Vigabatrin

Abstract

Vigabatrin was specifically designed to enhance γ-aminobutyric acid (GABA) function in the CNS. By increasing brain concentrations of this inhibitory neurotransmitter the drug appears to decrease propagation of abnormal hypersynchronous discharges, thereby reducing seizure activity. At this stage in its development, clinical experience with vigabatrin is limited primarily to patients with refractory seizure disorders. In this difficult-to-treat population, ‘add-on’ therapy with vigabatrin ⩾ 2 g/day has shown impressive efficacy, reducing seizure frequency by ⩾ 50% in approximately half of patients. Clinical efficacy does seem to vary with seizure type with the best response reported in adults with complex partial seizures with or without generalisation and in children with cryptogenic partial epilepsy or symptomatic infantile spasm. Vigabatrin appears to have a negative effect on absences and myoclonic seizures. Some disorders of motor control may also be amenable to enhanced GABAergic function. In the small number of patients with tardive dyskinesia treated to date, vigabatrin produced mild to moderate improvement in hyperkinetic symptom scores but Parkinsonism or schizophrenic symptoms occasionally worsened. The best response was reported in a study of patients who had been withdrawn from neuroleptic therapy. In a small but well-controlled comparative trial, vigabatrin was as effective as baclofen in reducing spasm and improving some parameters of spasticity in patients with spinal cord lesions or multiple sclerosis. Most adverse reactions to vigabatrin are mild and transient with central nervous system (CNS) changes being reported most frequently. Of particular note, serial evoked potential studies and the few available histology reports have not found evidence of intramyelinic oedema during therapeutic use, as was reported in rats and dogs on chronic high-dose treatment. Thus, vigabatrin is a promising new anticonvulsant drug. Current evidence supports a trial of this agent as adjunctive therapy in patients with refractory seizure disorders, and future investigation of vigabatrin monotherapy and its efficacy relative to established agents is awaited with interest. Wider experience should help to clarify which patients — by seizure type and concurrent CNS pathology — are likely to benefit from vigabatrin and ongoing monitoring should further clarify the potential detrimental effects, if any, of long term use. In the meantime, it is a welcome addition in the difficult setting of resistant epilepsy. Vigabatrin was specifically designed to increase brain GABA levels by inhibiting catabolism of this neurotransmitter. It replaces GABA as substrate for GABA transaminase (GABA-T), but enzymatic activation produces an intermediate which binds covalently to the active site, thereby consuming both enzyme and inhibitor in an irreversible reaction. Only the S(+)-enantiomer of vigabatrin is pharmacologically active. In vitro the effect on other enzymes is very minor and, in particular, vigabatrin does not inhibit GABA synthesis via glutamate decarboxylase (GAD). In v/vo, however, GAD activity is significantly decreased after repetitive dosing in rodents, which is perhaps related to feedback inhibition from the very high levels of GABA. Exposure of cultured mouse neurons to 10 μmol/ L of vigabatrin resulted in a 3.5-fold increase in GABA content. After withdrawal of the drug from culture, GABA-T activity returned to pre-exposure levels after 4 to 6 days, reflecting the time required to synthesise new enzyme. S-vigabatrin is taken up into cultured neurons by a high affinity uptake system, with the result that GABA-T was preferentially inhibited in neurons compared with cultured astrocytes (glial cells). It is probably only the GABA released from the neurotransmitter pool within GABAergic neurons which is important in inhibiting neuron activity, but the majority of GABA catabolism occurs within glial cells; thus, while whole brain levels of GABA reach a peak after about 4 hours, the anticonvulsant effect has a more complex temporal dependency. In rodents, the rise in GABA levels was dose-related and cumulative — the 50% seizure protective dose in audiogenic seizure mice was 990 mg/kg after a single dose and 30 mg/kg after daily administration. Cerebrospinal fluid (CSF) levels of GABA and its conjugate, homocarnosine, rose in a dose-related manner in patients treated with vigabatrin. The decrease in seizure frequency was also dose-related; however, a direct correlation between GABA levels in the CSF and seizure control was not found. Based on results of CSF assays, vigabatrin does not affect the concentration of other neurotransmitters or peptides within the CNS of treated patients. When administered to rodents by intraperitoneal or intracranial injection, vigabatrin was protective against various convulsive stimuli including isoniazid, strychnine, picrotoxin, methionine sulfoxide, pentetrazol (pentylenetetrazol), mercaptopropionic acid and maximal electroshock. The time-course and degree of protection varied between studies and between stimuli. Vigabatrin tended to have a short-lived proconvulsant effect in amygdala-kindled animals followed by a prolonged decrease in susceptibility to the motor component of these seizures. The drug was also anticonvulsant in animals with a genetic susceptibility, including audiogenic seizures and a model of febrile seizures in photoepileptic chicks. Vigabatrin had no effect on, or worsened, EEG seizure activity in rats with spontaneous nonconvulsive (petit mal) epilepsy. Apart from its anticonvulsant properties, treated animals typically exhibit decreased spontaneous locomotion, general sedation, hypothermia and moderate antinociception. Food intake is depressed. In patients with epilepsy treated with vigabatrin long term, no consistent effect was found on serial EEGs or on visual, somatosensory and brainstem auditory evoked potentials. The 2 enantiomers of vigabatrin differ somewhat in their pharmacokinetic properties, but R-vigabatrin (which is inactive) does not interfere with disposition of the active S-enantiomer nor does it undergo chiral inversion in vivo. After oral administration of a single 1.5g dose of racemate, S-vigabatrin reaches a maximum plasma concentration of 114.7 μmol/L at about 1 hour. Extrapolation from intravenous data in animals and urinary recovery data in humans suggests that the drug is almost completely absorbed. Food does not influence absorption. The apparent volume of distribution is 0.8 L/kg and the drug reaches the CNS of patients with epilepsy; CSF concentrations are approximately 10 to 15% of plasma levels. Two hours after administration of 3g of vigabatrin in chronically treated patients, the mean CSF concentration of vigabatrin (racemate) was 4.21 μmol/L. Plasma concentrations follow a biexponential decay and the half-life of elimination of both enantiomers of vigabatrin in healthy volunteers was approximately 7 hours. The drug does not bind to plasma proteins, is not appreciably biotransformed in the liver and does not appear to influence hepatic metabolism. It is eliminated unchanged in the urine with a renal clearance of approximately 1.3 ml/min/kg; 64.6% of the R-enantiomer and 48.9% of the S-enantiomer are recovered from urine within 48 hours of intake of a single 1.5g dose of racemate. In children, disposition of the S-enantiomer was not affected by age. Bioavailability may be somewhat decreased compared with the value in adults but the half-life of elimination and extent of renal clearance appear to be similar. Elimination of vigabatrin is impaired in individuals with a creatinine clearance < 60 ml/min. Excessive plasma concentrations of vigabatrin are associated with adverse CNS effects (excess sedation, confusion, ataxia); thus, dosage adjustment in patients with diminished renal function is appropriate. At this stage of clinical development, experience with vigabatrin as an antiepileptic agent has been confined to the treatment of patients with refractory epilepsy. Adjunctive therapy with vigabatrin in this population resulted in substantial improvement in seizure control (⩾ 50% reduction in frequency) in approximately half of adults given ⩾ 2 g/day. Some individuals with a less complete response nonetheless benefited in terms of improved quality of life via a decrease in seizure severity or duration and improvement in general well-being. A similar response rate was also reported in the few paediatric trials available. Response does seem to vary depending on the seizure syndrome with the best results reported in adults with complex partial epilepsy with or without generalisation, and in children with cryptogenic partial epilepsy or symptomatic infantile spasm. Seizure frequency tended to increase in patients with absences or myoclonic seizures. Because vigabatrin was used in conjunction with the patients’ previous antiepileptic drug regimen, additive effects cannot be excluded. However, in double-blind studies in this difficult-to-treat population, vigabatrin exhibited impressive efficacy and the response to treatment was sustained in the majority of responding patients when followed up for periods of > 1 year. Vigabatrin has been investigated in a small number of patients with tardive dyskinesia. A substantial (44%) improvement in hyperkinetic symptom scores was reported in 1 study of patients who had been withdrawn from neuroleptic drugs. Results in patients still receiving neuroleptic therapy were less impressive, with occasional worsening of Parkinsonism or schizophrenia. Treatment of individuals with degenerative ataxias has not been rewarding. Preliminary results in patients with spasticity secondary to multiple sclerosis or spinal cord lesions suggest that vigabatrin 2 to 3 g/day can ameliorate spasm and produce mild to moderate improvement in spasticity in a substantial number of patients. Equal therapeutic benefit was achieved during treatment with vigabatrin or with baclofen in a well-controlled comparative trial. Most adverse effects are related to the CNS. Sedation and fatigue are reported most frequently (13 and 9% incidence, respectively) with other CNS effects (e.g. headache, dizziness, confusion, ataxia, diplopia, memory impairment and insomnia) reported at rates of < 4%. Generally, all these effects are mild and transient but drug withdrawal or dosage reduction may be necessary in some instances, particularly elderly patients with reduced renal clearance. Limited experience in children suggests that 8 to 30% may exhibit agitation, insomnia and behaviour changes which are reversible with dosage reduction. Adverse psychiatric events which may be categorised as behaviour changes or anxiety are reported in 4 and 6%, respectively, of patients overall during the initial weeks of treatment, but tolerance to these effects appears to develop rapidly. Aggression or frank psychosis is a serious finding in a few patients, particularly in those with a history of behaviour disturbances, mental retardation or psychiatric illness. Rapid and extreme changes in seizure frequency secondary to vigabatrin have also been suggested as permitting or provoking the expression of psychotic behaviour in susceptible individuals. Long term treatment with vigabatrin is generally very well tolerated. In particular, serial evoked potential studies and the few available reports of brain histology in treated individuals do not support the development of intramyelinic oedema during clinical use of vigabatrin as was noted in nonprimate toxicology studies. As there are few data available on the use of vigabatrin in pregnancy, it is currently contraindicated in pregnant women. No consistent, clinically significant effect on the plasma concentrations of most concurrently administered anticonvulsants — notably carbamazepine, valproic acid (sodium valproate), phenobarbital (phenobarbitone), primidone, and clonazepam — was observed in patients treated with these drugs in combination with vigabatrin. Phenytoin is the exception. Plasma concentrations of this agent are decreased 20 to 30% by vigabatrin, which in a few individuals was felt to have compromised seizure control. The mechanism has not been elucidated. The usual adult dosage is 2 to 3 g/day given orally in 1 or 2 divided doses. As required by clinical response, the dosage may be increased in increments of 0.5 to 1 g/day to a total dosage of 4 g/day or 50 mg/kg/day. Lower starting dosages and slow upward titration may be appropriate in patients at risk for adverse psychiatric events. Higher dosages do not improve the clinical response and there is no direct correlation between plasma concentrations and clinical effect. Sedation is common in patients beginning treatment and they should be cautioned about activities requiring mental alertness. The recommended dosage in children is 1 g/day if aged 3 to 9 years and 2 g/day in older children. Dosage reduction should be considered in patients with a creatinihe clearance < 60 ml/ min because of the association between excessive plasma concentrations of vigabatrin and adverse CNS effects. Vigabatrin should be tapered off when therapy is discontinued in epileptic patients.