Antiepileptic Drugs, 5th Edition



Chemistry, Biotransformation, and Pharmacokinetics

Kenneth W. Sommerville PhD*

Stephen D. Collins MD, PhD**

* Medical Director, Neuroscience, Department of Marketed Products, Clinical Research, Abbott Laboratories, Abbott Park, Illinois

** Associate Medical Director, Department of Marketed Products, Abbott Laboratories, Abbott Park, Illinois

Tiagabine (Gabitril) is a nipecotic acid derivative developed specifically for the treatment of epilepsy. The synthesis and development of tiagabine were motivated by the therapeutic goal of preventing or reducing the uptake of γ-aminobutyric acid (GABA) by presynaptic neurons and glial cells.

GABA is extensively distributed in the mammalian brain, and as a major inhibitory transmitter it is thought to be an important part of most integrative central nervous system functions (1). Reduced GABAergic neurotransmission activity has been implicated in numerous neurologic disorders, including anxiety, pain, and epilepsy (2). The pharmacologic enhancement of GABAergic function is a possible therapeutic approach to alleviating these disorders. Enhancement of GABAergic function could conceivably be effected through direct receptor agonism, inhibition of the enzymatic breakdown of GABA, or action on GABA-coupled ion channels. However, inhibition of the uptake of synaptic GABA by neurons and glial cells offers the advantage of enhancing natural physiologic mechanisms (3).

Nipecotic acid and related cyclic amino acids exhibit anticonvulsant activity in mice, through inhibition of GABA uptake (4); however, the inability of these compounds to cross the blood-brain barrier renders them unsuitable for therapeutic use. Tiagabine, originally synthesized by scientists at Novo-Nordisk in Denmark, has a lipophilic anchor attached via an aliphatic chain to the amino acid nitrogen of nipecotic acid, and readily crosses the blood-brain barrier (5). Tiagabine has demonstrated clinical efficacy against a range of partial seizure types, both as adjunctive therapy (6) and as monotherapy (7, 8, 9).


Tiagabine HCl (hereafter referred to as tiagabine) is a whitish, odorless, crystalline powder, with the chemical name (-) - (R)-1-[4,4-bis(3-methyl-2-thienyl)-3-butenyl]-3-nipecotic acid hydrochloride, and the empiric formula C20H25NO2S2HCl; its structure is shown in Figure 72.1. Tiagabine has a molecular weight of 412.0 and a melting point between 193°C and 195°C. Tiagabine is soluble in water to 3% and is insoluble in heptane. The negative log of dissociation constant (pKa) of the -COOH group is 3.3, and the pKa of the -NH2 moiety is 9.4. Tiagabine has a partition coefficient of 39.3 in octanol and water at pH 7.4. The nipecotic acid moiety has an asymmetric carbon atom; the (R)-(-)-enantiomer is four times more potent than the (S)-(+)-enantiomer, and the name tiagabine refers to the (R)-(-)-enantiomer.

Tiagabine can be quantified in samples of human plasma through the use of a sensitive and precise high-performance liquid chromatography procedure developed at Abbott Laboratories in Illinois. The procedure involves separation on a C18 column using a mobile phase containing the ion-pairing reagent sodium octane sulfonate; the limit of detection is 2 ng/mL. Tiagabine is stable in plasma samples; no evidence of degradation is observed after 23 hours at room temperature or after 2 months at -20°C (10).


Tiagabine is rapidly absorbed, with the time to maximal plasma concentration (Tmax) <2 hours in healthy volunteers


under fasting conditions. The elimination half-life (t½) ranges from 4 to 9 hours, regardless of dose. Absorption and elimination of tiagabine are linear processes across the therapeutic range of tiagabine doses, in both single-dose and steady-state studies (11).


FIGURE 72.1. Chemical structure of tiagabine.

Tiagabine absorption was examined after single doses (dose range, 2 to 24 mg) administered to 58 healthy male subjects (11) in three phase I studies. Tiagabine was rapidly absorbed across all doses; Tmax occurred ≤1 hour after administration in most subjects. The maximum plasma concentration (Cmax) was proportional to the dose; no evidence of nonlinearity was found throughout the dose ranges studied (study I, 2 to 24 mg; study II, 2 to 10 mg; study III, 6 or 12 mg). Both the dose-adjusted area under the plasma concentration-time curve (AUC) and Cmax were independent of the administered dose; the dose-adjusted AUC averaged 105 ng-hr/mL/mg, and the dose-adjusted Cmax averaged 20.8 ng/mL/mg in study I. In most subjects, plasma concentration-time curves were biphasic and indicated possible enterohepatic recirculation. The recirculation created some uncertainty regarding the estimation of elimination t½; the harmonic mean t½ was 6.7 hours. The single-dose pharmacokinetic properties of tiagabine are summarized in Table 72.1.


Dose (mg)


Tmax (h)

Cmax (ng/mL)

T½ (h)a

AUC (ng · h/mL)



0.8 ± 0.5

43.4 ± 14.7


228 ± 71.4



1.0 ± 0.0

150 ± 26.2


727 ± 165



1.3 ± 1.0

241 ± 78.7


1450 ± 420



1.1 ± 0.8

552 ± 189


2190 ± 597

AUC, Cmax, and Tmax are reported as mean ± standard deviation; T½, half-life.

a Harmonic mean.

Adapted from Gustavson LE, Mengel HB. Pharmacokinetics of Tiagabine, a γ-aminobutyric acid-uptake inhibitor, in healthy subjects after single and multiple doses.Epilepsia 1995;36:605-611, with permission.

Tiagabine pharmacokinetics at steady state also demonstrate dose-proportional behavior. In a study of healthy male subjects receiving once-daily doses of tiagabine (dose range, 2 to 10 mg/day) for 5 days, dose-adjusted Cmax values (range, 16.0 to 26.5 ng/mL/mg) were generally independent of the dose administered (11). The dose-adjusted AUC also appeared to be independent of dose at day 1 and day 5, with accumulation ratios throughout the 5-day regimen clustered near 1.0. As in the single-dose study, tiagabine was rapidly absorbed: Tmax was reached 0.5 to 1.5 hours after administration under fasting conditions. Tmax values were independent of both the size of the dose and the number of doses administered (Table 72.2).


Tiagabine exhibits high bioavailability regardless of the formulation. The mean absolute bioavailability was 89.9±9.7% in a study comparing oral tablets and intravenous infusions in healthy subjects (12). The bioavailability of tiagabine administered as tablets, capsules, and oral solution is essentially identical (13).





Tmax (hr)

Dose-adjusted Cmax (ng/mL/mg)

T½ (hr)

Dose-adjusted AUC (ng · hr/mL/mg)



Day 1







Day 1






Day 5







Day 1






Day 9






Day 14





AUC, area under the plasma concentration-time curve, a measure of drug exposure; Cmax, maximum plasma concentration after a dose; Tmax, time to maximum plasma concentration after a dose; T½, half-life.

The intake of food concomitant with tiagabine administration appears to reduce the rate of tiagabine absorption, but not its extent. Whereas Tmax is extended more than twofold and the peak plasma concentration is lower in the fed state than in the fasting state, the AUC is similar (14). Clinical efficacy trials of tiagabine were conducted using administration in the fed state, to reduce variations in plasma concentration levels and possibly to reduce the occurrence of intolerability from the higher Cmax in the fasting state.


Tiagabine is available as a film-sealed tablet in the United States in nondivisible strengths of 2, 4, 12, 16, and 20 mg (15). An extended-release formulation has been used in a pilot study of neuropathic pain in patients with diabetes (16). Other formulations, including intravenous and liquid formulations, have been explored but have not been studied clinically. Because tiagabine is water soluble, further development of these formulations should be possible.


Plasma Protein Binding

In vitro studies have shown that >95% of tiagabine in plasma is bound to protein, primarily to albumin and α1-acid glycoprotein. In healthy volunteers, approximately 4% to 5% of tiagabine is unbound (17). Tiagabine is not displaced by phenytoin, carbamazepine, or phenobarbital, nor does it displace these drugs in plasma. Valproate reduces tiagabine binding by a small, but statistically significant, amount—from 96.3% to 94.8%; however, tiagabine has no effect on valproate binding. Salicylate and naproxen displace tiagabine, but the clinical implications of these changes are likely to be minimal. Tiagabine binding is not affected by various other drugs that were studied in vitro, including propranolol, verapamil, chlorpromazine, amitriptyline, imipramine, warfarin, ibuprofen, digitoxin, furosemide, tolbutamide, and haloperidol (18).

Central Nervous System

Tiagabine crosses the blood-brain barrier and increases GABA levels in both the extracellular fluid of the brain and the cerebrospinal fluid. An analysis of extracellular GABA levels using microdialysis probes mounted on deeply implanted hippocampal electrodes (in a preoperative evaluation of patients with intractable seizures) showed an increase of approximately 50% in GABA level beginning 1 hour after administration of a tiagabine dose. The increase in GABA level was sustained for several hours, with no observed drug-related adverse events (19).

In a trial for treatment of refractory partial seizures, samples of cerebrospinal fluid were obtained from 10 patients at baseline and after 3 months of treatment with tiagabine or with placebo. The cerebrospinal fluid samples obtained after treatment demonstrated a significant increase in GABA levels, over both baseline levels and levels observed in patients who received placebo (20).

Transporters, Transplacental Passage, and Transmission in Breast Milk

The synaptic action of GABA is terminated by rapid uptake into the presynaptic terminals and adjacent glial cells. The uptake is mediated by the transporters GAT-1, GAT-2, GAT-3, and BGT-1 (21). In vitro testing showed tiagabine has the greatest affinity for GAT-1 and relatively low affinity for GAT-2, GAT-3, and BGT-1. GAT-1 is likely to be the transporter responsible for the anticonvulsant action of tiagabine.

No clinical data are available on the transplacental passage or excretion in breast milk. The high protein binding


of tiagabine may predict limited crossing of the placenta or into breast milk.


Tiagabine is rapidly absorbed and extensively metabolized after oral administration. No active metabolites of tiagabine have been identified. In healthy volunteers receiving 4 mg of14C-tiagabine, approximately 63% of the total radioactivity was excreted through the feces, and 25% was excreted in the urine. Only 2% of the initial 4-mg dose was excreted unchanged in the urine. Isomers of 5-oxo-tiagabine, the major metabolite, were identified in the urine; these are products of the thiophene ring oxidation pathway (22). The 5-oxo-tiagabine metabolite is inactive.

The metabolism of tiagabine appears to involve primarily the 3A subfamily of cytochrome P450 (CYP) enzymes. Incubation of 14C-tiagabine with human hepatic microsomes and NADPH (the reduced form of nicotinamide-adenine dinucleotide phosphate) demonstrated formation of two 5-oxo-tiagabine isomers, as well as a significant correlation between the disappearance of 14C-tiagabine and CYP3A-mediated activity. Metabolism of tiagabine is inhibited by CYP3A-selective inhibitors, and tiagabine was metabolized slowly by cDNA-expressed CYP3A4. Correlation and selective inhibition studies, along with studies of other cDNA-expressed CYP forms (including CYP2D6, 2E1, 1A2, 2C9, and 2A6), provided little evidence that other CYP isoforms play a significant role in the metabolism of tiagabine (23).

In contrast to several other antiepileptic drugs (AEDs), tiagabine does not appear to induce or inhibit hepatic microsomal enzymes. Analysis of the pharmacokinetics of antipyrine, an indirect measure of hepatic microsomal enzyme activity, revealed no significant differences in clearance or t½ when it was administered before and after 14 days of tiagabine administration (11).


As in healthy volunteers, the pharmacokinetic behavior of tiagabine in patients with epilepsy is linear across all doses. However, the pharmacokinetics of tiagabine in patients with epilepsy is significantly affected by the concomitant administration of enzyme-inducing AEDs (24). The clearance of tiagabine in patients treated with enzyme-inducing drugs is significantly higher than in those treated with tiagabine in combination with noninducing drugs, such as valproate (25,26). This finding suggests that noninduced patients may require a lower dose of tiagabine than induced patients to produce the same serum concentration and therapeutic effect.

One study examined patients taking tiagabine (dose range, 24 to 80 mg daily) as an add-on to a stable regimen of one to three known enzyme-inducing AEDs. Tiagabine demonstrated linear pharmacokinetics across all dosages, with no significant differences in dose-adjusted Cmax, trough concentration (Cmin), and AUC values in induced patients. The harmonic mean t½ ranged from 3.8 to 4.9 hours, in contrast to a harmonic mean t½ of 7.1 hours in 30 historic control subjects not taking enzyme-inducing AEDs (Figure 72.2) (24).

In a population analysis involving 511 patients from clinical studies, age 11 to 77 years (mean age, 32.1± 12.3 years), the pharmacokinetics of tiagabine as an add-on therapy was evaluated using a one-compartment model with first-order absorption and elimination (26). No differences were observed in tiagabine pharmacokinetics when patients were analyzed by sex, race or ethnicity, or other selected demographic variables (including age and smoking), and, as in earlier studies, the pharmacokinetics of tiagabine was linear. However, the central clearance value in patients receiving concomitant treatment with enzyme-inducing AEDs was 67% higher than that observed in patients taking only noninducing AEDs (21.4 versus 12.8 L/hr). The effect of enzyme-inducing AEDs on the central clearance value for tiagabine was not additive; the average values for patients receiving a single concomitant enzyme-inducing AED were similar to those for patients receiving multiple enzyme-inducing AEDs.

These findings have important clinical implications. To maintain consistent serum concentrations, tiagabine dosage should be adjusted when either adding or removing adjunctive enzyme-inducing AEDs.



Tiagabine pharmacokinetics in children are similar to those observed in adults. A study of a single tiagabine dose evaluated pharmacokinetics in 25 children (ages 3 to 10 years) who were undergoing epilepsy treatment with either enzyme-inducing or noninducing AEDs (27). The children were given tiagabine at a dose of 0.1 mg/kg. As in the adult population, tiagabine was cleared more quickly in the group of children taking concomitant enzyme-inducing AEDs than in the group taking valproate, a noninducing AED. The harmonic mean t½ in induced pediatric patients was 3.2 hours, consistent with findings in induced adults, whereas the harmonic mean t½ in noninduced children was 5.7 hours, also consistent with adult findings. The plasma concentration-time curves for induced and noninduced pediatric patients are shown in Figure 72.3.




FIGURE 72.2. Mean tiagabine plasma concentrations after a morning dose (6 AM) and an evening dose (6 PM) of tiagabine in 23 patients taking concomitant enzyme-inducing AEDS. (From So EL, Wolff D, Graves NM, et al. Pharmacokinetics of tiagabine as add-on therapy in patients taking enzyme-inducing antiepilepsy drugs. Epilepsy Res 1998;22:221-226, with permission.)


FIGURE 72.3. Plasma concentration-time profiles for tiagabine in pediatric epilepsy patients treated with other antiepileptic drugs (circles, induced; squares, valproate). (From Gustavson LE, Boellner SW, Granneman GR, et al. A single-dose study to define tiagabine pharmacokinetics in pediatric patients with complex partial seizures. Neurology1997;48:1032-1037, with permission.)

An interesting finding that bears on dosage selection for children is that the pharmacokinetic parameters of volume of distribution over bioavailability (Varea/F) and clearance over bioavailability (CL/F) in pediatric patients were more strongly correlated with body surface area than with body weight. At least one study (28) has indicated that liver volume (presumably related to the rate of drug metabolism) correlates more strongly with body surface area than with weight: liver volume (measured by magnetic resonance imaging) normalized to weight decreases with increasing age, whereas liver volume normalized to body surface area remains constant with increasing age. This finding suggests that, at least for certain drugs, body surface area may be a more useful variable than body weight in calculating pediatric doses.

Elderly Patients

Elderly patients are a particularly important population in epilepsy therapy, because 25% of new epilepsy diagnoses are made in people >65 years old, primarily as the result of vascular insult. In addition, age-related changes in metabolism, renal and liver function, and body composition may play a role in the pharmacokinetics of specific drugs. However, tiagabine pharmacokinetics in elderly patients appears similar to that observed in younger adult populations.



In a comparison of healthy elderly volunteers (≥65 years; mean, 70.8; range, 65.9 to 75.7; n = 8), elderly patients with epilepsy who take enzyme-inducing AEDs (≥65 years; mean, 69.4; range, 66.3 to 72.5; n = 8), and healthy young volunteers (mean, 25.6; range, 20.2 to 31.0; n = 8), the pharmacokinetics of tiagabine after single- and multiple-dose administration was similar in healthy young and elderly volunteers (29). The exception was the AUC on the final day of the multiple-dose trial, which was significantly higher (103±29 ng-hr/mL/mg versus 72-20 ng-hr/mL/mg) in the elderly volunteers compared with the younger subjects.

As with other populations, Cmax, AUC, and t½ were significantly lower in elderly patients being treated with enzyme-inducing AEDs compared with healthy elderly or young volunteers. The overall pharmacokinetic parameters, for both healthy elderly volunteers and elderly patients with epilepsy who were taking inducing AEDs, were quite similar to those observed in corresponding groups of younger patients. These findings suggest that tiagabine dosage may not need to be adjusted on the basis of age, but as with younger patients, the dosage should be increased in patients receiving inducing AEDs concurrent with tiagabine administration. This study, however, included relatively young elderly patients; very old patients may require dose adjustments if their hepatic function is reduced.

Patients with Hepatic Impairment

The alterations in tiagabine pharmacokinetics that result from the concomitant use of hepatic enzyme-inducing AEDs, in addition to the very small fraction of tiagabine eliminated as unchanged drug in urine, suggest that hepatic function is important in the elimination of tiagabine. This suggestion was confirmed in a pharmacokinetic study comparing healthy subjects and subjects with varying degrees of hepatic impairment.

Volunteers with mild or moderate liver function impairment (Child-Pugh classification) were compared with healthy volunteers in a multiple-dose study with regard to their ability to metabolize tiagabine (30). Four subjects had mild hepatic impairment, three had moderate hepatic impairment, and six were matched healthy subjects. Each study subject received twice-daily oral tiagabine for 5.5 days, and serial blood specimens were obtained for 48 hours after the final dose. The mean AUC and Cmin values were significantly greater in both hepatic impairment groups compared with values observed in the group of healthy volunteers. The mean Cmax value was significantly higher in the mild impairment group, and it was higher (but did not reach statistical significance, probably because of the small group size) in the moderate impairment group than in the healthy group. Plasma concentration-time curves for patients with mild and moderate hepatic impairment are compared with the curve from healthy subjects in Figure 72.4.


FIGURE 72.4. Mean tiagabine plasma concentrations after a final dose in patients with normal hepatic function, mild hepatic impairment, and moderate hepatic impairment. (From Lau AH, Gustavson LE, Sperelakis R, et al. Pharmacokinetics and safety of tiagabine in subjects with various degrees of hepatic function. Epilepsia 1997;38:445-451, with permission.)

The mildly and moderately impaired hepatic function groups exhibited higher plasma tiagabine Cmin than the normal group, a finding indicating that elimination of tiagabine is slowed in patients with liver function impairment. There were also more adverse events reported by the impaired hepatic function groups. As is generally seen with tiagabine, many of these events involved the central nervous system. The plasma levels of unbound tiagabine were higher in the groups with impaired hepatic function, a finding consistent with the reduced levels of serum albumin and α1-acid glycoprotein that are typical of liver disease.

These findings strongly suggest that tiagabine should be used with caution in patients with epilepsy and impaired hepatic function, and tiagabine dosage should be reduced or intervals between doses should be increased to maintain safe and effective plasma concentrations. Such patients should be monitored closely for signs of neurologic side effects, and the dosing regimen should be adjusted appropriately.

Patients with Renal Impairment

In contrast to the situation with hepatic function impairment, renal impairment does not appear to alter tiagabine pharmacokinetics. Although drugs such as gabapentin and topiramate are excreted unchanged primarily in the urine, tiagabine is extensively metabolized and removed mostly


through fecal elimination. The fraction excreted by the kidneys (~25%) is composed almost completely of inactive metabolites of tiagabine.

A single- and multiple-dose study of tiagabine pharmacokinetics was conducted in 25 nonepileptic adult volunteers stratified according to degree of renal impairment (17). Group I (n = 5) had normal renal function, group II (n = 5) had mild renal impairment (creatinine clearance, 40 to 80 mL/min/1.73 m2), group III (n = 6) had moderate renal impairment (creatinine clearance, 20 to 39 mL/min/1.73 m2), group IV had severe renal impairment (creatinine clearance, 5 to 19 mL/min/1.73 m2), and group V (n = 5) was receiving hemodialysis. For groups I to IV, the study was a 7-day, single- and multiple-dose, one-period, open-label study. These groups received a single 4-mg dose of tiagabine on day 1, 4 mg every 12 hours on days 2 to 4, and a single 4-mg dose on day 5. For group V, the study was a single-dose, two-period, open-label study. In period I, a single 4-mg dose of tiagabine was administered approximately 2 hours before dialysis. In period II, a single 4-mg dose was administered before a dialysis-free period.

Although small differences were observed among the groups, no correlation existed between the degree of renal impairment and the pharmacokinetic parameters. The accumulation ratios were nearly identical (~1.5) in all groups receiving multiple doses, and there was no correlation between the incidence of adverse events and the degree of renal impairment. It can be concluded that renal impairment does not substantially alter the pharmacokinetics of tiagabine; therefore, special dosage adjustment is not necessary for patients with renal impairment.


FIGURE 72.5. Tiagabine plasma concentration-time profiles for a subject who received 2 mg tiagabine once daily for 5 days (day 1, open squares; day 5, solid squares) and for a subject who received 10 mg tiagabine once daily for 5 days (day 1, open circles; day 5, solid circles). (From Gustavson LE, Mengel HB. Pharmacokinetics of tiagabine, a γ-aminobutyric acid-uptake inhibitor, in healthy subjects after single and multiple doses. Epilepsia 1995; 36:605-611, with permission.)


Tiagabine exhibits linear pharmacokinetics for single and multiple doses in both healthy volunteers (11) and enzyme-induced patients with epilepsy (24). The dose-adjusted Cmax in healthy volunteers is in the range of 15 to 26 ng/mL/mg (11). Persons taking a single 2-mg dose have average Cmax values of 38 ng/mL, whereas a single 10-mg dose averages 265 ng/mL (Figure 72.5).

As expected, plasma concentrations with concomitant hepatic enzyme-inducing AEDs are lower than in healthy volunteers, but they exhibit linear pharmacokinetics (24). The dose-adjusted Cmax among 21 patients with epilepsy who were taking enzyme-inducing AEDs in total daily doses of 40, 56, or 80 mg ranged from 15.3 to 18.6 mg/mL/mg in the morning to 15.8 to 15.9 ng/mL/mg in the evening. Dose-adjusted Cmin values averaged 2.9 to 4.2 ng/mL/mg in the morning and 3.0 ng/mL/mg in the evening. Data from other sources reported with this study indicated the dose-adjusted Cmax to be 8 to 12 ng/mL/mg and the dose-adjusted Cmin to be 2 to 4 ng/mL/mg. These data may not reflect the concentrations at the true Tmax, which is often <1 hour. The diurnal variations are small and are unlikely to have clinical importance.

Extrapolating these data to the clinical setting, a single 8-mg dose should produce a Cmin concentration of approximately 16 to 32 ng/mL in an enzyme-induced patient. This relationship holds throughout the commonly prescribed dose range for tiagabine, as suggested in a clinical dose-response trial. The concentrations across increasing doses show linear increases at trough and at 1 and 2 hours after administration (Figure 72.6). The dose-adjusted values


from the same dose-response trial were similar for each dosage group (31).


FIGURE 72.6. Mean (±SD) plasma concentrations of tiagabine in patients with complex partial seizures treated with three dose levels of tiagabine, as evaluated after 4 weeks of fixed-dose treatment. Left, concentrations by absolute values; right, concentration by milligrams of tiagabine administered. (From Uthman BM, Rowan AJ, Ahmann PA, et al. Tiagabine for complex partial seizures: a randomized, add-on, dose-response trial. Arch Neurol 1998;55:56-62, with permission.)


Tiagabine has been shown to raise GABA levels in the nervous system, but the increase is transient (19). Serum concentrations of tiagabine may therefore correlate with clinical effects.

The relationship of serum concentrations with adverse events was described in pharmacokinetic studies of healthy volunteers (11). All six subjects given once-daily 6-mg doses of tiagabine for 14 days completed the study, whereas one of three randomly assigned subjects receiving once-daily 12-mg doses discontinued the study because of dizziness, difficulty in concentrating, incoordination, and somnolence. All six subjects receiving once-daily 10-mg doses for 5 days completed the study.

Among healthy subjects receiving single doses of 2, 8, 12, or 24 mg, there were no adverse events with Cmax concentrations <180 ng/mL (12). Persons with Cmax values in the range of 200 to 400 ng/mL had the same incidence of adverse events as those with Cmax values >400 ng/mL. The adverse events found with concentrations >400 ng/mL were, however, more severe. The adverse events tended to occur at the time of the predicted Cmax (1 to 2 hours) and generally resolved by 4 to 6 hours after administration.

The relation of serum concentration to efficacy was explored in a United States multicenter, placebo-controlled, add-on trial of 16, 32, and 56 mg/day versus placebo in enzyme-induced patients with partial seizures (31). Serum concentrations in groups with Cmin values of ≤20 ng/mL, 21 to 40 ng/mL, and >40 ng/mL were compared with placebo, as well as by median reduction and those with ≥50% reduction in 4-week rates of complex partial seizures. Both median reduction and the proportion of those with ≥50% reduction in seizure frequency increased with increasing concentrations of tiagabine. Of the group with the highest concentrations, 45% had a ≥50% reduction in complex partial seizures. This is a better result than the 29% with ≥50% seizure reduction in the group receiving doses of 56 mg/day (Table 72.3). These comparisons are limited, because the timing of Cmin in multicenter trials is estimated, and the three concentration groups were designated in a post hoc analysis. Nevertheless, a


significant correlation was found between efficacy in reducing complex partial seizures and increasing Cmin values.


Trough Level, ng/mL

Baseline 4-wk Seizure Frequency

Treatment Phase: Change in Seizure Frequency

Median Change

P vs Placebo

Median % Change

≥50% Seizure Reduction


Median (Range)

No. (%)

P vs Placebo

0 (Placebo)


7.4 (2.8-109)



4 (4)



8.6 (2.6-209)




11 (13)




10.2 (3.1-136)




7 (20)




7.5 (2.6-77)




9 (45)


From Uthman BM, Rowan AJ, Ahmann PA, et al. Tiagabine for complex partial seizures: a randomized, add-on, dose-response trial. Archives Neurol 1998;55:56-62, with permission.

These data do not provide a definitive therapeutic range of serum Cmin values. It is not clear whether the pharmacodynamic effects of tiagabine are longer lasting than the pharmacokinetic effects. Another placebo-controlled, multicenter trial in the United States compared tiagabine at daily doses of 32 mg, given either twice or four times daily, with placebo (32). As expected, the average Cmin values in the twice-daily group were in the range of 14 to 16 ng/mL, whereas values for the group receiving four daily doses were in the range of 29 to 30 ng/mL. There were large standard deviations for both groups, a finding reflecting the nature of these data in multicenter trials. The efficacy of the drug for these two groups was similar; the proportion attaining ≥50% reduction in seizures was 31% in the twice-daily group and 27% in the group receiving four daily doses. The serum concentrations in this study did not correlate as well with efficacy as in the dose-response study. Based on these results, a longer pharmacodynamic than pharmacokinetic effect is possible with tiagabine.


The pharmacokinetics of tiagabine are summarized in Table 72.4. Tiagabine is rapidly absorbed. The presence of food slows absorption about twofold; however, administration to fed patients does not alter bioavailability, because the overall AUC is similar to that of the fasting state. Tiagabine penetrates the blood-brain barrier and increases extracellular GABA levels in the central nervous system. Once absorbed, tiagabine is highly protein bound (>95%) and is rapidly and extensively metabolized, primarily through the action of the 3A subfamily of hepatic CYP enzymes; no pharmacologically active metabolites have been identified. Tiagabine and its metabolites are excreted mainly through the feces. About 25% of the total ingested dose is eliminated as metabolites in the urine. Less than 2% of the total administered dose is excreted unchanged. Tiagabine has a relatively short t½ of 5 to 9 hours, which is reduced to 3 to 5 hours when the drug is given with enzyme-inducing AEDs.

Tiagabine demonstrates linear (non-dose-dependent) pharmacokinetics across the therapeutic dose range ≤80 mg/day. The therapeutic effect does correlate to some extent with plasma Cmin values, but this relationship is limited by the short t½. Despite the short t½, a therapeutic effect has been shown with twice-daily administration. Tiagabine does not induce hepatic enzymes, but its clearance is greater in the presence of enzyme-inducing AEDs. This is the only clinically important drug interaction demonstrated so far. Doses of tiagabine may need to be higher to maintain required concentrations when the agent is given with enzyme-inducing AEDs (e.g., phenytoin, carbamazepine, primidone, and phenobarbital). The pharmacokinetics of tiagabine in elderly patients are similar to that observed in younger adults, although clearance is somewhat lower.




Tmax (fasting)

0.5-2 hr

Plasma protein binding


% metabolized

>98% (<2% excreted unchanged in urine)

Metabolic pathway

Primarily via CYP3A subfamily

Liver enzyme induction


Liver enzyme inhibition


t½ (enzyme-induced patients)

3-5 hr

t½ (non-induced patients)

5-9 hr

Effect of renal-function impairment


Effect of hepatic-function impairment

Increase in AUC0-12h, Cmin, Cmax

AUC0-12h, area under the plasma concentration-time curve, a measure of drug exposure; Cmax, maximum plasma concentration after a dose; Cmin, minimum plasma concentration after a dose; CYP3A, a subfamily of hepatic enzymes commonly involved in drug metabolism; Tmax(fasting), time to maximum plasma concentration after a dose.



Pediatric pharmacokinetics also appears to be similar to that in adults, although more closely correlated with body surface area than with body weight. Renal impairment does not alter the pharmacokinetics of tiagabine. However, tiagabine should be used with caution, possibly with lower doses or longer dosing intervals in patients, with hepatic impairment, which substantially decreases the rate of metabolism and clearance.


We gratefully acknowledge the review of this manuscript by Linda Gustavson, PhD.


  1. Meldrum B. Pharmacology of GABA. Clin Neuropharmacol1982;5:293-316.
  2. Krogsgaard-Larsen P, Scheel-Krüger J, Kofoed H. GABA-neuro-transmitters: pharmacochemical, biochemical and pharmacological aspects.Copenhagen: Munksgaard, 1979.
  3. Nielsen EB, Suzdak PD, Andersen KE, et al. Characterization of tiagabine (NO-328), a new potent and selective GABA uptake inhibitor. Eur J Pharmacol1991;196:257-266.
  4. Croucher MJ, Meldrum BS, Krogsgaard-Larsen P. Anticonvulsant activity of GABA uptake inhibitors and their prodrugs following central or systemic administration. Eur J Pharmacol1983; 89:217-228.
  5. Yunger LM, Fowler PJ, Zarevics P, et al. Novel inhibitors of GABA uptake: anticonvulsant actions in rats and mice. J Pharmacol Exp Ther1984;228:109.
  6. Ben-Menachem E. International experience with tiagabine add-on therapy. Epilepsia1995; 36[Suppl 6]:S14-S21.
  7. Alarcon G, Binnie CD, Elwes RDC, et al. Monotherapy antiepileptic drug trials in patients undergoing presurgical assessment: methodological problems and possibilities. Seizure1995; 4:293-301.
  8. Schachter SC. Tiagabine monotherapy in the treatment of partial epilepsy. Epilepsia1995;32[Suppl 6]:S2-S6.
  9. Brodie MJ, Bomhof MAM, Kalviainen R, et al. Double-blind comparison of tiagabine and carbamazepine monotherapy in newly diagnosed epilepsy: preliminary results. Paper presented at the 22nd International Epilepsy Congress, Dublin, 1997.
  10. Gustavson LE, Chu SY. A high performance liquid chromatographic procedure for the determination of tiagabine concentrations in human plasma using electrochemical detection. J Chromatogr1992;574:313-318.
  11. Gustavson LE, Mengel HB. Pharmacokinetics of Tiagabine, a γ-aminobutyric acid-uptake inhibitor, in healthy subjects after single and multiple doses. Epilepsia1995;36:605-611.
  12. Jansen JA, Oliver S, Dirach J, et al. Absolute bioavailability of tiagabine. Epilepsia1995;36[Suppl 3]:S159.
  13. Mengel HB, Gustavson LE, Soerensen HJ, et al. Bioavailability of tolerability of a tiagabine HCl tablet formulation versus capsules and oral solution in normal subjects.Epilepsia1991;32 [Suppl 3]:S6.
  14. Mengel HB, Gustavson LE, Soerensen HJ, et al. Effect of food on the bioavailability of tiagabine HCl. Epilepsia1991;32[Suppl 3]:S6.
  15. Gabitril (tiagabine hydrochloride). Package insert. Cephalon, Inc., Westchester, Pennsylvania, 2000.
  16. Kirby LC, Collins SD, Deaton RL, et al. Tiagabine (Gabitril) in the management of painful diabetic neuropathy. Poster presented at American Pain Society Meeting, Fort Lauderdale, Florida, 1999.
  17. Cato A, Gustavson LE, Qian J, et al. Effect of renal impairment on the pharmacokinetics and tolerability of tiagabine. Epilepsia1998;39:43-47.
  18. Bopp BA. Effect of selected other drugs on thein vitro protein binding of 14C-tiagabine in human plasma: protocol V93-007. Abbott-70569 drug metabolism report no. 35. Abbott Laboratories Division 46 report no. R&D/94/002. Abbott Park, Illinois: Abbott Laboratories, 1994.
  19. During M, Mattson R, Scheyer R, et al. The effect of Tiagabine HC1 on extracellular GABA levels in the human hippocampus. Epilepsia1992;33[Suppl 3]:S83.
  20. Ben-Menachem E, Hedner T, Hamberger A. Cerebrospinal fluid concentrations of tiagabine and 22 amino acids in patients with refractory partial seizures treated for 3 months with tiagabine. Epilepsia1996;37[Suppl 4]:S59.
  21. Borden LA, Miurali Dhar TG, Smith KE, et al. Tiagabine, SK and F 89976-A, Cl-966, and NNC-711 are selective for the cloned GABA transporter GAT-1. Eur J Pharmacol1994;269: 219-224.
  22. Bopp BA, Gustavson LE, Johnson M, et al. Pharmacokinetics and metabolism of [14C]tiagabine HC1 after oral administration to human subjects. Epilepsia1995;36[Suppl 3]:S158.
  23. Bopp BA, Nequist GE, Rodrigues AD. Role of the cytochrome P450 3A subfamily in the metabolism of [14C]tiagabine by human hepatic microsomes. Epilepsia1995;36[Suppl 3]:S159.
  24. So EL, Wolff D, Graves NM, et al. Pharmacokinetics of tiagabine as add-on therapy in patients taking enzyme-inducing antiepilepsy drugs. Epilepsy Res1998;22:221-226.
  25. Richens A, Gustavson LE, McKelvy JF, et al. Pharmacokinetics and safety of single-dose tiagabine HC1 in epileptic patients chronically treated with four other antiepileptic drug regimens. Epilepsia1991;32[Suppl 3]:S12(abst).
  26. Samara EE, Gustavson LE, El-Shourbagy T, et al. Population analysis of the pharmacokinetics of tiagabine in patients with epilepsy. Epilepsia1998;39:868-873.
  27. Gustavson LE, Boellner SW, Granneman GR, et al. A single-dose study to define tiagabine pharmacokinetics in pediatric patients with complex partial seizures. Neurology1997;48:1032-1037.
  28. Murry DJ, Crom WR, Reddick WE, et al. Liver volume as a determinant of drug clearance in children and adolescents. Drug Metab Dispos1995;23:1110-1116.
  29. Snel S, Jansen JA, Mengel HB, et al. The pharmacokinetics of tiagabine in healthy elderly volunteers and elderly patients with epilepsy. J Clin Pharmacol1997;37:1015-1020.
  30. Lau AH, Gustavson LE, Sperelakis R, et al. Pharmacokinetics and safety of tiagabine in subjects with various degrees of hepatic function. Epilepsia1997;38:445-451.
  31. Uthman BM, Rowan AJ, Ahmann PA, et al. Tiagabine for complex partial seizures: a randomized, add-on, dose-response trial. Arch Neurol1998;55:56-62.
  32. Sachdeo RC, Leroy RF, Krauss GL, et al. Tiagabine therapy for complex partial seizures: a dose-frequency study. Arch Neurol1997;54:595-601.