J7ournal ofNeurology, Neurosurgery, and Psychiatry 1993;56:1271-1275 1271 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1271 on 1 December 1993. Downloaded from Effect of vigabatrin on sedation and cognitive function in patients with refractory epilepsy R A Gillham, J Blacklaw, P J W McKee, M J Brodie Abstract Vigabatrin was the first 'designer' drug to be Twenty-four patients with refractory used for the treatment of epilepsy. It is an epilepsy on one or more antiepileptic irreversible suicide inhibitor of y-aminobu- drugs were given additional vigabatrin tyric acid (GABA) transaminase, the enzyme (1 g twice daily for six weeks, followed by responsible for the breakdown of the 1-5 g twice daily for a further six weeks) inhibitory neurotransmitter GABA.1 Thus, it and matched placebo in a double blind, increases the concentration of GABA in the randomised, crossover study. A battery brain. Efficacy trials have been largely con- of neuropsychological tests was adminis- fined to patients with refractory epilepsy tered at baseline and at weeks two, six already receiving other antiepileptic drugs. and 12 of both treatment periods. No sig- Overall, a 50% reduction in frequency of the nificant differences were found between seizures has been reported in around 50% of vigabatrin and placebo at any time point treated patients.24 The benefits were particu- Department of for any of the objective tests of cognitive larly apparent in patients with partial seizures Clinical Psychology, Institute of function. Patients, however, reported a (unpublished observations).5 Good, long term Neurological Sciences, greater degree of sedation after two and efficacy has been reported with the drug over Southern General six weeks on vigabatrin than during the a number of years.f5 Hospital and Epilepsy Research Unit, equivalent placebo phase (p < 0 01), The safety of vigabatrin has come under University although no such difference was appar- particular scrutiny because it produced Department of ent at 12 weeks. Follow up over a mean of intramyelinic oedema in the dog and rat,l617 Medicine and Therapeutics, 14-75 months in 12 responders, who con- although there is no evidence in support of a Western Infirmary, tinued on vigabatrin, revealed a signifi- similar lesion in man.4 '>'9 Vigabatrin is Glasgow, UK cant improvement (all p < 001) on each known to be sedative in some patients, partic- R A Gillham J Blacklaw of three composite scales (three psy- ularly during early treatment.4 Few objective P J W McKee chomotor tests, four memory tests, three data are available on cognitive function, M J Brodie self rating scales) compared with their although one non-randomised, non-blinded Correspondence to: scores during the double blind trial. study with a parallel control group did not Dr M J Brodie, Epilepsy Research Unit, Department Vigabatrin did not cause cognitive support a major deleterious effect with the of Medicine and impairment either acutely or in the long drug.20 In this study, our standard battery of Therapeutics, Western Infirmary, Glasgow GIl term. Phased introduction, however, cognitive function tests21-23 was administered 6NT, United Kingdom seems a prudent policy to allow tolerance to patients with refractory epilepsy who were Received 31 July 1992 and to early subjective sedation. taking part in an add-on, dose ranging, in revised form 15 January 1993. placebo controlled, crossover trial (unpub- Accepted 1 February 1993 (7 Neurol Neurosurg Psychiatry 1993;56: 1271-1275) lished observations). Measures were made in http://jnnp.bmj.com/ the short term and after prolonged adminis- Table 1 Clinical characteristics of24 patients with refractory epilepsy tration of vigabatrin in responders. Age Duration Patient Sex (years) Seizure type (years) Treatment 1 F 24 GTCS 16 CBZ 2 F 53 CPS 43 CBZ Methods 3 M 27 CPS/GTCS 8 CBZ/VPA PATENTS 4 F 38 CPS/GTCS 24 CBZ on September 24, 2021 by guest. Protected copyright. 5 F 24 CPS/GTCS 24 CBZ/PRIM Twenty-four patients (16 women, 8 men; 6 F 17 CPS/GTCS 6 CBZ 7 F 31 CPS 30 CBZ/PRIM aged 17-53 years) with refractory epilepsy 8 F 45 CPS 42 PHT/PB entered the trial (table 1). Fourteen suffered 9 F 46 CPS/GTCS 22 VPA/PB 10 F 27 CPS 20 PHT complex partial seizures with secondary gen- 11 F 36 CPS/GTCS 33 CBZ/VPA eralisation (PSG), eight had complex partial 12 F 34 CPS/GTCS 23 CBZ 13 F 20 CPS/GTCS 9 CBZ/VPA seizures (CPS) alone and two reported only 14 F 32 CPS 30 PHT/PRIM generalised tonic-clonic seizures (GTCS). 15 M 49 CPS 27 CBZ/PRIM 16 M 22 GTCS 20 CBZ/PHT Each averaged at least one seizure per week 17 M 43 CPS/GTCS 8 CBZ/PRIM during the three months before recruitment. 18 M 26 CPS 19 CBZ Eleven patients were being treated with anti- 19 F 24 CPS/GTCS 4 CBZ convulsant monotherapy (nine took carba- 20 M* 26 CPS/GTCS 6 CBZ 21 M* 35 CPS/GTCS 9 CBZ/PHT mazepine, one phenytoin, and one sodium valproate) and the other 13 took two drugs in 22 Mt 44 CPS/GTCS 9 CBZ/PRIM six 23 Ft 27 CPS 7 VPA combination (10 had carbamazepine, 24 Ft 39 CPS/GTCS 36 CBZ primidone, four sodium valproate, four two All Patients excluded from seizure analysis. phenytoin, and phenobarbital). gave tPatients who failed to complete the study. written consent to their participation in the GTCS = generalised tonic-clonic seizures; CPS = complex partial seizures; CBZ = carba- which had the approval of the local mazepine; VPA = sodium valproate; PHT = phenytoin, PRIM = primidone; PB= pheno- trial, barbital. ethical committee. 1272 Gillham, Blacklaw, McKee, Brodie J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1271 on 1 December 1993. Downloaded from PROTOCOL level and the task was discontinued when The study was a double blind, randomised, both were failed. crossover, add-on trial of two consecutive The backward digit span measured the doses of vigabatrin and matched placebo. maximum number of digits the patient could After an initial four week run-in, two 12-week recall in reverse order immediately following treatment periods (phase one: six weeks of oral presentation. He or she was allowed two 1 g vigabatrin/matched placebo twice daily; trials at each level and, again, the task was phase two: six weeks of 1-5 g vigabatrin/ discontinued when both were failed. matched placebo twice daily) were separated The paired associate learning test mea- by a four-week washout. Neuropsychological sured the number of trials to reach the crite- testing was performed immediately before rion of three correct answers, when learning each schedule and after two, six and 12 weeks three, unrelated word pairs. of treatment. Patients were given individual In the Rivermead behavioural memory test, appointment times, which were kept constant the 'screen score' of this standardised psycho- throughout the trial. Patients were asked to metric battery was used. take their medication at the same times each day. Seizures were recorded using standard Selfreporting scales description and frequency charts with which The general health questionnaire was the the patients were familiar. At each hospital standard 28 question version, which gave a visit, compliance with medication was 'cut off' point for psychiatric cases. checked by a tablet count. The sedation score assessed the patient's rating of level of alertness using a 10 cm line TEST BAT-tERY where zero was 'wide awake' and 10 'nearly The test battery consisted of measures of ver- asleep'. bal and performance IQ, psychomotor func- The patient was also presented with a list tion, visual attention, memory, and subjective of 10 commonly reported antiepileptic drug side effect rating scales as below: side effects-for example, dizziness, headache, and nausea. He or she rated each Intelligence quotients one on a 10 cm line with 'don't have it' at The verbal IQ was calculated from the score one end and 'very severe' at the other. The on the National Adult Reading Test. The distances from the left hand side of the scales performance IQ was calculated by a standard were measured in centimetres and summated method from the score on Raven's standard to produce a subjective side effect score. progressive matrices. DRUG ASSAYS Psychomotor tests Blood was withdrawn at each review appoint- The decision time was the time in milli- ment for measurement of antiepileptic drug seconds to respond to a light coming on by concentrations. Samples were centrifuged removing the finger from the base button in a immediately and plasma stored at -20° for choice reaction time task. The mean of 30 batch analysis. Vigabatrin was extracted into trials was recorded. The movement time was ethyl acetate from plasma heated with dansyl the time in milliseconds to move the finger chloride to form a fluorescent derivative at from the base button to extinguish a light. high pH and quantified using high pressure http://jnnp.bmj.com/ The mean of 30 trials was recorded. liquid chromatography with phenyl-GABA as In another test, an array of small rectangles the internal standard. The interassay coeffi- was displayed on a visual display unit. After a cient of variation over the range 1-100 mg/l brief period an extra rectangle was added to was 5% and the lower limit of detection was the array. The patient was required to indi- 0 1 mg/l. Concentrations of other anti- cate which it was. The 'threshold' was the epileptic drugs were measured by enzyme minimum time gap in frame units between immunoassay (Emit, Syva, Palo Alto, United the presentation of the array and the addi- States).