Double-blind crossover trial of lamotrigine (Lamictal) as add-on therapy in intractable epilepsy

222

Epilepsy Res., 4 (1989) 222-229 Elsevier

EPIRES 00275

Double-blind crossover trial of lamotrigine (Lamictal) as add-on therapy in intractable epilepsy

C . D . B i n n i e *'1, R . M . C . D e b e t s * , M. E n g e l s m a n * , J . W . A . M e i j e r * , H . M e i n a r d i * , J. O v e r w e g * , A . W . P e c k * * , A . V a n W i e r i n g e n * a n d W . C . Y u e n * * *Instituut voor Epilepsiebestrijding, Heemstede (The Netherlands), and **Wellcome Research Laboratories, Beckenham ( U.K.) (Received 21 November 1988; revision received 3 January 1989; accepted 1 March 1989) Key words: Antiepileptic drug; Controlled trial; Lamotrigine; Epilepsy

A double-blind, placebo-controlled trial is reported of lamotrigine as add-on treatment in therapy-resistant epilepsy. A within-patients serial design was used, with two 3-month treatment periods and an intervening 6-week washout/crossover period. An unblinded investigator adjusted lamotrigine dosage to achieve a plasma concentration within a previously predicted therapeutic range. All patients had therapy-resistant partial seizures, some in combination with other seizure types and were without serious neurological or intellectual deficit. Of 34 patients recruited only one was withdrawn because of an adverse experience (maculo-papular rash) probably related to the experimental drug and 30 completed the trial. The other 3 withdrawals were due to default during baseline, dispensing error and cholecystectomy. There was a modest statistically significant reduction in total and partial seizures on lamotrigine compared to placebo treatment. There was no difference in adverse experiences or abnormal biochemical or haematological findings between the lamotrigine and placebo periods. The plasma concentrations of concomitantly administered antiepileptic drugs were not affected by lamotrigine treatment. It is concluded that lamotrigine shows promise as an antiepileptic drug with low toxicity.

INTRODUCTION Lamotrigine (3,5-diamino-6-(2,3-dichlorophenyl)-l,2,4-triazine; Lamictal2) is a novel antiepileptic drug (AED) chemically unrelated to any established agent and has a spectrum of action in animal models similar to that of phenytoin and carbamazepine. The oral EDs0 in mice and rats is in the range of 1.9-3.9 mg/kg, with corresponding plas1 Present address: Maudsley Hospital, Denmark Hill, London SE5, U.K. 2 Lamictal, Wellcome Foundation Ltd., Beckenham, Kent, U.K. Correspondence to: Dr. C.D. Binnie, Maudsley Hospital, Denmark Hill, London SE5 8AZ, U.K.

ma concentrations of the order of 1.5/ag/ml 8. In volunteers the drug was well tolerated in doses up to 240 mg daily producing blood levels in the predicted therapeutic range of 1-3 ~g/ml. The drug is completely bioavailable, with a long half-life (24 h) and linear kinetics3. Some preliminary evidence of efficacy in man was obtained by suppression of photosensitivity and interictal epileptiform EEG discharges 1,5 and by 2 studies of 7 days administration in patients with frequent partial or tonic-clonic seizures2'6. In normal volunteers the mean elimination half-life was approx. 24 h but single dose studies in patients already receiving AEDs showed the half-life to be reduced to a mean of 15 h in those taking enzyme-inducing drugs and in-

0920-1211/89/$03.50 (~) 1989 Elsevier Science Publishers B.V. (Biomedical Division)

223 creased to a mean of 59 h in those on sodium valproate monotherapy 1'6. These effects were confirmed by the results of a further study2. The linear kinetics of lamotrigine make it possible to estimate with reasonable accuracy by a computer stimulation the half-life of lamotrigine given the dose, weight of patient and steady state trough lamotrigine concentration; Binnie et al. 2 confirmed that after 7 days dosing the estimated half-life was 9-14 h in patients receiving only enzyme-inducing drugs and 12-27 h when these were combined with sodium valproate. Adverse experiences in the 7day studies chiefly reflected sedation and occurred only at estimated peak plasma concentrations above 3.5 btg/ml. In setting up double-blind, placebo-controlled add-on trials, adjustment of lamotrigine dosing on the basis of plasma concentrations and/or co-medication was therefore a major consideration. The objectives of the present study were to assess the efficacy of lamotrigine as add-on therapy in patients with poorly controlled partial seizures, to assess adverse reactions in these patients, to examine possible interactions between lamotrigine and other AEDs with reference to plasma A E D concentrations, and to examine the effects of tapered withdrawal of lamotrigine on seizure frequency (carry-over or rebound effects). MATERIALS AND METHODS

Design The study used a double-blind, placebo-controlled 2 periods within-patients crossover design, in which either lamotrigine or placebo was added to existing A E D medications. The trial comprised 5 phases: Baseline: a period of 8 weeks to familiarise the patient with trial procedures and to provide a reference point if the subsequent phases showed an order effect. Treatment period 1:12 weeks during which placebo or lamotrigine was administered. Washout 1:6 weeks. Treatment period 2:12 weeks. Washout 2:6 weeks. In the lamotrigine/placebo sequence group lamotrigine dosage was tapered over the first 2

weeks of washout 1; in the placebo/lamotrigine group tapering of the experimental drug occurred over the first 2 weeks of washout 2. In the remainder of the washout periods all patients received placebo. Throughout the treatment and washout periods the patients received experimental medication as 2 white capsules twice daily and changes in dosing were effected by altering the strength of the capsules supplied. Each clinic was supplied with lamotrigine capsules in 10 different strengths as blister packs of 28 capsules (1 week's supply), each pack marked with a unique random number. An unblinded investigator (C.D.B.) with knowledge of the patients' concomitant medication and plasma concentrations of lamotrigine and co-medication, instructed the blinded investigators to dispense trial medication from specified packs. Patients were allocated by block randomisation to the 2 treatment sequences by the unblinded investigator who kept the randomisation codes and the code numbers of the packs of capsules. The unblinded investigator could be asked to break the code individually if a patient developed a serious adverse experience the treatment of which required knowledge of medication taken. Dosage was targeted at achieving peak lamotrigine plasma concentration of 3/~g/ml. On the basis of previous studies the predicted dose required to achieve this was as follows: in patients on enzymeinducing AEDs, 200 mg daily; on inducing A E D + valproate, 100 mg daily; on valproate monotherapy, 75 mg daily. For the first week half these predicted doses were administered. At the end of week 1 blood was taken for trough plasma lamotrigine estimation, and the dose doubled. When the patient attended at the end of the second week the unblinded investigator was able on the basis of the plasma level at 7 days to prescribe a dose targeted on the intended blood level for week 3. Further plasma lamotrigine estimations were obtained at the ends of weeks 2, 3, 4 and 8 and further adjustments of dosage made as necessary. In the event of adverse experiences which could be dose-related side effects, the unblinded investigator was consulted and prescribed either a reduced dose of lamotrigine or (during the placebo period)

224 an alternative pack of placebo.

Patients Patients were recruited from 3 epilepsy out-patient clinics of the Instituut voor Epilepsiebestrijding. They were aged 16-65 years with a confident diagnosis of epilepsy uncomplicated by pseudoseizures. It was required that the seizures should be easily recognisable by patients or relatives, classifiable by the international classification of seizures4 and must include partial seizures, together possibly with other types. The patients were required to have a seizure frequency of at least 4 per month and all patients had kept a seizure diary for at least the previous 3 months. It was required that the seizures should be resistant to drugs of first choice, appropriate to the type of epilepsy, administered to the limits of tolerance, that the co-medication should have been unchanged for the previous 3 months and unlikely to be changed for the duration of the study. Exclusion criteria were: severe organic or psychiatric disease other than epilepsy, severe mental subnormality or progressive neurological disorder, haematological, biochemical or renal abnormalities on laboratory screening considered to be of clinical significance and not attributable to enzyme induction by AEDs; status epilepticus in the preceding 6 months or more than once in the past 2 years; use of any investigative (unmarketed) AED in the preceding 6 months, or use of more than 2 antiepileptic drugs (though this proved unduly restrictive and was relaxed) or any other chronic medication other than AEDs and oral contraceptives; non-compliance or failure to document seizures or adverse experiences; pregnancy, lactation or current exposed risk of pregnancy (females of reproductive age were excluded unless using effective contraception). Withdrawal criteria were: withdrawal of informed consent, severe or unacceptable adverse experiences, including haematological or biochemical abnormalities, or intercurrent illness. The protocol was approved by the Ethical Committee of the University Hospital of Leiden and all patients participated on the basis of informed consent obtained orally and in writing.

Procedures On admission to the trial all patients underwent a full general physical and neurological examination together with biochemical and haematological screens and urinalysis. A full history was taken with particular reference to seizures and adverse experiences. Patients were seen at weeks 1, 2, 4, 8 and 12 of each treatment period and at weeks 1, 2 and 6 of the washouts. At each of these visits vital signs, seizure counts and adverse experiences were checked together with haematological screens and estimations of plasma AED concentrations. Additionally, biochemistry, urinalysis and urinary enzymes were checked at weeks 4, 8 and 12 of each treatment period and at week 6 of both washouts.

Statistical analysis Total seizure frequency, defined as the total count of all seizure types, was the primary efficacy variable. As the seizure counts were unlikely to be normally distributed, data were analysed using non-parametric methods (Wilcoxon rank-sum test) for a 2 period model proposed by Koch7. The model assumes treatment effects to be additive, however, treatment effect on seizure frequency is likely to be multiplicative. Logarithmic transformation of the raw data renders the effects additive on the transformed scale to conform to the model. Hence, logarithmic transformation of the data was

TABLE I

Summary statisticsfor demographic data Age (yrs) Range Mean (+ S.D.) No. of Males Females

16-51 37.1 + 10.26

22 8

Height (cm) Range Mean (_+ S.D.)

164-194 176 + 8

Weight (kg) Range Mean (+ S.D.)

52.5-102.0 70.7 _+ 12.1

225 defaulted during the baseline period, another had to be excluded because of an error in dispensing, 1 was withdrawn because of a maculo-papular rash appearing whilst on lamotrigine, and another (randomised to the lamotrigine-placebo group) was admitted to hospital for a cholecystectomy during the placebo period. Thirty patients thus completed the study and could be included in the analysis of

TABLE II

Summary statistics of seizure history Etiology of seizures Idiopathic/unknown Symptomatic

22 8

Age of onset of epilepsy (yrs) Range Mean (_+ S.D.)

1-40 14.3 + 10.7

Duration of seizures at entry to trial (yrs) Range Mean (+ S.D.)

6-49.5 22.8 + 11.0

History of status epilepticus Yes No

7 23

TABLE IV

Individual total seizure countsfor each patient Patient no.

9 16 4 1

No. of AEDs being taken 1 2 3 4

2 13 13 2

not performed to normalise the data, but to satisfy the requirements of the statistical model. RESULTS Thirty-four patients were recruited, 1 of whom

TABLE III

Number of patients showing improvement Totalseizures Partialseizures Seizure days

(%)

P

L

P

L

P

L

1- 25 26- 50 51-100

5 4 0

7 10 2

4 4 0

8 10 2

7 1 0

10 6 2

Total No. showing no change

9

19

8

20

8

18

2

P = placebo; L = lamotrigine.

Change (%)*

(i)Patien~ withfewerseizuresonlamotrigine

No. of uncontrolled seizure types 1 2 3 4

Improvement

Treatment Totalno. of seizures on sequence Lamotrigine Placebo

2

4

1 3 6 7 9 11 13 14 16 17 18 19 20 21 22 27 28 32 34

L/P P/L P/L L/P L/P P/L P/L P/L L/P L/P P/L P/L L/P P/L P/L L/P P/L P/L L/P

29 31 10 101 44 26 14 4 14 17 63 20 42 6 10 14 37 444 4

40 50 17 145 61 31 34 5 18 18 95 22 44 11 17 17 65 697 12

28 38 41 30 28 16 59 20 22 6 34 9 5 45 41 18 43 36 67

~0 Patien~ withfewerseizures onplacebo 4 5 10 12 15 25 29 30 33

L/P P/L P/L L/P L/P L/P P/L L/P L/P

18 68 10 89 25 19 20 18 20

13 62 9 63 17 16 18 15 10

28 9 10 29 32 16 10 17 50

Oii)Patien~showingno changeintotalseizures 23 31

L/P P/L

17 8

17 8

0 0

P = placebo; L = lamotrigine. * % change is expressed relative to the larger of the total seizure counts in the two treatment periods.

226 efficacy. A preliminary analysis of adverse experiences after the first 4 weeks of treatment period 1 was performed in the first 16 patients admitted (8 on placebo, 8 on lamotrigine) in order to check safety. No changes in the conduct of the study resuited from this preliminary survey, the results of which were not communicated to the investigators. Demographic and epileptological features of the participating patients are summarised in Tables I and II.

by Koch7 indicates that the treatment x period interaction was not statistically significant (P > 0.20). A comparison of total seizure frequency between the first 4 weeks of each washout period shows no effect of the preceding treatment (P > 0.20). Separate analyses of partial seizures and of total seizure days on which seizures occurred gave results similar to the above (P < 0.01 and P = 0.06, respectively).

Safety Efficacy

A total of 39 adverse experiences were reported in 29 patients. Nine of these were classified as serious (Table V). Only 2 of these occurred on lamotrigine, and in 1 instance had been previously reported by the same patient on placebo. A maculopapular rash appearing after 20 days lamotrigine administration (final dose 100 mg daily) appeared to be drug related and resolved after withdrawal. The adverse experiences leading to dose reduction of trial medication are summarised in Table VI. The differences between treatments were minimal and not significant. Clinical laboratory evaluation, vital signs and general physical and neurological examination did not show any significant differences between lamotrigine and placebo, nor indeed any trends suggestive of an adverse effect of the drug.

There was a significant period effect with total seizure counts tending to be higher in treatment period 1 (P < 0.05) and there was a significant treatment effect in favour of lamotrigine (P > 0.02). The median percentage reduction in seizure count attributable to lamotrigine was 17% with a confidence interval of 0-30%. Table III shows the numbers of patients experiencing the indicated percentage change in total seizure frequency between treatment periods, and Table IV shows individual seizure counts and treatment sequences. Nineteen out of 30 patients showed an improvement on lamotrigine with respect to placebo, and it will be seen that 12 patients had a least 25% fewer seizures on lamotrigine, whilst 4 showed a similar improvement on placebo. Analysis of log transformed seizure totals using the method proposed

TABLE V

Serious adverse reactions reported during the trial Patient no.

Adverse experience (preferred terms)

PLO/L TG sequence group 2 Rash 29 Headache Nausea 32 Headache Asthenia L TG/PLO sequence group 9 Vision abnormal 20 Depression Depression Depression

When reported (phase~week)

L TG dose (mg)

Duration on L TG (days)

Intensity

Action taken

T2/28 T1/16 T1/16 T1/16 T1/16

100 PLO PLO PLO PLO

20 0 0 0 0

Moderate Moderate Severe Moderate Severe

Withdrawn Dose reduction Dose reduction

T2/30

PLO

0

Moderate

Dose reduction

B/0 B/8 T1/9

PLO PLO 100

0 0 7

Moderate] Moderate~ Moderate)

Psychological counselling

227 TABLE VI

Adverse experiences requiring dose reduction of trial medication Patient no.

Adverse experiences (preferred term)

Intensity

Seriousness

Phase~week

Dose reductionfrom to (rag)

Mod Mod Sev Mod Sev

Non Ser Ser Ser Ser

T1/30 T1/16 T1/16 T1/16 T1/16

Placebo Placebo Placebo Placebo Placebo

Mod Mod Mod Mod Mod Mod Mod Sev Sev Mild Mod Mild Sev Mild

Non Non Non Non Non Non Non Non Non Non Non Non Non Non

T2/28 T2/28 T2/28 T2/34 T2/34 T2/34 T2/34 T1/12 T1/12 T1/ll T2/28 T2/38 T2/38 T2/34 T1/12

100- 75 100- 75 100- 75 100- 50 100- 50 100- 50 100- 50 200-150 200-150 100- 75 200-100 300-100 300-100 200-150 200-150

Dummy dose reductions on PLO 9 29 32

Diplopia Headache Nausea Headache Asthenia

Dose reductions on L TG 6

6

8 9 11 14 22 23

Ataxia Dysarthria Chills Dysarthria Chills Pain breast Thinking abnorm Headache Diplopia Patient demand Diplopia Dizziness Headache Somnolence Dizziness

Intensity: Mod = moderate; Sev = severe. Seriousness: Non = non-serious; Ser = serious.

Therapeuticdrugmonitoring Mean plasma concentrations of the most common co-medication are summarised in Table VII. There is no evidence of any effect attributable to lamotrigine. Carbamazepine concentrations were significantly higher overall in treatment period 2 (S statistic = 4.5, df 1.25, P < 0.05). The mean plasma levels of lamotrigine in patients with a lower seizure count on that drug were 1.74 _+ 0.57

against 1.76 + 0.49 pg/ml in non-responders. Estimated trough plasma levels were 1.47 pg/ml in both instances. In patients on enzyme-inducing drugs the modal LTG dose over the 12-week treatment period was in the range of 50-400 mg/day (mean 242 + 81) and the mean plasma concentrations (at random intervals after dosing) at weeks 4, 8 and 12 were in the range of 0.8-2.0 pg/ml (mean 1.5 + 0.4)

TABLE VII

Mean plasma concentrations of concomitant antiepileptic drugs

Carbamazepine Phenobarbitone Phenytoin Sodium valproate Clobazam

Mean during placebo (PLO)

Mean during lamotrigine (L TG)

S.E. of difference (PLO-L TG)

95% interval estimate of difference (PLO-L TG)

6.63 16.86 12.99 68.9 56.4

7.03 14.52 11.46 68.6 56.4

0.407 1.349 0.860 4.6 8.37

-1.24; -1.40; -0.27; -10.6; -19.8;

0.44 6.08 3.33 11.2 19.8

228 TABLE VIII Comedication, lamotrigine dose and plasma concentrations Comedication regimen

Modal L TG dose ( r a g )

Mean L TG concentration (l~g/ml)

Inducing AEDs (n = 22) Range 50-400 Mean + S.D. 242 + 81

0.8-2.0 1.5 + 0.4

lnducing A E D + VPA (n = 9) Range 50-150 Mean + S.D. 114 + 38

1.5-3.0 2.3 + 0.5

(Table VIII). Where enzyme inducers were combined with valproate, the dosage range was 50-150 mg/day (mean 114 + 38) and the plasma concentrations were 1.5-3.0/~g/ml (mean 2.3 + 0.5). Blood was drawn for lamotrigine estimations typically some 8 h after dosing (on a 12-h schedule), and the values obtained were thus near trough levels. Peak to trough differences are necessarily less in patients with a longer half-life. The method of targeting on calculated peak levels (in order to avoid side effects) required, therefore, higher trough levels in the patients with longer half-lives due to valproate co-medication. DISCUSSION AND CONCLUSIONS Lamotrigine produced a clinically modest but statistically significant reduction in seizure counts in patients with poorly controlled partial seizures.

REFERENCES 1 Binnie, C.D., Van Emde Boas, W., Kasteleijn-Nolst Trenit& D.G.A., De Korte, R.A., Meijer, J.W.A., Meinardi, H., Miller, A.A., Overweg, J., Peck, A.W. and Van Wieringen, A., Acute effects of lamotrigine (BW430C) in persons with epilepsy, Epilepsia, 27 (1986) 248-254. 2 Binnie, C.D., Beintema, D.J., Debets, R.M.C., Van Erode Boas, W., Meijer, J.W.A., Meinardi, H., Peck, A.W., Westendorp, A.-M. and Yuen, W.C., Seven day administration of lamotrigine in epilepsy: add-on trial, Epilepsy Res., 1 (1987) 202-208.

It is not possible from the response in this therapy resistant population to predict how effective the drug might prove in a less refractory group. The efficacy demonstrated in the present study cannot be attributed to a kinetic interaction with other AEDs as the blood concentrations of co-medication were unchanged. Lamotrigine was well tolerated over the 3month treatment period. The only important adverse experience attributable to the drug was a skin rash leading to withdrawal. Otherwise adverse experiences or changes in laboratory safety parameters on lamotrigine did not differ significantly from those on placebo. The method of targeting plasma lamotrigine levels by means of an unblinded observer proved very successful. Although a complex procedure, requiring good communications and rapid on-site estimation of the experimental drug, its use avoided the problems of underdosing or intoxication (with consequent breaking of the blinding) which would otherwise probably have resulted from the effects of co-medication on lamotrigine metabolism. This trial, having provided the opportunity for further study of these interactions in 30 patients under steady-state conditions, has enabled subsequent studies to employ a simpler design with lamotrigine dosing determined by co-medication. The resuits in the present study highlighted the importance of investigation of possible pharmacokinetic interactions of any new A E D before it is used in add-on clinical trials, and where these are found the use of an unblinded investigator to monitor blood levels and adjust dosage of the experimental drug or, where appropriate, of the co-medication.

3 Cohen, A.F., Land, G.S., Breimer, D.D., Yuen, W.C., Winton, C. and Peck, A.W., Lamotrigine, a new anticonvulsant: pharmacokinetics in normal humans, Clin. Pharmacol. Ther., 42 (1987) 535-541. 4 Dreifuss, F.E., Proposal for revised clinical and electroencephalographic classification of epileptic seizures, Epilepsia, 22 (1981) 489-501. 5 Jawad, S., Oxley, J., Yuen, W.C. and Richens, A., The effect of lamotrigine, a novel anticonvulsant, on interictal spikes in patients with epilepsy, Br. J. Clin. Pharmacol., 2 (1986) 191-193. 6 Jawad, S., Yuen, W.C., Peck, A.W., Hamilton, M.J., Ox-

229 ley, J.R. and Richens, A., Lamotrigine; single-dose pharmacokinetics and initial 1 week experience in refractory epilepsy, Epilepsy Res., 1 (1987) 194-201. 7 Koch, G.G., The use of non-parametric methods in statistical analysis of a two-period change-over design, Biomet-

rics, 28 (1972) 577-584. 8 Miller, A.A., Sawyer, D.A., Roth, B., Wheatley, P.L., Leach, M.J. and Lamb, R.J., Anticonvulsant studies on BW430C, a novel potential antiepileptic drug, Epilepsia, 25 (1984) 655-656.