide-treated and placebo-treated patients, respectively, who achieved a dose of 2,400 mg/day and 60% and 57.8%, respectively, of those who achieved a dose of 3,200 mg/day. The time to first onset of dizziness, nausea, and nystagmus, which tended to occur during the titration phase, was significantly earlier in the rufinamide group than in the placebo group. There was no significant difference in the time to first onset of vomiting or rash between the two treatment groups. As expected, central nervous system (CNS) AEs comprised the largest number of TEAEs. However, in a post hoc analysis, there was a decreased incidence of CNS TEAEs among rufinamide-treated patients receiving three concomitant AEDs compared with those receiving only two (Table 4), which suggests little or no additional CNS tolerability burden from adding rufinamide to existing AED regimens.

Table 4.   Treatment-emergent CNS AEs by number of concomitantly administered AEDs (safety population)

AE^a	Two concomitant AEDs, n (%)		Three concomitant AEDs, n (%)		Overall^b n (%)
	Rufinamide (n = 78)	Placebo (n = 79)	Rufinamide (n = 64)	Placebo (n = 60)	Rufinamide (n = 176)	Placebo (n = 180)
^aOnly those AEs occurring in more than one patient in either treatment group are listed. ^bOverall includes all patients (i.e., those receiving 1–4 concomitant AEDs).
Any	45 (57.7)	27 (34.2)	29 (45.3)	23 (38.3)	95 (54.0)	62 (34.4)
Dizziness	26 (33.3)	6 (7.6)	10 (15.6)	6 (10.0)	47 (26.7)	15 (8.3)
Headache	14 (17.9)	11 (13.9)	9 (14.1)	7 (11.7)	29 (16.5)	23 (12.8)
Somnolence	11 (14.1)	5 (6.3)	8 (12.5)	6 (10.0)	22 (12.5)	13 (7.2)
Tremor	5 (6.4)	4 (5.1)	1 (1.6)	2 (3.3)	8 (4.5)	7 (3.9)
Balance disorder	3 (3.8)	1 (1.3)	4 (6.3)	1 (1.7)	8 (4.5)	4 (2.2)
Convulsion	3 (3.8)	2 (2.5)	3 (4.7)	1 (1.7)	8 (4.5)	3 (1.7)
Coordination abnormality	1 (1.3)	2 (2.5)	3 (4.7)	0	7 (4.0)	4 (2.2)
Sedation	2 (2.6)	3 (3.8)	0	2 (3.3)	2 (1.1)	5 (2.8)
Lethargy	2 (2.6)	2 (2.5)	1 (1.6)	1 (1.7)	4 (2.3)	3 (1.7)
Memory impairment	2 (2.6)	0	2 (3.1)	0	4 (2.3)	0
Hypoesthesia	1 (1.3)	2 (2.5)	0	0	1 (0.6)	2 (1.1)
Speech disorder	2 (2.6)	0	0	0	2 (1.1)	1 (0.6)
Nystagmus	1 (1.3)	0	3 (4.7)	0	4 (2.3)	0

No patient died during this study. Six rufinamide-treated patients (3.4%) and seven placebo-treated patients (3.9%) had treatment-emergent SAEs (TESAEs). Epilepsy-related TESAEs occurred in identical percentages of patients in the two treatment groups: complex partial seizures (0.6% in each group) and convulsion (1.1% in each group).

Discontinuations due to TEAEs occurred in 27 patients (15.3%) in the rufinamide group and 11 placebo patients (6.1%). TEAEs that led to discontinuation in more than one rufinamide-treated patient were dizziness (n = 7); diplopia (n = 3); convulsion (n = 3); headache (n = 3); and nausea, vomiting, somnolence, and pruritic rash (n = 2 for each). The TEAEs that led to discontinuation in more than one placebo-treated patient were dizziness (n = 3), vomiting (n = 2), and sedation (n = 2).Two patients in each group (1.1%) discontinued treatment due to a TESAE: rufinamide (one with complex partial seizures and one with weight decrease and convulsion) and placebo (one with chronic cholecystitis and one with convulsion).

Mean changes in laboratory values over time were small, generally similar across treatment groups, and were not suggestive of a clinically meaningful effect of rufinamide. The incidence of a shift from normal baseline value to abnormal end-point value in hematology and chemistry was low (<5%) for most parameters, and generally did not differ between rufinamide and placebo groups. The only instances of such shifts that occurred in >5% of either group and at a rate in the rufinamide group ≥2 times that in placebo were: low absolute lymphocyte count (9.3% vs. 2.3%), low absolute neutrophil count (5.0% vs. 2.3%), and low lymphocyte concentration (8.1% vs. 4.0%). A further analysis of the low absolute lymphocyte count in the 19 patients that presented this shift revealed that in nine of the patients the change was transient and minimal, and in one case the change was already present at screening. The percentage of patients in the rufinamide group who had a clinically notable decrease in serum potassium concentration during the MP (8.9%) and the combined titration phase and MP (13.0%) of the study was greater than that observed in the placebo group (4.5% and 5.1%, respectively). However, only five rufinamide-treated patients (3.1%) had low potassium concentrations at study end point versus two (1.2%) placebo-treated patients, and just one patient was considered to have developed hypokalemia as a treatment-related AE.

No clinically meaningful change was observed with rufinamide treatment in vital signs, weight, ECGs, physical examinations, or neurologic examinations.

PK/PD results

The oral bioavailability of rufinamide was lower at higher doses, and clearance increased with the subject’s body weight; when the patient’s body weight increases by a factor of 4, clearance is predicted to increase by a factor of 2. The apparent clearance in a typical subject weighing 76 kg was 7.1 L/h at 2,400 mg/day, and 9.5 L/h at 3,600 mg/day. The median trough rufinamide concentration (mg/L) at days 40, 68, and 96 was 15.1, 15.4, and 14.0, respectively, showing there is no effect of prolonged dosing on the PK of rufinamide.

Neither the age nor sex of the patient, nor the coadministration of carbamazepine, lamotrigine, levetiracetam, or phenobarbital had an effect on the apparent clearance of rufinamide. However, a decrease in clearance was observed with coadministration of valproate, which was proportional to valproate serum concentration (23% reduction in clearance at 100 mg/L valproate concentration).

Discussion

Treatment with rufinamide resulted in a statistically significant reduction in total partial seizure frequency compared with placebo. This result was robust, supported by sensitivity analyses and subgroup analyses. The secondary efficacy variables—responder rate (≥50% reduction) and total partial seizure frequency rate—also demonstrated statistically significant results favoring rufinamide. Several exploratory efficacy variables, including ≥75% responder rate and increase in the number of seizure-free days, were also associated with significantly better results for rufinamide.

With respect to efficacy by seizure type, rufinamide was significantly superior to placebo for complex partial seizures, the most common seizure type, and numerically superior to placebo for simple partial seizures and secondarily generalized partial seizures. The median reduction in secondarily generalized partial seizures of 40% in this study is consistent with that previously observed at identical rufinamide dosage (38%, p = NS vs. placebo) (Brodie et al., 2009). Although some patients did have an increase of ≥25% in their seizure frequency, the percentage of patients with this effect was equivalent in both rufinamide and placebo patients, suggesting that the addition of rufinamide to an AED regimen is not associated with an increase in seizures at a rate greater than placebo.

This randomized study extends, in an additional large study population, results from a similar previous study (Brodie et al., 2009). Brodie et al. recorded similar outcomes: a statistically significant median reduction in partial seizure frequency of 20.4% in rufinamide-treated patients, compared with a 1.6% rise in seizure frequency in placebo controls; and a greater proportion of rufinamide patients having a ≥50% decrease in partial seizure frequency compared with placebo controls (28.2% vs. 18.6%). Important differences between the two studies include the number and type of concomitant AEDs received by study patients. In the Brodie study, patients were receiving up to two concomitant AEDs (70% received two), whereas patients in this study were on up to three (44% received two, and 35% received three), underscoring the particularly treatment-refractory condition of the population included in this study. In addition, in the Brodie study, 62% of rufinamide-treated and 58% of placebo-treated patients received carbamazepine, compared with only 20% of patients in this study. In contrast, newer AEDs such as levetiracetam, lamotrigine, and topiramate were among the most commonly used concomitant AEDs in this study. Therefore, the results of this study extend prior findings even though patients were receiving primarily second-generation AEDs and a greater number of concomitant AEDs.

Withdrawals due to AEs occurred at a higher rate in the rufinamide group, and five AEs occurred in rufinamide-treated patients at a rate ≥5% higher than in placebo: dizziness, fatigue, nausea, somnolence, and diplopia. The majority of AEs were of mild or moderate severity. A similar number of patients treated with rufinamide had an SAE compared with placebo, and only two patients from each group discontinued due to an SAE. In addition, there was no clinically meaningful impact on laboratory parameters evident from adding rufinamide to established AED regimens.

Rufinamide PK parameters estimated by sparse sampling in this study were similar to those estimated in previous studies in healthy patients and in patients with epilepsy. The primary purpose of the PK analysis in this study was to obtain data on rufinamide and second-generation AEDs when concomitantly administered. Overall, there were no significant PK effects on either rufinamide or any second-generation AED when coadministered. The PK results confirmed previous findings, showing lower oral bioavailability of rufinamide at higher doses, increased clearance of rufinamide with increasing body weight, and no effect of prolonged rufinamide dosing on the PK of rufinamide. Clearance of rufinamide was decreased with valproate coadministration, as previously observed. The reduction in seizure frequency observed in this study is consistent with that predicted at average rufinamide plasma concentrations of ∼15 mg/L (Perucca et al., 2008).

In summary, this study demonstrates that rufinamide is effective as adjunctive therapy in reducing total partial seizure frequency in treatment-refractory adolescent and adult patients, and confirms the known safety and tolerability profile of rufinamide in this patient population.

Acknowledgments

This study was sponsored by Eisai Medical Research, which was solely responsible for the design and conduct of the study as well as the collection, management, and analysis of the data.

Editorial support was provided by B. Kadish, of PAREXEL, and was funded by Eisai Inc.

Disclosure

Victor Biton reports receiving research grant fees from Eisai and Pfizer for his role as a participating principal investigator; he also reports receiving research grant fees from Forest, Icagen, Impax, Janssen, King, Medivation, Sepracor, UCB, Vernalis, Vertex, and Wyeth. Gregory Krauss reports receiving consulting fees from UCB; honoraria from Samsung Hospital; and research grant fees paid to his institution from Eisai, UCB Pharma, and Icagen. Blanca Vasquez-Santana reports receiving research grant, consulting, and speaking fees from Sepacor, GSK, UCB, and Eisai. Francesco Bibbiani, Allison Mann, Carlos Perdomo, and Milind Narurkar are salaried employees 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.