Gabapentin in generalized seizures

EPILEPSY RESEARCH ELSEVIER

Epilepsy Research 25 (1996) 191-197

Gabapentin in generalized seizures David Chadwick a Deborah B . L e i d e r m a n b, * Wilhelm Sauermann c Jeannine Alexander by Elizabeth Garofalo b,l a Department of Medical and Surgical Neurology, Walton Hospital, Rice Lane 6B, Liverpool L9 1AE, UK b Parke-Davis Pharmaceutical Research, Division of Warner-Lambert Company, 2800 Plymouth Road, Ann Arbor, M148105, USA c DATAMAP GmbH, LiOrracher Str. 16, D-79115 Freiburg, Germany

Received 14 August 1995; revised 28 February 1996; accepted 7 March 1996

Abstract

The efficacy of gabapentin (Neurontin ®) in generalized seizures was evaluated in this 14 week, double-blind, placebo-controlled, parallel-group, add-on, multicenter study. A total of 129 patients with refractory generalized seizures were randomized to receive either placebo or 1200 m g / d a y gabapentin as add-on therapy. Patients received their standard regimens of antiepileptic drugs (AEDs) during a 12 week baseline period, and gabapentin or placebo was added-on in the subsequent 14 week evaluation period. Results of both an intent-to-treat (ITT) and evaluable-patient analyses showed that gabapentin provided greater reduction in the frequency of generalized tonic-clonic seizures than did placebo; however, the differences between treatments were not statistically significant. Gabapentin did not affect the frequency of absence or myoclonic seizures. Adverse events were reported by 67% of gabapentin-treated patients and by 56% of placebo-treated patients. The most frequently occurring adverse events among patients receiving gabapentin were somnolence, fatigue, and dizziness. Gabapentin is well tolerated by patients with generalized seizures. The results of this study show a trend toward an effect of gabapentin in reducing the frequency of generalized tonic-clonic seizures and suggest that further exploration of high dose gabapentin in generalized epilepsy is warranted. Keywords: Gabapentin; Epilepsy; Clinical trials; Anticonvulsant; Generalized seizures

* Corresponding author. I E. Garofalo for the Gabapentin in Generalized Seizures Study Group. This study group comprised: Marian Saunders (Warner-Lambert Company, Eastleigh, UK); David Bates, Niall Cartlidge (Royal Victoria Infirmary, Newcastle-Upon-Tyne, UK); Graham Schapel, Roy Beran (The Queen Elizabeth Hospital, Woodville, Australia); Ken Cumming (Withington Hospital, Manchester, UK); Timothy Betts (University of Aston, Birmingham, UK); Elinor Ben-Menachem (Sahlgren's Hospital, Gothenburg, Sweden); Stephen Wilson (Atkinson Morley's Hospital, Wimbledon, UK); Frank Vajda, Samuel Berkovic (Austin Hospital, Heidelberg, Australia); R.A. Mackenzie (The Prince Henry Hospital, Little Bay, Australia); Peter Newman, Michael Saunders (Middlesbrough General Hospital, Middlesbrough, UK); Victor H. Patterson (Royal Victoria Hospital, Belfast, UK); Stephen W. Brown (The David Lewis Centre for Epilepsy, Cheshire, UK); George Yuill, David Shepherd (North Manchester General Hospital, Manchester, UK); Peter Cleland (Sunderland District General Hospital, Sunderland, UK); G.S. Venables (Royal Hallamshire Hospital, Sheffield, UK); H.G. Boddie (North Staffordshire Royal Infirmary, Stoke-on-Trent, UK). 0920-1211/96/$15.00 Copyright © 1996 Published by Elsevier Science B.V. All rights reserved. PII S 0 9 2 0 - 1 2 1 1 ( 9 6 ) 0 0 0 2 0 - 4

192

D. Chadwick et al./Epilepsy Research 25 (1996) 191 197

1. Introduction Gabapentin (Neurontin®), 1-(aminomethyl)cyclohexaneacetic acid, is a new antiepileptic drug (AED) whose mechanism of action, while as yet undetermined, appears novel. The drug's pharmacologic profile in animal models is unique, and it does not significantly interact with known AED binding sites [6]. Gabapentin binds to a specific receptor associated with the L-amino acid transport system, and its anticonvulsant activity may be related to its effects on brain amino acid concentrations or metabolism [7]. Because most patients with refractory epilepsy have partial seizures, new AEDs typically undergo initial clinical testing in this population. Gabapentin's efficacy in partial epilepsy has been demonstrated [1,8,9], and it is licensed in 19 countries for use as add-on therapy in that indication. However, approximately 30% of patients with epilepsy have either idiopathic or symptomatic generalized epilepsy [4], in which generalized seizures predominate [2]. It is therefore important, in defining the spectrum of activity of a new AED, that clinical trials be undertaken in patients with generalized epilepsies, particularly when the mechanism of action of the new drug is novel. The objective of the study reported here was to assess the safety and efficacy of 1200 m g / d a y gabapentin as adjunctive therapy in patients with refractory generalized epilepsy.

2. Methods This was a randomized, double-blind, placebocontrolled, parallel-group, multicenter add-on study. Eligible patients were 12 years or older and had medically uncontrolled generalized seizures, with a minimum of 1 generalized seizure per week despite treatment with 1 or 2 standard AEDs, not including benzodiazepines. Dosages of standard AEDs were held constant for 3 months prior to the baseline phase. Patients were randomized to treatment at screening and entered a 12 week baseline observation phase, during which they received their prescribed regimens of standard AEDs. Study medication was added to existing therapies during a 14 week efficacy assessment phase, which comprised a 2 week drug

titration period (600 m g / d a y gabapentin, or placebo) and 12 weeks of treatment at full dose (1200 m g / d a y gabapentin, or placebo). Following the efficacy assessment phase, blinded treatment was continued for a further 6 - 1 0 weeks during the interim phase, to allow collection of case report forms before the treatment code was broken for individual patients. At this point, patients were offered the opportunity of receiving gabapentin therapy in an open-label study. Throughout the baseline, efficacy assessment, and interim phases, investigators monitored adverse events, routine biochemical and hematological parameters, and serum concentrations of concomitant AEDs. Seizure counts were documented in a diary completed by patients a n d / o r their care-givers. The primary outcome measure was the reduction in the frequency of generalized tonic-clonic (GTC) seizures; other types of generalized seizures were of secondary interest. Seizure frequency, defined as the number of seizures occurring per 28 days of observation, was compared between the 12 week baseline period and the 14 week treatment period. The primary efficacy variable was responder rate, defined as the percentage of patients with at least a 50% reduction in the seizure frequency during the treatment period as compared with baseline. Seizure reduction was also expressed as percentage change in seizure frequency ( P C H = 100 ( T - B ) / B ) and as a response ratio (RRatio = (T - B ) / ( T + B)), where B is baseline seizure frequency and T is the seizure frequency during treatment. For both variables, negative values indicate a reduction, and positive values indicate an increase in seizure frequency from baseline. A 50% reduction in seizure frequency corresponds to an RRatio of - 0 . 3 3 . The RRatio is symmetrically distributed in a range from - 1 to + 1, which allows the use of parametric statistical methods. This is not the case for percent change in seizure frequency, the distribution of which is often highly skewed. PCH cannot be calculated for patients who had no seizures of a specified type during baseline, while RRatio can be determined for patients who had at least one seizure during either the baseline or treatment period. It is for this reason that N differs between RRatio and PCH in Table 3 through 5 below. All statistical testing was 2-sided, and differences between treatment groups were considered signifi-

D. Chadwick et a l . / Epilepsy Research 25 (1996) 191-197

cant if P < 0.05. The planned sample size of 120 randomized patients was considered sufficient to provide 80% power to detect a difference between treatments, assuming a responder rate of 15% with placebo, 45% with gabapentin, and a dropout rate of 20%. RRatio for GTC seizures was analyzed by an analysis of variance (ANOVA) with treatment, center, and treatment-by-center interaction in the model. Fisher's Exact Test (2-sided) was used to compare responder rates for GTC seizures between treatment groups. The primary analysis was an intent-to-treat (ITT) analysis, including all patients who received at least one dose of study medication. In reviewing patient clinical data and seizure descriptions, it became evident that a number of patients had been enrolled who had a high probability of having partial epilepsy or symptomatic generalized epilepsy. Therefore, a second analysis was conducted, excluding data for pa-

193

tients who had any of the following: partial seizures, localized cerebral pathology on imaging, or an identified etiology consistent with partial epilepsy. This evaluable-patient analysis thus included data only for patients with probable idiopathic generalized epilepsy. Safety data were summarized for all patients who received study medication. Safety summaries include data collected during the 14 week treatment phase as well as the interim phase, which was of a 6 - 1 0 week duration.

3. Results 3.1. Patient characteristics and disposition A total of 129 patients received study medication. In the ITT population, patient characteristics were

Table 1 Patient characteristics at screening Characteristic

Evaluable-patient population

ITT population Placebo ( N = 7 1 )

Gabapentin (N = 58)

Placebo ( N = 36)

Gabapentin (N = 28)

Sex (no. (%) of patients) Men Women

28 (39.4) 43 (60.6)

27 (46.6) 31 (53.4)

14 (38.9) 22 (61.1)

10 (35.7) 18 (64.3)

Age (years) Mean Range

29 13-61

30 16-62

30 13-61

30 17-60

22 2-57

20 3-36

17 4-57

28 4.7 2.2 0-32.7

17 4.8 3.7 0-18.3

1(2.8) 9 (25.0) 18 (50.0) 7 (19.4) 1 (2.8)

0(0.0) 7 (25.0) 19 (67.9) 2 (7.1) 0 (0.0)

Duration of epilepsy (years) Median Range

20 3-42

Baseline seizure frequency per 28 days: GTC seizures N Mean Median Range

57 7.3 3.3 0-103.3

40 7.4 3.9 0-54.3

Number of concurrent AEDs a (no. (%) of patients) 0 1(1.4) 0 1 15 (21.1) ll 2 37 (52.1) 36 3 15 (21.1) 11 4 3 (4.2) 0

(0.0) (19.0) (62.1) (19.0) (0.0)

a Entry criteria specified that patients were to be taking 1 or 2 AEDs, "not including benzodiazepines." Total number of AEDs shown here includes benzodiazepines prescribed for daily administration, but does not include benzodiazepines taken PRN.

194

D. Chadwick et al. / Epilepsy Research 25 (1996) 191-197

comparable between treatment groups (Table 1). The majority of patients were taking 2 AEDs, indicating that this was a highly refractory patient population. The AEDs taken most frequently were valproate (71% of patients), carbamazepine (55%), and phenytoin (33%). Carbamazepine/valproate was the most frequently used combination, taken by 25% of patients. Twenty-four percent of patients were taking a benzodiazepine in combination with other drugs. Characteristics of the evaluable-patient population were similar to those of the ITT population. The small difference in baseline GTC seizure frequency between the ITT and evaluable populations was attributable to the exclusion of patients with clear multifocal brain disease and refractory secondarily generalized seizures. Patient disposition is shown in Table 2. The percentage of patients electing to enter the open-label extension was higher for the placebo treatment group than for the gabapentin group. This is likely due to the fact that the blind was broken for individual patients before they were offered the option of receiving open-label gabapentin (see Methods). Thus, patients who had not responded well to double-blind gabapentin might have been less likely to choose to enter the open-label study than placebo-treated patients, who had not yet had the opportunity to attempt gabapentin therapy.

3.2. Efficacy The results of the ITT analysis of GTC seizures are given in Table 3. The number of patients contributing data to the analysis is somewhat smaller than the number of patients who entered the efficacy assessment phase because not all patients had GTC

Placebo

Gabapentin

57 10 (17.5)

40 11 (27.5) ~'

Responder rate N No. (%) of responders

Response ratio N

57 0.057 ( 0 . 0 6 1 ) --0.034

Mean (S.E.M.) Adjusted mean b

4(1 -0.155 (0.066) --0.181 ~

Percent change j?om baseline N

Median

53 - 15.2

39 29.3

S.E.M. = standard error of the mean; S.D. = standard deviation. '~ P = 0.317, Fisher's Exact Test, 2-sided. b A N O V A with treatment-by-center interaction. ~ P = 0.169, A N O V A .

seizures during the study. The responder rate for GTC seizures was higher among patients in the gabapentin group (28%) than among placebo-treated patients (18%), although the difference between treatment groups was not statistically significant. A similar trend toward effectiveness was seen with the response ratio for GTC seizures, though the difference was again not statistically significant. There were no significant center effects or treatment-bycenter interactions.

Table 4 Efficacy variables for GTC seizures: evaluable-patient population Placebo

Gabapentin

28 4 (14.3)

17 7 (41.2) ~

28 0.002 (0.096) 0.002

17 - 0 . 1 8 2 (0.126) -0.182 b

25 - 15.2

16 -41.9

Responder rate N No. (%) responders

Table 2 Patient disposition (no. (%) of patients)

Entered efficacy assessment phase Withdrawn during double-blind phase Adverse event Lack of efficacy Lack of compliance Lost to follow-up Completed double-blind phase Entered open-label study

Table 3 Primary efficacy variables for GTC seizures: ITT population

Response ratio Placebo

Gabapentin

N

71 (100.0) 6 (8.5) 5 (7.0) 0 (0.0) 0 (0.0) 1 (1.4) 65 (91.5) 60 (84.5)

58 (100.0) 4 (6.9) 2 '~ (3.4) 1 (1.7) 1 (1.7) 0 (0.0) 54 (93.1) 36 (62.1)

Mean (S.E.M.)

'~ Includes one patient who died of asphyxiation during a seizure.

Adjusted mean ( A N O V A )

Percent change fi-om baseline N

Median

S.E.M. = standard error of the mean. *' P = 0.072, Fisher's Exact Test, 2-sided. b p = 0.2496, A N O V A .

D. Chadwick et al. / Epilepsy Research 25 (1996) 191-197

Results for the evaluable-patient analysis (Table 4) also showed a clear trend toward a treatment effect in the gabapentin group, though once again, the difference between treatments failed to achieve statistical significance. The lack of statistical significance is possibly attributable to the low power of the comparison, due to the small number of evaluable patients. The group of patients excluded from the evaluable-patient analysis included a small number of patients who were having documented partial seizures. The remainder of the patients were experiencing multiple types of generalized seizures and appeared to have symptomatic generalized epilepsy, which is commonly highly resistant to treatment. Indeed, these excluded patients did not respond to gabapentin therapy as strongly as evaluable patients. Among excluded patients, for GTC seizures, percent change in seizure frequency from baseline was

Table 5 Efficacy variables for absence and myoclonic seizures: ITT population Placebo

Gabapentin

56 13 (23.2)

39 11 (28.2)

195

Table 6 Overview of adverse events Placebo Gabapentin ( N = 71) ( N = 58) No. (%) of patients with adverse events No. of total reports Maximum intensity of adverse events (no. No AEs Mild Moderate Severe No. of deaths No. (%) of patients withdrawn due to AEs

40 (56.3) 39 (67.2) 95 101 (%) of patients) 31 (43.7) 19 (32.8) 8 (11.3) 14(24.1) 22 (31.0) 16 (27.6) 10 (14.1) 9 (15.5) 0 6 (8.5)

1 4 (6.9)

- 2 8 % for patients treated with gabapentin (N = 23), compared with - 1 6 % for patients who received placebo ( N = 28); responder rate was 17% for gabapentin-treated patients ( N = 23), and 21% for the placebo group (N = 29). Absence seizures and myoclonic seizures did not respond to gabapentin therapy (Table 53).

3.3. Safety

Absence seizures

Responder rate N No. (%) of responders RRatio N Mean (S.E.M.)

56 -0.078 (0.061)

Percent change from baseline N 53 Median - 16.5 Myoclonic

39 -0.117 (0.093)

36 - 16.4

seizures

Responder rate N No. (%) of responders RRatio N RRatio (mean (S.E.M.))

14 1 (7.1)

14 +0.174(0.112)

Percent change Jrom baseline N 13 Median + 2.3 S.E.M. = standard error of the mean.

15 1 (6.7)

15 +0.140 ( 0 . 1 3 4 )

14 + 5.5

The incidence of adverse events and their severity are summarized in Table 6. Sixty-seven percent of gabapentin-treated patients experienced adverse events, compared with 56% of patients who received placebo. Most adverse events were of mild or moderTable 7 Most frequent adverse events a (no. (%) of patients) Adverse event

Placebo ( U = 71)

Gabapentin ( X = 58)

Somnolence Fatigue Dizziness Convulsions Ataxia Weight increase Nausea a n d / o r vomiting Emotional lability Amblyopia Rash Thrombocytopenia Headache

3 (4.2) 4 (5.6) 3 (4.2) 8 (11.3) 5 (7.0) 5 (7.0) 4 (5.6) 3 (4.2) 2 (2.8) 2 (2.8) 0 (0.0) 6 (8.5)

7 (12.1) 6 (10.3) 6 (10.3) 5 (8.6) 4 (6.9) 4 (6.9) 4 (6.9) 4 (6.9) 3 (5.2) 3 (5.2) 3 (5.2) 2 (3.4)

a Events reported for > 3 patients in either treatment group.

196

D. Chadwick et al. / Epilepsy Research 25 (1996) 191 197

ate intensity. The rate of withdrawal due to adverse events was low. One patient death occurred, which was due to asphyxiation during a seizure. The most frequent adverse events are summarized in Table 7. Adverse events that occurred in 10% or more of gabapentin-treated patients were fatigue, somnolence, and dizziness; these events occurred less [¥equently among placebo-treated patients. No significant changes in concomitant AED serum concentrations were noted, and there were no significant differences between treatment groups in clinical laboratory values.

4. Discussion This study represents a first attempt to study the effects of gabapentin as add-on therapy in a heterogeneous group of refractory patients with generalized seizures. The patient recruitment period, which extended from 1986 to 1991, spanned the period during which the International Classification of Epilepsies and Epileptic Syndromes was being discussed and finalized [2-4]. The study enrolled patients who would now be classifed as having a symptomatic generalized epilepsy and others who would be classified as having an idiopathic generalized epilepsy. There is no doubt that the absence of a generally accepted international classification and the heterogeneity of the patient population enrolled in this trial contribute to the difficulties in interpretation of the results. The clinically relevant outcome of this study was a trend toward decreased frequency of GTC seizures among patients treated with add-on gabapentin, as compared with patients treated with add-on placebo. This trend was observed in both the ITT and the evaluable-patient analyses. The failure to achieve statistical significance may be due to the small number of patients contributing data. In the analyses of GTC seizures, the effective sample size was smaller than that originally planned because not all patients enrolled had GTC seizures. Sample size was further diminished for the evaluable-patient analyses, following exclusion of data for patients who likely had partial epilepsy or symptomatic generalized epilepsy. The relatively low dose of gabapentin used may

also have contributed to the lack of significant findings. In the treatment of partial seizures, higher doses of gabapentin seem to be more effective; and gabapentin monotherapy trials currently underway use doses up to 4800 mg/day. The dosage of 1200 m g / d a y administered in the present study may well have been sub-optimal. In this study, gabapentin did not affect the frequency of absence or myoclonic seizures. These findings are consistent with the negative results from a smaller pediatric study in childhood absence epilepsy that used EEG quantification of generalized spike wave activity for efficacy analyses [5]. Gabapentin was exceptionally well tolerated in this patient population, and indeed it would be difficult to differentiate the gabapentin treatment group from the placebo treatment group on the basis of adverse events. These results confirm those from other studies, which show no dose-limiting toxicities with gabapentin. The limitations of this study preclude a definitive conclusion about the efficacy of gabapentin in generalized seizures. The trend toward a positive effect of gabapentin in reducing the frequency of GTC seizures suggests that further exploration of high dose gabapentin in generalized epilepsy is warranted.

References [1] Anhut, H., Ashman, P., Feuerstein, T.J., Sauermann, W., Saunders, M., Schmidt, B. and The International Gabapentin Study Group, Gabapentin (Neurontin) as add-on therapy in patients with partial seizures: a double-blind, placebo-controlled study, Epilepsia, 35 (1994) 795 801. [2] Commission on Classification and Terminology of the International League Against Epilepsy, Proposal for revised clinical and electroencephalographic classification of epileptic seizures, Epilepsia, 22 (1981) 489-501. [3] Commission on Classification and Terminology of the International League Against Epilepsy, Proposal for classification of epilepsies and epileptic syndromes, Epilepsia, 26 (1985) 268 278. [4] Commission on Classification and Terminology of the International League Against Epilepsy, Proposal lbr revised classifi cation of epilepsies and epileptic syndromes, Epilepsia, 30 (1989) 389 399. [5] Leiderman, D., Garofalo, E. and LaMoreaux, L., Gabapentin patients with absence seizures: two double-blind, placebo controlled studies, Epilepsia, 34, Suppl. 6 (1993) 45.

D. Chadwick et aL / Epilepsy Research 25 (1996) 191-197

[6] Taylor, C.P., Mechanism of action of new anti-epileptic drugs. In: D. Chadwick (Ed.), New Trends in Epilepsy Management: the Role of Gabapentin. Int. Congr. and Symp. Ser. No. 198, Royal Society of Medicine Services, London, 1993, pp. 13-40. [7] Taylor, C.P., Emerging perspectives on the mechanism of action of gabapentin, Neurology, 44, Suppl. 5 (1994) S 10-S 16.

197

[8] UK Gabapentin Study Group, Gabapentin in partial epilepsy, Lancet, 335 (1990) 1114-1117. [9] US Gabapentin Study Group No. 5, Gabapentin as add-on therapy in refractory partial epilepsy: a double-blind, placebocontrolled, parallel-group study, Neurology, 43 (1993) 22922298.