A survey comparing lamotrigine and vigabatrin in everyday clinical practice

Seizure

1996;

5: 267-270

A survey everyday GRAHAM

comparing lamotrigine clinical practice

SCHAPEL

Department Correspondence & Neurosurgery,

& DAVID

of Neurological

in

CHADWICK

Science,

to: Professor David Rice Lane, Liverpool

and vigabatrin

The Walton

Chadwick, L9 1AE,

Centre

Department UK

for Neurology

of Neurological

& Neurosurgery,

Science,

The

Walton

Liverpool, Centre

U.K.

for Neurology

The objective of this study was to compare the efficacy and tolerability of two new antiepileptic drugs, lamotrigine (LTG) and vigabatrin (GVG) in everyday clinical practice. A comprehensive retrospective survey of a computerized data baseand hospital casenotes wascarried out at the Mersey Regional Epilepsy Clinic (MREC), Liverpool, which services a population of 3 million in the North West of England. The study comprised 333 out-patients with refractory epilepsy exposed to LTG and GVG forming a subsetin a total population of 2250 patients with epilepsy held on a comprehensivedatabase.The main outcome measureswere duration of treatment with each drug described by a Kaplan-Meier survival curve, seizure control determined by a 50% decreasein seizure frequency and freedom from seizures,and incidenceof adversedrug effects leading to discontinuation. The Kaplan-,Meier curve indicated a 57% probability of patients continuing to take LTG and 43% GVG after 40 months.A 50% improvement in seizure control followed the addition of LTG in 45% of patients, with 10% seizure free, compared with 32% and 6%, respectively after the addition of GVG. LTG was discontinued becauseof adverse events (most frequently skin rash) in 15% of patients compared to GVG in 25% (particularly becauseof personality disturbanceand psychiatric disorder). Both LTG and GVG are effective new AEDs in patients with refractory epilepsy, treated in a tertiary referral out-patient setting. LTG has a broader spectrum of antiepileptic efficacy for patients with both partial and idiopathic generalized seizures,whereasGVG should be reserved for patients with partial seizuresat low risk of psychiatric disorder.

INTRODUCTION New antiepileptic drugs (AEDs)

are licensed as a result of pre-registration studies which compare a potential AED against placebo in add-on clinical trialsi4. However, such trials are of limited value in informing clinicians about the use of a new drug in everyday clinical practice5. Preregistration studies inevitably take place in a highly selected population of people with a long history of refractory epilepsy, and there must be questions about extrapolating results from this population to a larger group of less refractory patients with less severe epilepsy. Furthermore, the studies which will ultimately define the place of new AEDs in clinical practice will be comparative studies against existing first choice drugs. A number of comparative monotherapy studies of older drugs have been reported”12. These studies are, however, difficult to perform and take prolonged periods of time to recruit and to analyse. We have adopted a rather different 1059-1311/96/040267

+ 04

$12.00/O

approach, namely comprehensive survey of two newly licensed AEDs, lamotrigine (LTG) and vigabatrin (GVG), within a tertiary referral epilepsy centre.

PATIENTS AND METHODS

A survey of LTG and GVG treatment was performed between September 1993 and February 1994 at the Walton Centre for Neurology and Neurosurgery. A cohort of 333 patients prescribed LTG and GVG was identified from the total number of 2250 patients, data of whom was held on a computerized data base of the Mersey Regional Epilepsy Clinic (MREC) in Liverpool, servicing a population of three million in the North West of England. The identification of patients prescribed LTG and GVG in this population was thus comprehensive, all patients having their intended AED therapy documented on the data base at each visit. We can be 0

1996

British

Epilepsy

Association

268

confident that all patients in whom there was intent to treat with either drug have been identified. A proforma was designed to define demography, epilepsy and seizure type and aetiology, pre-existing or associated neurological, intellectual and psychiatric dysfunction, duration of therapy, dosage and concomitant AED therapy, adverse effects leading to discontinuation, and response of seizure control following treatment with LTG or GVG. Seizure control was evaluated by estimating the average monthly seizure frequency over three months before the introduction ot LTG or GVG and for the last three months of treatment with the maximal dose of each new drug. Data were initially obtained by viewing the computer data base and by perusal of the hospital case notes. Subsequently, where data were missing, they were retrieved by questioning patients, family and general practitioners over the telephone. The completed data record forms were checked, collated and then entered on a computerized data base for further analysis, using summary statistics and the construction of a Kaplan-Meier survival curve.

RESULTS

The population studied was one with severe refractory epilepsy typical of that seen in a tertiary referral centre. Severe epilepsy, defined as more than one minor seizure (not tonic-clonic) per week or more than one generalized or secondary generalized tonic-clonic seizure per month, was present in over 90% of patients (Table 1). About two thirds of patients were already taking one, a third were taking two, and a few three conventional AEDs, while over a third had neurological, learning or psychiatric dysfunction before LTG or GVG was added to their treatment (Table 1). Of the lamotrigine exposures, 154 (69%) were in patients with localization related epilepsy, 65 (29%) in patients with generalized epilepsy, and four in patients whose epilepsy was not classifiable. For vigabatrin the figures were 182 (83%), 32 (15%) and 3, respectively. There were 223 patient-treatments with LTG for up to 56 months and 217 patient-treatments with GVG for up to 72 months in 333 patients: 116 patients (35%) were exposed to LTG only, 108 patients (32%) to GVG only and 109 patients (33%) to both drugs, either serially or in combination. LTG was continued in preference to GVG by 36 patients and GVG was

G. Schapel Table 1: Comparison of response-related oatients treated with LTG and GVG

Characteristic

Epilepsy Severe*

severiry

Number AEDs

of baseline

1 2 3 Associated

& D. Chadwick characteristics

of

LTG % of patients n=223

GVG % of patients n = 217

92

91

63 36 1

62 34 4

34

37

disability

Neurological dysfunction, learning disability and/or nsvchiatric disorder

Expressed according to seizure frequency during the three months before addition of LTG and GVG, where ‘severe’ indicates more than one minor seizure per week or more than one major seizure per month.

continued in preference to LTG by 17 patients, both drugs were continued in combination by 31 patients and both were discontinued by the same patient on 25 occasions. From the Kaplan-Meier curve the probability of patients continuing to take LTG at 40 months was 57% and that for GVG was 43% (Fig. 1). In fact, LTG was still being taken by 67% of patients and GVG by 51%

0.4 1 0

1 10

Fig. 1: Kaplan-Meier GVG in 333 patients

1 I 1 t 1 1 20 30 40 50 60 70 Duration of treatment (months) survival curve for duration -, LTG; ---, GVG.

1

8a

on LTG and

Number at risk LTG

CI-’

Number CI’ at risk GVG

Immediately before 20 months

69

55-70%

93

52-66%

Immediately before 40 months

18

44-70%

37

36-52%

CI’, Confidence interval for Kaplan-Meier at each time point

survival function

Lamotrigine

vs. vigabatrin

in everyday clinical practice

269

at the time of audit. Both LTG and GVG were taken as monotherapy by 12% of patients. The addition of LTG was followed by a reduction in the number of other AEDs in 37 patients (17%) and by GVG in 38 patients (18%). The mean number of AEDs before the addition of LTG or GVG was 1.5, compared with 1.3 at the census point demonstrating that the policy was largely one of substituion. A 50% improvement in seizure control followed the addition of LTG in 45% of patients (Fig. 2), with 10% seizure free (Fig. 3), compared to 32% (Fig. 2) and 6% (Fig. 3), respectively for GVG. A 50% seizure reduction was recorded in 14 patients following substitution of LTG for GVG, in seven patients following substitution of GVG for LTG and in 13 patients who were prescribed LTG and GVG in combination. LTG was discontinued because of an adverse event in 15% of patients, skin rash being the most frequent adverse event, accounting for 5% but also including nausea, ataxia and visual disturbance (Table 2). GVG was discontinued in 25% of patients because of adverse events, including personality disturbance, drowsiness and weight gain, psychiatric adverse events occurring in 8% of patients. The drop out rates due to lack of efficacy (as opposed to apparent exacerbation of seizures) were 5.5% for LTG and 11% for GVG. The median daily dose of LTG was 400 mg. without concomitant sodium valproate (VPA) and 2000 mg. in combination with VPA while that of GVG was 200mg. There was evidence of preference for combining LTG with VPA and GVG with carbamazepine (CBZ), possibly related to the broader spectrum of efficacy observed for LTG in both localization-related and generalized epilepsy syndromes.

20

t8 I5 $ ‘9

I2165

22i223

IO

8 ‘Z 5

‘j

0 All

epilepsy

Localization epilepsy

This retrospective survey of the use of two new AEDs in a large tertiary referral out-patient population with severe epilepsy was nonrandomized and is therefore subject to potential biases which may influence the results’3. A selection bias could have occurred since prescribers clearly stopped exposing patients with generalized epilepsies to GVG relatively early in the period during which these two drugs became available. A further bias was that imposed by the earlier availability (by 1 year) of GVG which may have resulted in the administration of LTG to a more therapeutically refractory patient population. The recording of adverse events in the present study was unblinded so that there may have been bias in attributing unwanted symptoms of specific types to LTG and GVG. However, the strength of the study is that it was comprehensive and included all patients exposed to LTG and GVG within an individual tertiary referral centre without any exclusion, the results being presented on an intent-to-treat basis. This

event

(%) of patients % of events GVG n =217

epilepsy

Localization epilepsy

related

CNS Psychiatric Increase

Fig. 2: Numbers above bars represent those from which responders are calculated. 0, Lamotrigine; n , vigabatrin.

in seizures

Death Gastro-intestinal Rash

Genralized epilepsy

%

%

DISCUSSION

Adverse

All

Genralized epilepsy

Fig. 3: Numbers above bars represent those from which responders are calculated. Cl, Lamotrigine; H, vigabatrin.

Table 2: Percentage adverse events

3132

related

Total

withdrawn leading

due to to withdrawal

LTG

n = 223

9

5

8

0.5

4 3 1

1 1 2

0.5

5

25

15

270

study is not a substitute for randomized controlled trials of LTG vs. GVG, but it does reflect the clinical effectiveness of these two AEDs in the real world of treatment of patients with severe epilepsy refractory to previous treatment with conventional AEDs. It also serves to generate hypotheses concerning the use of LTG and GVG which can subsequently be tested by prospective controlled studies. The study was undertaken in a refractory population with severe epilepsy as shown by the pre-treatment seizure frequency, the high incidence of associated neurological and intellectual impairment and by a relatively high mortality during follow-up (10 patients died). The retention times on both drugs were significant, and the mean number of drugs being taken by patients fell somewhat, indicating that the drugs appear to be clinically useful. Patterns of responsiveness and adverse events are informative. The results of the present study suggest that LTG is as effective as GVG for the control of partial seizures, despite the information about responder rates provided by placebocontrolled add-on trials’. This result further emphasizes that caution needs to be exercised in interpreting the results of phase II and III add-on trials when comparing new AEDs administered to different populations of refractory patients’. The results of the present study support a different spectrum of efficacy for LTG and GVG. They suggest that LTG is a broad-spectrum AED, analogous to valproate, effective in patients with both localization-related and generalized epilepsy syndromes, whereas GVG should be prescribed selectively for patients with localization-related epilepsy3. There also appears to be a higher incidence of psychiatric side-effects leading to discontinuation of GVG than LTG and a clearly drug-related excessive incidence of skin rash resulting in the withdrawal of LTG3.

G. Schapel 81 D. Chadwick

REFERENCES 1. Morrow,

Drug

J.I. and

Richens.

Euuhcution

A. Drugs

O.J.). MacMillan Gram, L. and controlled clinical

3.

(Suppl. 7): Sl-S6. Patsalos. P.N. and A review of their

4.

Drug Therapy. CNS Dncgs 1994: Schmidt, D. Evaluation of clinical drug trials. In: New Anriepileplic

Duncan. current

to

study

Phurtnaco-epidemiology Livingstone.

2: 40-77. efficacy Drugs

New

York,

1985; 313: Mattson.

in antiepileptic (Epilepsy

Res

Richens,

A.)

E. and The

drugs. potential.

use

of

phramaco-

beneficial drug effects. (Ed. Strom, B.L.) Churchill 1989: pp. 307-324.

Mattson. R.H.. Cramer. Comparison of carbamazepine. and primidione in partial tonic-clonic

7.

J.S. New antiepileptic status and clinical

wppl 3) (Eds Pisani. F.. Perucca. Elsevier Science 1991: pp. 69-77. Strom, B.L. and Malmon. K.L. epidemiology

6.

In: Eur/.v

(Eds O’Grady, J. and Linet. Press. London. 1990: pp. 395-406. Schmidt, D. Innovative designs of trials in epilepsy. Epikpsiu 1993: 34

2.

5.

in epilepsy.

in Man.

J.A..

Collins. phenobarbital, and secondarily

seizures. New 145-151. R.H.. Cramer.

England

J.E.

er ul.. phenytoin generalised

Journal

J.A.

and

In:

of Medicine

Collins.

J.F.

(The

Department of Veterans Affairsd Epilepsy Cooperative Study No. 264 Group). A comparison of valproate with carbamazepine for the treatment of complex partial seizures and secondarily generalised tonic-clonic seizures in

New

adults.

England

Journal

of Medicine

1992:

327:

767-771. 8. Richens. A., Davidson, Easter. D.J. A multicentre carbamazepine in adult

D.L.W., trial onset

ology. 9. Steiner,

Psychiarry

Neurosrrrgery and T.J.. Silveira,

Thames Lamictal rigine (Lamictal) diagnosed epilepsy. IO. Tallis. R. and valproate and Il. Heller, A.J..

S.

and

Study Group). and phenytoin

Epilepsiu

Cartlidge. of sodium epilepsy.

N.E.F. valproate

Journal Neur1994: 57: 682-687.

Yuen.

A.W.C.

Comparison monotherapy 1994: 35 (Suppl.

Easter. D. Multicentre phenytoin. Epilepsia Chesterman. P..

and and

(North of lamotin newly 7): 61.

comparative 1994: 35 (Suppl. Elwes, R.D.C.

study of 7): 62. et cl.,

Phenobarbitone. phenytoin. carbamazepine or sodium valproate for newly diagnosed adult epilepsy: a randomized comparative monotherapy trial. Jotrrtml oj

Neurology 12.

Brodie. blind

newly 13. Altman. Chapman

Neurosurgery

M.J.. comparison diagnosed D.G. and

Psychiarry

Richens. of

A. and lamotrigine

epilepsy.

Prucrical Hall.

London,

Lancer Statistics 1991:

1995: 58: 44-50. Yuen. A.W.C. Double and carbamazepine 1995:

345:

for

Medical

pp.

92-101.

476-479.

Research.

in