Tiagabine for Complex Partial Seizures: Title and subTitle BreakA Randomized, Add-on, Dose-Response Trial FREE

TIAGABINE IS a new potent and specific inhibitor of γ-aminobutyric acid (GABA) uptake into glial and neuronal elements in vitro.¹ This antiepileptic drug (AED) blocks GABA reuptake into presynaptic neurons without itself being transported back into the cells.² The pharmacokinetics of tiagabine and drug-drug interactions have been described elsewhere.^{3 - 4} The first phase 2 trial of this drug demonstrated that daily treatment with a mean dose of 32 mg of tiagabine significantly reduced the frequency of complex partial seizures (CPS) and tonic-clonic seizures.⁵ The present study was designed to further evaluate the efficacy and tolerability of tiagabine (Gabitril) as adjunctive therapy for patients with medically refractory CPS.

Patients with partial epilepsy were entered into the study at 21 treatment sites. Eligibility criteria included (1) age between 12 and 77 years; (2) good health except for epilepsy; (3) occurrence of at least 6 CPS alone or in combination with any other seizure type in the 8 weeks preceding the screening visit (with each of the two 4-week segments containing at least 1 CPS); (4) electroencephalographic evidence of a unilateral or bilateral abnormality consistent with CPS; and (5) availability of at least 1 neuroimaging study of the brain to rule out the presence of any progressive lesions. Female patients could not be pregnant or lactating. The patient had to be receiving a stable regimen of 1 to 3 hepatic enzyme-inducing AEDs: phenytoin, carbamazepine, phenobarbital, or primidone. Valproate could be used in combination with any of the hepatic enzyme-inducing AEDs, but valproate monotherapy was not allowed. The institutional review board of each study site approved the protocol, and written informed consent was obtained from all participants or their guardians.

This randomized, double-blind, placebo-controlled study had a parallel-group, add-on design and lasted 32 weeks (Figure 1). The randomization ratio to treatment groups was 3:2:3:2 for placebo and 16 mg, 32 mg, and 56 mg of tiagabine (Gabitril), respectively. No changes were allowed in the total daily dosage of concomitant AEDs throughout the study.

At the screening visit, a complete medical examination was performed, an electrocardiogram obtained, and complete neurological assessment done. Blood chemistry tests, hematology evaluations, and urinalyses were performed (and repeated at subsequent visits), and plasma concentrations of concomitant AEDs were measured. Patients were instructed in the use of a seizure diary.

At each monthly visit, plasma levels of concomitant AEDs were measured, an interval seizure history was taken, and adverse events were monitored. To be eligible for double-blind treatment, a patient had to have at least 8 CPS during the 12 weeks, with at least 1 occurring during 2 of the three 4-week periods, either alone or in combination with simple partial seizures (SPS) and/or secondarily generalized tonic-clonic seizures (SGTCS).

During the first 4 weeks, the tiagabine dose was titrated to the level at which the patient had been randomized; sham titration was performed for the placebo group. Initially, patients in all 3 tiagabine groups received 8 mg/d (2 mg 4 times a day). Dosage increases were as follows: the 16-mg group had a single increase of 8 mg/d 2 weeks after the first dose; the 32-mg group had a weekly increase of 8 mg/d for 3 successive weeks; and the 56-mg group had a weekly increase of 16 mg/d. If an adverse event occurred, the protocol permitted reduction of the patient's dose for up to 3 days, after which it was increased again; patients who could not tolerate the dosage rechallenge were withdrawn from the trial. The titration period was followed by a 12-week fixed-dose period; after 4 weeks in the fixed-dose period (visit 8), tiagabine levels were measured before dosing and then at 1 and 2 hours after dosing. The fixed-dose period was followed by a 4-week tapering period, during which the tiagabine dose was decreased at the same rate that was used for dose escalation. Compliance was assessed by tablet count in the blister packs that patients brought back to each visit.

Adverse events were evaluated and recorded at every visit and were categorized using the COSTART dictionary.⁶ Also, telephone contact was made at specified intervals.

The primary data set for analysis was derived from the intent-to-treat group, consisting of all randomized patients for whom at least 1 interval seizure history was obtained during the first 16 weeks of double-blind treatment (the dose-escalation and fixed-dose periods). The primary outcome variable was the change in 4-week median CPS frequency from baseline phase to double-blind treatment phase. The primary method of analysis was the van Elteren method⁷ of linearly combining Wilcoxon rank sum test results from individual study sites; pairwise comparisons were performed between each tiagabine group and the placebo group. The proportions of patients having a 50% or greater reduction in CPS frequency were compared using the Cochran-Mantel-Haenszel test with study sites as the strata.

The frequency of SPS and SGTCS was analyzed similarly, except that study sites were not incorporated into the models. Only patients reporting these seizure types during baseline or treatment were included in these analyses. Data on new seizure types experienced by individual patients during the double-blind treatment phase were captured and reported. All tests were 2 tailed, with significance set at P≤.05. The Statistical Analysis System⁸ was used to perform all analyses.

The treatment groups were comparable in baseline demographics, medical histories, and concomitant AED regimens (Table 1). Of 322 patients admitted to the protocol for screening, 297 were randomized to the double-blind phase; however, 5 were excluded from the intent-to-treat analyses because no double-blind assessments were done or their centers lacked patients in all treatment groups. All 297 patients were evaluated for adverse events. Of these, 247 patients completed treatment, with the 56-mg tiagabine group having a significantly lower completion rate than the placebo group. The reasons for premature termination were similar in the 4 groups (Table 2).

Peak plasma concentrations of tiagabine were observed in most patients receiving tiagabine at 1 hour after dosing. Mean tiagabine concentrations before and after dosing are shown in Figure 2. Changes in serum concentrations of individual concomitant AEDs were similar in the 4 treatment groups (Table 3).

Results for the dose-response analysis, measured by change in median 4-week frequency of CPS between baseline and double-blind treatment, are shown in Table 4. Compared with the placebo group, both the tiagabine 32- and 56-mg groups had a statistically significant reduction in median seizure rates. Similarly, the proportions of patients with a 50% or greater reduction in seizure frequency were statistically significantly greater in these tiagabine groups than in the placebo group.

Similar trends were noted for the secondary outcome variables. Of the 292 patients in the intent-to-treat data set, 166 had 1 or more SPS in the baseline phase (Table 5). For those patients, reductions in median SPS frequencies were statistically significant in all 3 tiagabine groups, as were the proportions of patients with 50% or greater reductions in seizure frequencies. During double-blind treatment, increases in SPS frequency were reported in 31 placebo-treated patients, 9 taking 16 mg of tiagabine, 19 taking 32 mg of tiagabine, and 9 taking 56 mg of tiagabine. Six patients with SPS during double-blind treatment had not experienced them during the baseline period; 4 patients, however, reported a history of SPS. One of the 2 patients with no history of SPS experienced 16 SPS (4-week rate, 4.0) and the other experienced 2 SPS (4-week rate, 0.5) during double-blind treatment. One patient with a history of SPS reported 2 SPS (4-week rate, 2.0) during the discontinuation but not during the baseline or treatment phases. Further analyses examining the possibility of increased SPS frequency during the 4-week tapering-off period showed no statistically significant differences in 4-week seizure rates in the baseline and tapering-off periods.

Secondarily generalized tonic-clonic seizures occurred in 106 patients at baseline. The median 4-week frequencies of SGTCS at baseline and during the double-blind treatment were 2.0 and 1.8, respectively, in the placebo group; 1.6 and 0.9 in the 16-mg tiagabine group; 1.8 and 1.0 in the 32-mg group; and 1.5 and 0.8 in the 56-mg group. Fifteen patients with no SGTCS at baseline had them during double-blind treatment: 3 in the placebo group, 4 taking 16 mg of tiagabine, 5 taking 32 mg of tiagabine, and 3 taking 56 mg of tiagabine. All had histories of SGTCS. Five patients without SGTCS during the baseline or double-blind treatment phases reported them during the discontinuation phase: 1 each in the placebo group, 16-mg tiagabine group, and 32-mg tiagabine group, and 2 in the 56-mg tiagabine group. Of these 5 patients, 4 had a history of SGTCS. Some patients in all groups had an increase in SGTCS frequency during double-blind treatment: 16 of 35 in the placebo group, 9 of 28 in the 16-mg tiagabine group, 11 of 32 in the 32-mg tiagabine group, and 10 of 26 in the 56-mg tiagabine group.

When changes in CPS frequency were correlated with trough plasma levels of tiagabine, the 4-week CPS frequency was found to decrease with increasing plasma tiagabine concentrations (Table 6). The median decrease in patients with the highest tiagabine levels was 4.9 and the proportion experiencing a 50% or greater reduction in seizure frequency was 45%; in comparison, patients in the 56-mg dose group had a median decrease of only 2.8 seizures, and 29% experienced 50% or greater seizure reduction (Table 4).

The most frequent adverse events involved the nervous system, reported by 63% of the placebo group, 69% of the 16-mg tiagabine group, 70% of the 32-mg group, and 77% of the 56-mg group. Symptoms with a significantly higher incidence in tiagabine-treated patients were dizziness, tremor, abnormal thinking, and depressed mood (Table 7). Treatment was discontinued prematurely due to adverse events in 33 patients. The numbers of patients discontinued because of adverse events related to the central nervous system compared with the numbers discontinued because of all adverse events were 1 of 7 in the placebo group, 2 of 4 in the 16-mg tiagabine group, 10 of 13 in the 32-mg group, and 8 of 9 in the 56-mg group. No clinically significant changes in vital signs or laboratory values were found. Monitoring of patients for common hematologic and chemistry variables showed no significant differences in the proportions of patients in the placebo group or any of the 3 tiagabine dose groups whose values were outside the normal range. No patient was discontinued from the study because of an abnormal laboratory value.

A clear dose-response relationship was observed between tiagabine dose levels and reduction in CPS frequency, with higher doses of tiagabine (32 and 56 mg/d) resulting in a significantly greater decrease in 4-week seizure frequency than that observed in the placebo group. The frequency of SPS was also significantly reduced; however, reductions in SGTCS did not reach statistical significance. Possibly the threshold of efficacy exists at some dose level between 16 and 32 mg/d in patients taking hepatic enzyme-inducing AEDs. The efficacy of tiagabine in reducing CPS frequency in our study was consistent with the results of another, smaller phase 2 crossover trial⁵ and data from a parallel-group, multicenter adjunctive study.⁹ To our knowledge, the present study is the largest clinical trial of tiagabine reported to date.

The threshold for clinically important adverse events appeared to be similar for the efficacy threshold for tiagabine, ie, somewhere between 16 and 32 mg. Fifteen percent of patients in the 32-mg group and 16% in the 56-mg group discontinued the study prematurely because of adverse events, compared with 7% of patients treated with 16 mg of tiagabine and 8% of placebo-treated patients. None of the between-group differences was statistically significant. Tremor and abnormal thinking (a COSTART term referring to difficulty in concentrating or mental dullness¹⁰ ) appeared to be dose related, while dizziness and depressed mood did not. When abnormal thinking occurred in tiagabine-treated patients, it was most often reported as mild. Tremor was usually limited to a fine tremor affecting the upper extremities; only 1 patient withdrew from the study as the result of this adverse effect. Also, there were no clinically significant drug effects related to vital signs, or hematology, chemistry, and urinalysis results. The adverse event data were similar to those reported in other clinical trials of tiagabine.^{5 ,9}

The serum concentration of tiagabine peaked at 1 to 2 hours after dosing, a finding consistent with data reported in phase 1 dose-ranging studies in healthy volunteers. Three studies indicated that the pharmacokinetics of tiagabine involve linear processes of absorption and elimination for dosing regimens up to 24 mg of tiagabine daily.³ A similar study of tiagabine in which patients were taking hepatic enzyme-inducing AEDs also showed linear pharmacokinetics up to 80 mg/d.⁴ The short half-life of tiagabine in patients taking enzyme-inducing AEDs suggests that 4-times-a-day dosing is appropriate, but that longer dosing intervals could be considered for study. A longer-acting formulation, an extended-release tablet, is being developed by Novo Nordisk (Copenhagen, Denmark). Recent research on cerebral GABA concentrations suggests that the pharmacodynamic effect of tiagabine may outlast its pharmacokinetic effect.¹¹

Other AEDs coadministered in the study included at least 1 enzyme-inducing drug, so that tiagabine would be absorbed and metabolized at a similar rate in all patients. Patients receiving monotherapy with drugs that do not induce hepatic enzyme activity (eg, valproate) were excluded, because the pharmacokinetics and metabolism of tiagabine would have been different than in patients taking enzyme-inducing AEDs, and these potential confounding variables may have interfered with interpretation of study results. Changes in plasma concentrations of concomitant AEDs were unlikely to confound interpretation of efficacy results, as these changes were similar in all treatment groups.

In conclusion, our data indicate that tiagabine reduces the frequency of both CPS and SPS in patients taking 32 and 56 mg daily in 4 divided doses, with the greatest reductions occurring in patients taking the higher doses. Our results also show that treatment with tiagabine can be safely discontinued by tapering over a 4-week period without withdrawal seizures occurring. The most common adverse events associated with use of the drug at higher doses are dizziness and tremor, usually of mild or moderate severity. Tiagabine appears to be effective and well tolerated as an adjunctive drug for the treatment of seizures with partial onset.

Accepted for publication July 29, 1997.

This study was supported by a grant from Abbott Laboratories; the study drugs were also provided by Abbott Laboratories.

In addition to Drs Uthman, Rowan, Ahmann, Leppik, and Schachter, the following investigators entered patients into the study: Gregory L. Barkley, MD, Henry Ford Hospital, Detroit, Mich; Donna C. Bergen, MD, Rush-Presbyterian-St Luke's Medical Center, Chicago, Ill; Lawrence W. Brown, MD, Medical College of Pennsylvania, Philadelphia; Victor Dostrow, MD, University of Mississippi Medical Center, Jackson; Fritz Dreifuss, University of Virginia School of Medicine, Charlottesville; James Ferrendelli, Washington University, St Louis, Mo; Jacqueline French, MD, Graduate Hospital, Philadelphia; Gregory L. Holmes, MD, Children's Hospital, Boston, Mass; Jack A. Madsen, MD, University of Utah Medical Center, Salt Lake City; Denise Malkowicz, MD, Medical College of Pennsylvania; Fumisuke Matsuo, MD, University of Utah Medical Center; Richard Mattson, MD, West Haven VA Medical Center, West Haven, Conn; John M. Pellock, MD, Medical College of Virginia, Richmond; R. Eugene Ramsay, MD, University of Miami Comprehensive Epilepsy Center, Miami, Fla; Jeremy Slater, MD, University of Miami Comprehensive Epilepsy Center; Elson L. So, MD, Mayo Clinic, Rochester, Minn; David M. Treiman, MD, UCLA School of Medicine, Los Angeles, Calif; Braxton B. Wannamaker, MD, Carolina Neurological Clinic, Charleston, SC; and Mark Yerby, Oregon Comprehensive Epilepsy Program, Portland.

Reprints: Basim M. Uthman, MD, VAMC Neurology Service (127), 1601 SW Archer Rd, Gainesville, FL 32608.