Monotherapy Antiepileptic Drug Trials in Patients Undergoing Presurgical Assessment: Methodological Problems and Possibilities

Seizure 1995; 4:293-301

Monotherapy antiepileptic drug trials in patients undergoing presurgical assessment: methodological problems and possibilities

G. ALARCON, C.D. BINNIE, R.D.C. ELWES & C.E. POLKEY

Institute of Epileptology, The Maudsley Hospital, Denmark Hill, London SE5 8AZ Correspondence to: Dr. G. Alarcon, EEG Department, The Maudsley Hospital, Denmark Hill, London SE5 8AZ, UK

Key words: clinical trials; placebo; presurgical assessment of epilepsy; anticonvulsants; seizures.

INTRODUCTION reduce apparent efficacy. (2) Induction of metabolism by the comedication may produce lower It may fairly be claimed that up to the last decade than expected blood levels and result in a failure no antiepileptic drug (AED) had undergone to demonstrate any effect of the experimental rigorous testing. Coatsworth I in a comprehensive drug. (3) Efficacy and adverse events may depend review of AED assessment reported three con- on a pharmacodynamic synergy with the trolled trials up to 1971, and Richens (1976) was comedication. able to identify some 17. Gram and coworkers A number of alternative trial designs are now found 51 by 1982, but the majority suffered gross available and increasing attention has recently methodological deficiencies. In the past decade been directed to ethically acceptable monothe- the situation has changed dramatically, the rapy designs7. One approach, first adopted by number of controlled AED trials currently in Bourgois et al 8, in the development of Felbamate, progress must exceed the total of those completed is the performance of monotherapy trials in before 1983. The development programmes of patients whose AEDs had been withdrawn to the new AEDs registered in recent years have facilitate capture of seizures by telemetry, as part necessarily been innovative, and methods of of a preoperative assessment protocol. The AED testing are still undergoing rapid evolution- conduct of such trials is difficult and complex. The ary change. Conventional phase II AED trials are patients are in an unstable state and there may be usually placebo controlled studies employing only a brief interval between capturing sufficient designs of either single period parallel groups or seizures to meet the needs of preoperative multiple periods within patients. Since it is, in assessment and the onset of serial seizures general, unacceptable to withhold effective treat- demanding immediate control. To be of demons- ment from a patient with epilepsy, during trable efficacy under these circumstances, the test conventional AED trials the experimental drug is drug must act swiftly. If the onset of drug action is added to the presumably partially effective delayed by one or two days, the opportunity to medication already being taken (add-on trials). prove an effect may be missed. This may happen Although widely regarded as an ethical necessity, if the half life is long or efficacy depends on an this approach poses a number of problems which active metabolite or if there are constraints on the often makes this design difficult to interpret: (1) rate of dose escalation. In many patients, it is Drug interactions may confound both therapeutic impossible to achieve monotherapy if seizure and adverse effects. Elevation of blood levels of frequency increases rapidly before existing medi- comedication may cause intoxication, which may cation is fully withdrawn8. be wrongly attributed to the experimental agent4"5 We report two variants of this design: an open and efficacy can be wrongly ascribed to a new trial of remacemide hydrochloride (Fisons CR product because it causes an increase in con- 2083), and a controlled trial of tiagabine (Abbott centrations of comedication6. Conversely in- M90-511 TIA-102) in which the test treatment creased clearance of the comedication may was introduced prior to withdrawal of other

1059-1311/95/040293 + 09 $12.00/0 @ 1995 British Epilepsy Association 294 G. Alarcon et al medication. This last approach may avoid many an active metabolite in the rat and has also been of the difficulties described above, but will identified in the dog and man m'~l. In healthy prolong the period of telemetry in patients who volunteers the drug has been generally well respond to the active treatment. For reasons tolerated at single doses of up to 400mg or at which will be apparent, both were inconclusive 150 mg q.i.d, for six days. Minor adverse events but they are reported here as illustrating some of (gastrointestinal upset, nausea and lightheaded- the methodological possibilities and problems of ness) were observed at higher doses. The this novel type of trial design. mechanism of action of remacemide hydroch- Ioride has not been fully elucidated but it appears to differ from that of other AEDs. The compound REMACEMIDE HYDROCHLORIDE TRIAL has been shown to enhance the survival of mice in a hypoxic environment and to exert a neurop- The drug rotective effect against global and focal cerebral ischaemia in rodents and higher species 12. When administered orally in rodents, remacemide hydrochloride protects against maxi- mal electroshock induced seizures with a potency Patients similar to that of phenytoin, phenobarbitone or carbamazepine and greater than valproate. It is Ten male patients aged between 19-41 years also effective against absence-like seizures in (average, 25.7) entered the study. Table 1 rats 9. The desglycinyl derivative of remacemide is summarizes the demographic and pretreatment

Table 1: Patient characteristics and seizure history of patients taking part in the remacemide trial Mean value or frequency

Number of patients: 10 Age (years): Mean 25.7 Range 19-41 Sex: Males l0 Females 0 Duration of epilepsy (years): Mean 19.4 Range I 1-35 Etiology: Febrile convulsions 2 Birth injury I History of head injury l History of haemorrhage l Idiopathic 1 Unknown 4 Number of seizures per month: Mean 22.4 Range 4-84 Types of seizures experienced: Simple partial 6 Complex Partial 9 Complex partial with secondary generalisation 6 Generalised tonic clonic 1 Number of seizure types: Three types 3 Two types 5 One type 2 Standard Anticonvulsants: Carbamazepine 6 Vigabatrin 5 Sodium valproate 4 CIobazam 4 Phenytoin 2 Primidone 2 Clonazepam 2 Monotherapy AED trials in patients undergoing presurgical assessment 295

characteristics of the patients. Patients were Throughout this phase patients were resi- capable of giving informed consent in writing, and dent in the telemetry monitoring facility of the had confident diagnosis of epilepsy supported by Maudsley Hospital, under continuous nursing a previous EEG workup. Exclusion criteria were supervision and video monitoring. The duration the existence of any significant medical history or of Phase A was five days, or until 12 hours laboratory abnormalities other than those at- following discontinuation of the drug--whichever tributable to the concomitant AEDs: alcohol was shorter. At the end of Phase A, patients consumption greater than 21 units per week: were assessed for their eligibility to enter history of chronic drug abuse, heavy smoking Phase B. (more than 15 cigarettes a day); chronic use of non-steroidal anti-inflamatory drugs or antihis- tamines; history of poor compliance and/or pseudoseizures. Phase B

Study design Phase B was the continuation of the studv from the end of Dav 5 (i.e. Phase A) to dav 28. After a pre-study screening and seizure docum- Only patients completing all 5 days of Phase A entation over at least four weeks, the withdrawal were to be considered for Phase B. During of AEDs and presurgical EEG telemetry, eligible Phase B patients were discharged home but patients received remacemide hydrochloride were to be evaluated on at least four occasions. 150mg q.i.d, for a maximum of 28 days. The The patient kept a daily record of seizure study treatment was administered initially as frequency and type, adverse events, and con- monotherapy but in some patients continuing comitant medications throughout this phase. seizures required the reinstitution of their stan- During Phase B, patients would be on re- dard antiepileptic drugs. The study was carried macemide hydrochloride monotherapy, or on out in two phases: Phase A (first five days) and remacemide hydrochloride plus standard AED Phase B (6 to 28 days). comedication, or on remacemide hvdrochloride monotherapy at day 6, followed by addition of standard AED at any point during Phase Phase A B. Remacemide hydrochloride was reduced gradually over four days at the end of phase Patients satisfying the study entry criteria re- B in order to reduce the risk of withdrawal ceived remacemide hydrochloride, as above. seizures.

Table 2: Total seizure frequency during Phase A of the remacem~de trial Patient Baseline Phase A number (Day -1) Day 1 Day 2 Day 3 Day 4 Day 5

1 7 4.6* .... 2 2 6 7.7* ------3 7 7.2* .... 4 3 I 3 2 0 19.2" 5 I 0 0 I I I 6 11 2 1 Manv Many -- 7 7 3 I 1 _.t~~ '* 8 1 2 I 2 1.5" -- 9 ...... It) II 2 I - - -

Comparison of Baseline vs Day I: P = 0.138 Comparison of Baseline vs Day 2: P = 0.198 * Estimated by linear interpolation from incomplete 24 hour periods of at least 12 hrs. 296 G. Alarcon et al

Table 3: Summary of adverse events reported during the remacemide trial Number of patients experiencing event

Baseline Phase A Phase B Follow-up

Musculo-skeletal: Myalgia 0 0 0 Nervous system: Dizziness 0 0 0 Headache 2 0 1 Hyperaesthesia 0 0 0 Hypoaesthesia 0 0 0 Paraesthesia 0 0 0 Vision: Vision abnormal 0 0 0 Psychiatric: Anxiety 0 0 0 Confusion 0 0 0 Somnolence 0 0 0 Thinking abnormal I 0 0 Digestive system: Vomiting (I I 0 0 Nausea 0 I 0 0 Respiratory system: Bronchitis 0 0 1 0 Hyperventilation 0 1 0 0 General symptoms: Fatigue I 0 0 0 Fever 1 0 0 0 Influenza-like symptoms 1 0 0 0 Malaise 1 2 0 0 Pain 0 0 0 1 Other medical events: 2 1 0 0 Total number of patients I0 10 1 l0 Number of patients experiencing events in each period 6 7 I I

Results differences between the two study days and baseline. It was not possible to analyse simple Efficacy partial and generalized seizure frequencies due to insufficient data. Seizure frequency was recorded at baseline and throughout the study period up to and including day 5 (Table 2). Additional data after this time Safety were available only for patient 5, who completed Phase B of the study. Seizure frequency for this Of the 10 patients who received remacemide patient did not appear to alter significantly during hydrochloride, only one patient received the drug this phase. Total seizure frequency on days 1 and for the complete study period of 28 days. The 2 ~vere compared with baseline (day -1) using a remainder stopped study treatment between one Wilcoxon Matched Pairs Signed Rank test. and five days after the start of dosing, due to Statistical analysis of data on days 3, 4 and 5 was adverse events or deterioration of epilepsy (Table not possible due to insufficient sample size 3). Nine patients experienced adverse events (caused by patient withdrawals). A total of nine during the study. Six patients experienced events patients were included in an analysis of total during the baseline period, seven during Phase A seizure frequency, which compared day 1 of the of the treatment period, one patient during Phase study with baseline (day -1) and a total of seven B and one patient during the follow-up period. patients were included in an analysis of total The most commonly experienced event during seizure frequency which compared day 2 of the the treatment period was headache (three pati- study with baseline (day -1). Although a ents). No serious events were reported during the reduction of median seizure frequency was seen study. Seven patients experienced events which on days 1 and 2 when compared with baseline, were thought to be possibly related to test this difference did not achieve statistical sig- treatment and included myalgia, malaise and nificance. Analysis of complex partial seizure dizziness, hyperaesthesia, vomiting, confusion, frequency and secondarily generalized seizure hypoaesthesia, nausea, anxiety, paraesthesia and frequency also showed no statistically significant hyperventilation, and abnormal vision. Monotherspy AED trials in patients undergoing presurgical assessment 297

TIAGABINE TRIAL human chorionic gonadotrophin pregnacy test at day -1. Patients taking primidone or phenobar- The drug bital had this medication discontinued prior to admission to the study. The plasma phenobarbital Tiagabine hydrochloride (HCI) is a potent and concentration was required to be -< 10/zg/ml specific inhibitor of GABA uptake by glial and prior to day -1. neuronal elements in vitro. Following administra- Exclusion criteria included: history of relevant tion by intraperitoneal injection it has potent psychiatric illness, progressive CNS disease, other anticonvulsant activity in animal models. In medical disease, non-compliance with medication volunteers, adverse events can be observed at or medical advice, recent history of alcoholism, doses above 12 mg/day, including dizziness, drug abuse or addiction, consumption of 30 or headache, somnolence, abnormal vision, poor more units of alcohol per week, significant concentration, incoordination, stupor, confusion laboratory abnormalities other than those at- or nausea 13. tributable to antiepileptic drugs.

Patients Study design

Eleven patients entered the study with ages The study consisted of an open phase (28 days), a between 17 and 52 (mean, 32.8) years. The mean double-blind treatment phase (seven days) and a age for the patients in the tiagabine HCL and in washout phase (24 hours). During the screening the placebo group was 38.3 and 23.3 years, visit of the open phase, patients provided written respectively. Patient characteristics are shown in informed consent and a medical history, under- Table 4. Patients were capable of signing an went a physical examination and laboratory tests. informed consent, were hospitalized during the Within four weeks following the Screening Visit, double-blind and washout phases, and had a eligible patients were admitted to the hospital, diagnosis of epilepsy with complex partial se- where they remained for the duration of the izures supported by observed ictal events docu- study. On day -1 of the open phase (the day mented by reliable witnesses and confirmed by at before starting the double-blind treatment least one of the following: (1) ictal electroen- phase), an interval medical and seizure history, cephalogram (EEG), (2) interictal EEG demons- and laboratory tests were obtained. During the trating epileptiform activity, (3) computed tom- seven-day double-blind phase, randomized pati- ography or magnetic resonance imaging showing ents were dosed every four hours with the a focal cerebral lesion consistent with complex tiagabine HCI or matching placebo. Assignment partial seizures. Additional requirements in fe- to the treatment group was balanced (1:1) male patients were at least one of the following: between placebo and tiagabine HC1. Patients had (1) being surgically sterilized, (2) at least one their dose of study treatment titrated by a year postmenopausal, (3) an intrauterine device physician blinded to medication to achieve either in place for at least one month, (4) use of an oral the maximum well tolerated dose or 66 mg/day contraceptive for at least one menstrual cycle, (or equivalent number of placebo capsules). and for the duration of the study; (5) a negative Concomitant AEDs were discontinued with start

Table 4: Patient characteristics of the patients who took part in the tiagabine trial Tiagabine Placebo Total Characteristic N = 7 N = 4 N = 11

Age (years): mean (sd) 38.3 (9.96) 23.3 (6.13) 32.8 (11.3) range 24-52 17-31 17-52 Sex: males 5 3 8 females 2 1 3 Race: Black 0 1 0 Caucasian 7 3 10 Weight (Kg): mean (sd) 75.4 (15.5) 74.8 (14.9) 75.2 (14.5) range 50-96 63-95 50-96 Height (cm): mean (sd) 173.1 (9.92) 179.8 (10.6) 175.5 (10.2) range 157-184 167-189 157-189 298 G. Alarcon et at

of study medication. Tiagabine HCi (or placebo) placebo) experienced two convulsive seizures was then administrated in six daily doses and within three hours. One patient (on Tiagabine) titrated with gradually increasing doses, starting experienced more than seven seizures in a 48 with 2 mg/dose and increasing 1 mg/dose every hour period with a baseline rate of 0.32 three to five doses. For the next eight days, seizures/day. One patient (on placebo) was patients' interval seizure histories, adverse ev- discontinued because of protocol violation (not ents, AED and study drug plasma concentrations, keeping a seizure diary during the open phase). and laboratory values were assessed periodically. Two patients (on Tiagabine) were discontinued On day 8 study medication was stopped. Follow- because of an adverse event. ing a 24-hour washout phase (day 9) patients received a final evaluation, their original AEDs were reinstituted, and patients were discharged Drug administration from the study. Escape criteria to stop drug administration during the double-blind phase Tiagabine or placebo were administered ap- were established a priori and included: convulsive proximately every four hours on the dosing days. status epilepticus, three convulsive seizures Of the patients on tiagabine HCI, only two were within any 24 hours or two convulsive seizures administered the maximum study drug dose of 66 within any three hours. In addition, escape mg/day, both of whom successfully completed the criteria depending on each patient's baseline rate study. Four other patients received maximum (number of seizures in the open phase divided by study drug doses of 42 mg/day, 41 mg/day, 28 28) were established. mg/day and 25 mg/day. All patients stopped all Baseline rate -<0.25: drug administration was marketed antiepilepsy drugs (AEDs) on or prior stopped if the patient had more than four to day 1 except for one patient who took seizures/day or five seizures/48 hours. vigabatrin thoughout the study. Baseline rate >0.25 and <--0.50: drug administration was terminated if the patient had more than five seizures/day or more than seven seizures/48 hours. Efficacy Baseline rate >0.50 and <-I.00: drug administration was terminated if the patient had more than While the sample size was too small for statistical six seizures/day or more than eight seizures/48 significance to be attained, the results suggest hours. efficacy of tiagabine HCi for complex partial, Baseline rate >1.00: drug administration was simple partial, combined partial and secondarily terminated if the patient had six or more times generalized tonic-clonic seizures because the the baseline rate seizures/day or four or more patients in the tiagabine HCI group experienced times the baseline rate seizures/48 hours. fewer seizures than the placebo group in the Eligible patients had the option of entering an double-blind period. Based on median seizure open-label tiagabine HCL extension study (Ab- rates, there was no seizure type in either group bott M91-595 TIA-104). that lessened in frequency during the double- blind period but the tiagabine HCI group had less increase in seizures of all types (Table 5). Results Complex partial seizures. During the open phase, tiagabine HCI-treated patients experienced 0.178 Eleven patients were enrolled, seven received more seizures per 24 hours than placebo-treated tiagabine HCI and four received placebo. Three patients based on median seizure rates. During patients completed the study. the double-blind phase, tiagabine HCl-treated patients experienced a median of 0.571 fewer seizures per 24 hours than placebo-treated Premature discontinuation patients. Both groups, however, did have an increase in seizures over baseline. Eight patients were prematurely discontinued from the study. Five patients were discontinued Simple partial seizures. During the open phase, because the number of seizures exceeded the tiagabine HCl-treated patients experienced 0.054 escape criteria. Two patients (one on Tiagabine fewer seizures per 24 hours than placebo-treated and one on placebo) experienced three convul- patients. During the double-blind phase, tiaga- sive seizures within 24 hours. Two patients (on bine HCl-treated patients experienced median Monotherapy AED trials in patients undergoing presurglcal assessment 299

Table 5: Seizure frequency of 24-hour seizure rates in the tiagabine trial Seizure rate per 24h Tiagabine Placebo

N Mean sd Median N Mean sd Median

Complex partial: Baseline 7 0.368 0.163 0.321 4 0.170 0.135 0.143 Double blind phase 7 1.088 0.723 1.154 4 1.771 0.357 1.725

Simple partial: Baseline 5 0.229 0.258 0.071 2 0.125 0.126 0.125 Double blind phase 5 0.635 0.949 0.286 2 2.657 0.486 2.657

Combined partial: Baseline 7 0.389 0.154 0.321 4 0.188 0.150 0.161 Double blind phase 7 1.357 0.908 1.254 4 2.103 0.711 2.014

Secondarily generalized tonic-clonic: Baseline 4 0.152 0.074 0.161 4 0.018 0.021 0.018 Double blind phase 4 0.911 0.958 0.711 4 1.699 0.394 1.580

2.371 fewer seizures per 24 hours than placebo Safety treated patients.

Combined partial seizures. During the open Nine patients (82%) experienced at least one phase, tiagabine HCl-treated patients ex- treatment-emergent adverse event. The most perienced 0.160 more seizures per 24 hours than frequent adverse events were related to the placebo-treated patients. During the double-blind nervous system (Table 6). At least one treatment- phase, tiagabine HCl-treated patients ex- emergent adverse event was reported by 7/7 perienced median 0.760 fewer partial seizures per (100%) patients receiving tiagabine HCl and by 24 hours than placebo treated patients. 2/4 (50%) patients receiving placebo. All adverse Secondarily generalised tonic-clonic seizures. events were considered to be mild to moderate in severity. Ten patients reported headache and During the open phase, tiagabine HCl-treated nine patients reported cheek pain on day -1. patients experienced 0.143 more seizures per 24 These were attributed to the insertion of foramen hours than patients on placebo. During the ovale electrodes. Two of the 11 patients (18%) double-blind phase, tiagabine HCl-treated pati- were prematurely discontinued during the study ents had a median of 0.869 fewer seizures per 24 because of adverse events. Both patients had hours than patients on placebo. been on tiagabine HC1. One patient was discontinued because of moderate nausea on day 3 and Table 6: Summary of adverse events reported during the another patient was discontinued because of tiagabine trial moderate depression on day 5 and withdrawn Number of patients experiencing consent. the event

Tiagabine Placebo Total n = 7 n = 4 n = 11 DISCUSSION Any adverse event 7 2

Digestive system: Conventional double-blind add-on trials can Mouth ulceration 1 0 be expected to show little difference between Nausea 1 0 placebo and novel drugs unless the latter Nervous system: are significantly more effective than the exis- Depression 1 0 1 ting standard treatment j4. While a difference Diplopia 1 0 1 greater than 50% is regarded as indicative of Dizziness 1 1 2 Headache I 2 3 effectiveness in most add-on trials, this outcome Insomnia 0 1 1 is unlikely even if the tested drug is known to Paraesthesia 2 1 3 be effective, as shown by Schmidt 15 who found 'Feeling outside an improvement in only 15% of patients when this world' l 0 l a second marketed drug was added to the regime. 300 G. Alarcon et al

Monotherapy trials during presurgical assessment ment allow only a short time available to show are potentially a good alternative to complement therapeutic effects. This is an important limitation add-on studies both to establish efficacy and to since absolute seizure frequency can increase justify early initiation of more conventional during the trial even if the novel drug is effective monotherapy trials. Pharmacological interactions and in this situation it is difficult for the blinded are avoided and since no other therapeutic agents physician to justify continuing the trial. Objective are simultaneously present, there is not a basal escape criteria must be applied to ensure with- seizure control over which the new agent has to drawal of patients at risk of harm from continuing demonstrate further efficacy. This implies that in the trial (see Material and Methods of agents which are at least as active and effective in Tiagabine trial). These criteria should be prag- the same patients as already marketed drugs can matic and take account of the hazards associated in principle show an effect. with different types of seizure and of the patients' There are however a number of problems, both habitual seizure type and severity. theoretical and practical, that complicate the As with add-on trials, presurgical monotherapy interpretation and conduct of monotherapy trials trials will include patients with particularly during preoperative assessment. refractory epilepsy. If a drug does not show Sudden withdrawal of antiepileptic medication efficacy in presurgical patients it could still be can be associated with an increase in the effective in other milder epilepsy. Both tiagabine incidence of seizures ~6 and it is in general not HCI and remacemide HC1 have proven efficiency clear the extent to which this might be due to in traditional design trials ~°'~7"18. halting effective therapy or to rebound with- Experiencing a greater than usual seizure drawal effects. The influence of withdrawal frequency over a short period of time is seizures on results can be minimized by using necessarily disagreeable. Moreover, in many parallel placebo controlled designs. Although the centres invasive procedures are often used during absolute incidence of seizures on the test drug can presurgical assessment. These factors combine to exceed baseline values after withdrawal of stan- give rise to adverse events which can be difficult dard antiepileptic medication, even if the novel to distinguish from those induced by the novel drug is effective, such increment would be drug. For instance most of our patients had expected to be higher in patients on placebo. undergone implantation of foramen ovale electr- Indeed we found that seizure frequency increased odes under general anaesthesia which is often during tiagabine treatment after withdrawal of associated with headache, nausea and facial pain standard antiepileptic medication but the incre- lasting for several days, and often overlapped ment was higher in patients on placebo for all with trial day 1. Headache is often of several days seizure types, which suggests that tiagabine duration and was thus indistinguishable from possesses a degree of effectiveness although the headaches possibly induced by the experimental small number precluded formal statistical analy- drug. Some other adverse events are more likely sis. No such conclusion would have been possible to be drug specific (drowsiness, tremor, ataxia) if tiagabine had been tested on an open-label and can be appropriately monitored to study drug basis, as in the case of remacemide hydrochloride. safety during monotherapy trials. The pressure of Consequently., no comparisons can be drawn time creates the need to escalate the dose of the between the relative efficacy of tiagabine and experimental dose faster than would be normal remacemide hydrochloride since the study de- practice; this may lead to an incidence of signs were disimilar. Whether suppression of drug-induced adverse experiences greater than withdrawal seizures is predictive of efficacy in would be encountered with routine use. This epilepsy may be disputed. An alternative ap- problem would be alleviated by introducing the proach which avoids this last problem would be to study treatments more gradually prior to with- randomize patients and start experimental treat- drawal of comedication. ment whilst gradually tapering off standard Sample size in monotherapy trials during antiepileptic medication during several days surgical assessment will in general be smaller than before surgical assessment. Phenobarbitone, in more conventional trials. This means that whether adminstered as such or arising as a seizure reduction would have to be greater in metabolite of primidone, has a long half life order to achieve statistical significance. For (24-48 hours) and primidone administration was instance, in order to obtain an 80% chance of consequently stopped between 48 and 72 hours finding existing differences at a 5% significance before day 1. level, seizure reduction will have to be 0.78 times Monotherapy trials during presurgical assess- the standard deviation for a 10 patient sample, Monotherapy AED trials in patients undergoing presurgical assessment 301

whereas it would have to be 0.25 for 100 patients 5. Treiman, D.M. and Ben-Menachem, E. Inhibition of and 0.18 for 200 patients. carbamazepine and phenytion metabolism by nafimidone, a new antiepileptic drug. Epilepsia 1987: 28: 699-705. In summary, monotherapy trials during presur- 6. Hansen, J.M., Kristensen, M. and Skovsted, L. Sulthiame gical assessment offer a relatively new methodol- (Ospolot) as inhibitor of diphenylhydantoin metabolism. ogy to study efficacy and safety of antiepileptic Epilepsia 1968: 9: 17-22. drugs during Phase II with minimal interactions 7. Pledger, G.W. and Kramcr, L.D. Clinical trials of with standard anticonvulsants. Placebo controlled investigational antiepileptic drugs: monotherapy designs. Epilepsia 1991: 32: 716-721. parallel double blind designs are preferable 8. Bourgeois, B., Leppik, I., Sackellares, J.C. el al. because absolute seizure frequency can increase Felbamate: a double-blind controlled trial in patients during the trial due to withdrawal seizures, even if undergoing presurgical evaluation of partial seizures. the novel drug is effective. A number of Neurology 1993: 43: 693-696. difficulties outlined above could largely be over- 9. van Luijtelaar, E.L.J.M. and Coenen, A.M.L. Effects of remacemide and its metabolite FPL 12495 on spike-wave come by gradually introducing and possibly discharges, electroencephalogram and behaviour in rats titrating to the tolerated limits the experimental with absence epilepsy. Neuropharmacology 1995: (in treatments prior to withdrawal of comedication press). and tapering off standard antiepileptic medication 10. Palmer, G.C., Clark, B. and Hutchinson, J.B. Antiepilep- over several days before surgical assessment. tic and neuroprotective potential of remacemidc hydroch- Ioride. Drugs of the Future 1993: 18: 1021-1042. With an effective treatment, patients randomized I1. Palmer. G.C., Murray, R.J., Wilson, -T.C.M., et al. to the active compound at therapeutic doses Biological profile of the mctabolites and potential should show no, or less, withdrawal seizures. mctabolites of the anticonvulsant remacemide. Epilepsy Anticonvulsant pharmacokinetics should be care- Research 1992: 12: 9-21). fully considered and the timing of withdrawal of 12. Bannan, P.E., Graham, D.I., Lees, K.R. and McCulloch, J. Neuroprotectivc effect of rcmacemide hydrochloride in standard anticonvulsants planned so that sub- focal cerebral ischemia in the cat. Brain Research 1994: therapeutic levels are not reached before surgical 664: 271-275. assessment in patients on placebo. If the novel 13. Pierce, M.W., Suydak, P.D., Juvstavson, L., Mengel, drug is effective, admission time and duration of H.B., McKelvy, J.F. and Maut, T. Tiagabine. In: The New invasive recordings would be lengthened but this Antiepileptic Drugs. (Eds E. Perucca, F. Pisant. A. Richens.) Epilepsy Research 1991: (Suppl. 3): 157-160. would be amply justified if some patients were 14. Temkin, N.R. and Wilensky, A.J. New AEDs: are the controlled on the experimental agent and there- compounds or the studies incffective? Epilepsia 1986: 27: fore avoided surgery. 644-645. 15. Schmidt, D. Two antiepilcptic drugs lor intractable epilepsy with complex partial seizures. Journal of Neurology. Neurosurgery and Psychiatry 1982: 45:1119- REFERENCES 1124. 16. Marks, D.A., Katz, A., Scheyer, R. and Spencer, S.S. 1. Coatsworth, J.J. Studies on the Clinical E~cacy of Clinical and electroencephalographic effects of acute Marketed Antiepileptic Drugs. NINDS Monograph No 12. anticonvulsant withdrawal in epileptic patients. Neurology Washington, U.S. Government Printing Office, 1971. 1991: 41: 508-512. 2. Richens, A. Drug Treatntent of Epilepsy. London, 17. Brodie, M., Lassen, L. Ch. et al. Randomized, Double- Kimpson, 1976. blind, placebo-controlled, parallel-group study of safety 3. Gram, L., Bentsen, K.D., Parnas, J. and Flachs, H. and efficacy of tiagabine administered three times daily as Controlled trials in epilepsy: a review. Epilepsia 1982: 23: adjunctive treatment for partial seizures. Epilepsia 1994: 491-519. 35(Suppl. 7): 61. 4. Patsalos, P.N., Shorvon, S.D., Elyas, A.A. and Smith, G. 18. Duncan, J., Sander, J.W., Lyby, K. et al. Long-term safety The interaction of denzimol (a new anticonvulsant) with and efficacy evaluation of patients with partial seizures carbamazepine and phcnytoin. Journal of Neurology. treated with tiagabine in two open extension studies. Nearosurgerv and Psychiatry 1985: 48: 374-377. Epilepsia 1994: 35(Suppl. 7): 74.