Carbamazepine-10,11-Epoxide in Epilepsy A Pilot
Study
Torbj\l=o"\rnTomson, MD; Ove Almkvist, BSc; Bengt Y. Nilsson, MD; Jan-Olof Svensson, Engr; Leif Bertilsson, PhD \s=b\ The effects of carbamazepine-10,11\x=req-\ epoxide, an active metabolite of carbamazepine, were evaluated in seven outpatients with frequent epileptic seizures. The study included an initial 4-week period with the carbamazepine dose optimized for each patient. Patients were then crossed over, dose by dose, to carbamazepine-10,11-epoxide and followed up for another 4 weeks. Dosing was single
blind. The evaluation of the anticonvulsant effect was hampered by marked fluctuations in plasma levels during treatment with carbamazepine-10,11-epoxide. There was, however, no significant change in seizure control. During epoxide treatment, no subjective side effects were
reported despite epoxide plasma
concen-
trations up to 57 \g=m\mol/L. Neuropsychological assessment revealed a significant improvement in finger motor speed and logical reasoning during the carbamazepine-10,11-epoxide period. Subnormal serum sodium levels in two patients were normalized after switching from carbamazepine to the epoxide. Continued investigations with this active metabolite of carbamazepine in epilepsy are therefore
justified. (Arch Neurol. 1990;47:888-892)
Accepted for publication November 3, 1989. From the Departments of Neurology (Dr Tomson) and Clinical Neurophysiology (Dr Nilsson) S\l=o"\dersjukhuset;Department of Psychiatry, St G\l=o"\ran'sHospital (Dr Almkvist); and Department of Clinical Pharmacology, Huddinge University Hospital, Karolinska Institute (Dr Bertilsson and Mr Svensson), Stockholm, Sweden. Reprint requests to Department of Neurology, S\l=o"\dersjukhuset,S-100 64 Stockholm, Sweden (Dr
Tomson).
is Carbamazepine the treatment
first-line drug in of epilepsy1 and trigeminal neuralgia2 and is now also frequently used in bipolar depression.3 Carbamazepine is metabolized to carbamazepine-10,ll-epoxide, which is an active metabolite that appears in plasma at concentrations of 10% to 40% of the parent compound." The possible contribution of this metabo¬ lite to the effects of carbamazepine therapy has recently been debated intensely.58 It is a common opinion that carbamazepine-10,ll-epoxide to a large extent contributes to the side ef¬ fects of carbamazepine therapy.67·915 Schoeman et al16 have suggested that neurotoxic side effects are a rule in pa¬ tients treated with carbamaze¬ pine when carbamazepine-10,ll-epoxide levels exceed 9 µ /L. Rosenfeld et al12 claimed that all patients with epoxide concentrations of 15 /¿mol/L or a
during carbamazepine treat¬ ment were either clinically intoxicated or unable to complain owing to mental more
retardation. These and other authors recommend routine monitoring of carbamazepine-10,ll-epoxide plasma lev¬ els to control for side effects of car¬ bamazepine therapy. The high epoxide concentrations in these studies oc¬ curred almost exclusively in patients treated with carbamazepine in combi¬ nation with other drugs, eg, phenytoin, which induce the epoxidation of car¬ bamazepine. Hence, it was not possible to determine if the observed side ef¬ fects were caused directly by the drug given in addition to carbamazepine or by the high epoxide levels. A reliable assessment of the effects
of carbamazepine-10,ll-epoxide re¬ quires direct administration of the ep¬ oxide to patients. In the only previous study of carbamazepine-10,ll-epoxide given as such to patients, we found a potent pain-relieving effect in trigeminal neuralgia, and no side effects were observed.17 We now report the results of a pilot study of carbamazepine10,11-epoxide in the treatment of epi¬
lepsy.
PATIENTS AND METHODS Patients
The study included seven adult outpa¬ tients (aged 21 to 56 years) with complex partial seizures with or without secondary generalization. All patients had at least five seizures per month before the study despite optimal treatment with carbamazepine given in combination with one or two other antiepileptic drugs (Table 1). The carba¬ mazepine dosage was unchanged for at least 6 weeks before the study. The duration of epilepsy ranged from 6 to 41 years. Patients with brain tumors and progressive brain disorders were excluded.
Study Design The study had a single-blind crossover design. Patients were followed up during a first 4-week period with their original opti¬ mal carbamazepine treatment dosed four times a day (7 am, noon, 5 pm, and 10 pm; pe¬ riod A). During the subsequent week they were switched over, dose by dose, from carbamazepine to carbamazepine-10,11epoxide therapy four times a day. This was followed by a 1-week washout period of possible carbamazepine effects and there¬ after by a 4-week observation period with carbamazepine-10,ll-epoxide (period B). The study design is illustrated in Pig 1. Pa¬ tients were kept unaware of changes in the
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
Table
1.—Dosage and Plasma Concentrations of Carbamazepine and Carbamazepine-10,11-Epoxide During Two Treatment Periods in Seven Epileptic Patients* Plasma Concentration, umol L
Patient No.
Treatment Period
Active
Drug Therapy, mg/d Carbamazepine, 1200 Carbamazepine-10,11-epoxide, Carbamazepine, 1200 Carbamazepine-10,11-epoxide, Carbamazepine, 800 Carbamazepine-10,11-epoxide, Carbamazepine, 1400 Carbamazepine-10,11 -epoxide, Carbamazepine, 1200 Carbamazepine-10,11 -epoxide, Carbamazepine, 1200 Carbamazepine-10,11 -epoxide, Carbamazepine, 600 Carbamazepine-10,11-epoxide, observations before morning dose at 7 am.
Concomitant Therapy
Carbamazepine-
Carbamazepine
10,11-epoxide
35
6.2
1200 1200 800
32 1400
24 1200
31
6.9
1200
29 600
8.5
22t
Plasma
Drug, mg/d Phenytoin, 200 Phenytoin, 200 Phenytoin, 50 Phenytoin, 50 Phénobarbital, 50 Phénobarbital, 50 Phénobarbital, 25 Phénobarbital, 25 Phenytoin, 200 Phenytoin, 200
Clonazepam, 1 Clonazepam, 1 Valproate sodium, Valproate, 3000
Concentration,
µ
/L
10
45
0.024 3000
*Mean of all tSingle observation.
therapy by use of carbamazepine-10,ll-epoxide-matched placebo tablets during pe¬ riod A and carbamazepine-matched pla¬ cebo tablets during period B. Concomitant
therapy (Table 1) was unchanged through¬ study, which altogether lasted 10
out the
weeks. Patients were seen at weekly intervals in the outpatient clinic and in addition hospi¬ talized during 2 days at the end of each 4week period for registration of clinical effects. Drugs were given to the patients in
specially designed dispensers containing
the separate dosages for 1 week. Compli¬ ance was tested by checking the dispensers at each visit. Blood samples for drug analysis were ob¬ tained immediately before the morning dose at each weekly visit, and in addition every second hour (7am to 7 pm) during the last day of the two 4-week periods. The plasma concentrations of carbamazepine and carbamazepine-10,ll-epoxide were de¬ termined with high-performance liquid
chromatography.18
Drugs
Carbamazepine was given as plain Tegretol tablets (Ciba-Geigy Corp). Carbamazepine-10,ll-epoxide in unstable is gas¬ tric juice. Enteric-coated tablets provided by Ciba-Geigy Corp" were therefore used. Evaluation of Effects
Seizures were recorded by the patient or, when appropriate, relatives on a calendar that was checked at every visit. All sponta¬ neous reports of side effects were recorded, and in addition patients were actively ques¬ tioned at each visit about neurotoxic symp¬
following a special questionnaire (available on request). The following tests were performed during the last 2 days of the two treatment periods with the patient in hospital. The different tests were made at the same time of day on the two occasions. Biochemical Tests.—Hemoglobin, hematocrit, red and white blood cell counts, diftoms
as illustrated by patient 1. Carbamazepine and carbamazepine-10,11-epoxide were administered four times daily. Trough morning plasma levels of carbamazepine (open squares) and carbamazepine-10,11-epoxide (solid squares) are indicated in top panel.
Fig 1.—Design and results of study
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
ferential and platelet counts formed. Electrolytes, albumin,
were
per¬
creatinine, aminotransferases, -glutamyltransf erase, alkaline phosphatase, folate, thyrotropin, and immunoglobulins were all measured in serum.
Psychological Assessment.—Based on re¬ concerning behavioral effects associated with antiepileptic drugs20 and especially carbamazepine,2122 six computercent studies
administered tests of load23-24 were selected to
varying assess
status.
mental
cognitive
Finger Motor Speed.—The subject was instructed to tap a button as fast as possi¬ ble in five trials of three conditions: the in¬ dex finger of the dominant hand, the index finger of the nondominant hand, and alter¬ nating the right and left index fingers. The mean number of taps for five 10-second pe¬ riods was used to measure performance. Simple Reaction Time.—The subject had to hit a button when a time-randomized signal appeared on the screen. The mean latency during a 5-minute testing time was used to measure performance. Complex Reaction Time.—The subject was instructed to respond as fast as possi¬ ble by pushing a button when the word meaning (one of four color words) and the print color (one of four colors) corresponded to each other, and otherwise not to respond. The mean reaction time during a 10-minute testing period was used to measure perfor¬ mance.
Short-term Memory Scanning.—Fol¬ lowing a paradigm developed by Sternberg,25 the subject had to push as fast as possible one of two buttons as to whether a random probe digit matched (positive re¬ sponses) one of the digits of a preestablished set of digits or not (negative respons¬ es). The mean latency for positive responses was, timed for three different set sizes (1,3, and 5 digits in the set). Symbol Digit Substitution.—In this test, a relative of the Digit Symbol Test of the Wechsler Adult Intelligence Scale-Re¬ vised,26 the subject had to match digits to symbols according to an existing code scheme as fast as possible. The mean la¬
tency of five trials
was
measured after
a
training period. Logical Reasoning.—The subject had to decide whether a verbal premise, eg, A pre¬ cedes B, was false or true compared with a picture of the letters, eg, A. The number of correct judgments of 64 items and the mean latency were registered. This test is a computerized version of the Baddeley ver¬ bal reasoning task.27 Patients were also subjected to a neurophysiological examination including elec¬ troencephalogram, visual evoked poten¬ tials, and visual contrast sensitivity. De¬ tails of this evaluation will be reported elsewhere. Informed consent from all patients.
was
obtained
RESULTS
The steady-state plasma concentra¬ tions of carbamazepine and carbamazepine-10,ll-epoxide during the two treatment periods are given in Table 1.
Fig 2.—Total number of seizures of seven patients treated with carbamazepine (left) and carba¬ mazepine- 10,11 -epoxide (right). Note that seizures during epoxide treatment of patient 7 refer to the crossover and washout period, after which the patient dropped out of the study. No significant difference in seizure frequency was found (Wilcoxon Signed-Rank Test).
For the whole patient group, mean steady-state concentrations were 32 µ /L of carbamazepine and 7.3 µ /L of carbamazepine-10,ll-epoxide during carbamazepine treatment. The mean carbamazepine-10,ll-epoxide concentration was 25 µ /L dur¬ ing epoxide therapy. Morning plas¬ ma concentrations were stable during carbamazepine treatment, whereas there were marked day-to-day varia¬ tions in the steady-state concentra¬ tions of carbamazepine-10,ll-epoxide during treatment with carbamazepine-10,ll-epoxide (Fig 1). Morning steady-state carbamazepine-10,11epoxide levels above 30 µ /L oc¬
curred
on
at
least
one
occasion in five
patients (patients 2 through 6). High¬ est steady-state concentrations of carbamazepine-10,ll-epoxide were 37 µ /L (patient 4), 57 µ /L (patient 5), and 49 µ /L (patient 6). Diurnal plasma concentrations (7 am to 7 pm) of both carbamazepine and carbamazepine-10,ll-epoxide were stable during carbamazepine treatment. In contrast,
diurnal fluctuations in plasma levels were variable and often pronounced when the patients were treated with
carbamazepine-10,ll-epoxide. In some patients the time of peak concentra¬ tion appeared to be unrelated to the time of dosing. The peak concentration sometimes occurred immediately be¬ fore the morning dose and was fol¬ lowed by a steady decline in the plasma
concentration
over the day despite re¬ peated epoxide dosing. Steady-state
concentrations of the concomitant an-
tiepileptic drug therapy were stable throughout the study. On his own request, patient 7 dis¬ continued the trial prematurely, be¬ cause of four generalized seizures in the washout period after changing to carbamazepine-10,ll-epoxide. Side ef¬ fects were, however, evaluated accord¬ ing to the protocol before discontinua¬ tion immediately after the washout period and are included in the results.
The number of seizures for each pa¬ tient during the two treatment periods is shown in Fig 2. There was no signif¬ icant difference in seizure control be¬ tween the treatment periods (Wilcoxon Signed-Rank Test). None of the patients experienced subjective side effects at any time of the study. There was a trend toward improved perfor¬ mance in 9 of the 10 neuropsychological tests during treatment with car-
bamazepine-10,ll-epoxide compared carbamazepine (Table 2). The improvement of finger motor speed with dominant hand and logical rea¬ soning was significant (P < .05, paired t test). There was also a significant in¬ crease (P < .05 Wilcoxon Signed-Rank Test) in serum sodium levels when patients were changed to carbamazepine-10,ll-epoxide therapy (Fig 3). There was no significant difference with
in results of other biochemical tests.
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
Table
Carbamazepine-
Test
Finger
motor
speed, taps/10
2.—Neuropsychological Testing* No.
Carbamazepine
10,11-epoxide
s
Dominant hand
7
Nondominant hand
7
55.8 ± 3.4 51.3 ± 7.3
58.0 ±
3.0t
7 72.1 ±20.5 77.9 ± 12.3 Alternating hand Simple reaction time, latency, ms 707 ± 87.7 696 ± 95.6 Complex reaction time, latency, ms 6 Short-term memory scanning, latency; ms 513 ± 92.8 1 digit 5* 3 digits 696 ± 125.6 640 ± 141.8 5 713 ± 153.3 5 digits 779 ± 141.8 5 Symbol digit substitution, 7 38.9 ± 14.2 34.8 ± 10.5 latency, (s) 7.76 ± 2.13 6.22 ± 2.07f Logical Reasoning, latency, s 5 * Values are mean ± SD. Best performance was with carbamazepine-10,11-epoxide, except for short-term memory scanning, one digit.
tP< .05. ÍPatients excluded because of partial seizure at the time of testing.
COMMENT
Plasma clearance of carbamazepine10,11-epoxide was about fourfold higher than the clearance of carba¬ mazepine in a previous single-dose study.28 We therefore expected carbamazepine-10,ll-epoxide plasma lev¬
els to be much lower than those we ac¬ tually found in this study since the epoxide and carbamazepine were given in the same dosage. A saturable firstpass elimination of carbamazepine10,11-epoxide may be a partial expla¬ nation for the unexpectedly high epox¬ ide concentrations. Higher epoxide doses were used in this study than in
previous pharmacokinetic investiga¬ tions, and studies of carbamazepine10,11-epoxide and valpromide interac¬
tions indicate that a saturable firstpass elimination may exist.29 Another obvious factor contributing to the high levels was the formulation of carbamazepine-10,ll-epoxide. The absorption from the tablet appeared to be erratic and unpredictable. This was probably an effect of the enteric coat¬ ing allowing the epoxide to be absorbed only after the tablet had passed into the small intestine. This, together with a short plasma half-life of carbamazepine-10,ll-epoxide,28 resulted in the pronounced diurnal fluctuations in epoxide plasma concentrations. These fluctuations and the day-to-day varia¬ tions in carbamazepine-10,ll-epoxide concentrations may well be seizure provoking. The pharmacokinetic pro¬ file of the enteric-coated formulation was therefore considered unacceptable for an extended study of patients with epilepsy. The assessment of the antiepileptic effect of carbamazepine-
10,11-epoxide was clearly complicated by the shortcomings of the enteric-
coated
—
formulation.
Nevertheless, patient who dropped out after the washout period, there was no tendency toward deterioration of seizure control during treatment with carbamazepine-epoxide. apart from
Fig 3. Serum sodium of seven epileptic pa¬ tients during treatment with carbamazepine and carbamazepine-10,11-epoxide. Serum sodium level was significantly higher during treatment with epoxide (P < .05, Wilcoxon Signed-Rank Test). Shaded area shows nor¬ mal range of serum sodium levels.
one
If evaluation of the anticonvulsant effect was hampered, the unexpect¬
edly high carbamazepine-10,ll-epox-
ide concentrations on the other hand offered an excellent opportunity for direct evaluation of the neurotoxic po¬ tential of this compound. The plasma levels reached during carbamazepine10,11-epoxide therapy were several times higher than those generally found in patients receiving carba¬
mazepine therapy.4 Higher plasma
concentrations have rarely been re¬ ported." The plasma levels were also considerably higher than those pro¬ posed to be toxic during carba¬ mazepine therapy.1216 It is therefore remarkable that none of the patients reported side effects. These subjective assessments were supported by the re¬ sults of the neuropsychological tests. In fact, a trend toward improvement was consistently found when patients switched to carbamazepinewere 10,11-epoxide therapy. The differences were statistically significant for two of the tests despite the small patient number. Fatigue and learning effects were minimal in response-speed vari¬ ables of these neuropsychological tests.23 It is therefore unlikely that the nonrandomized treatment sequence, carbamazepine followed by carbamazepine-10,ll-epoxide, influenced the results. It may be argued that we studied patients with an extremely high toler¬ ance to side effects. However, the car¬ bamazepine dose that patients re¬ ceived when entering the study was the
highest tolerable for each individual. All patients had previously experi¬ enced neurotoxic, dose-dependent side effects with higher carbamazepine dosage. Another interesting observation that low serum sodium levels were normalized when patients were switched from carbamazepine to car-
was
bamazepine-10,ll-epoxide treatment. Hyponatremia is a not-infrequent dose-dependent side effect of carba¬ mazepine therapy.30 This indicates that the parent drug and not the epox¬ ide is responsible for the water intox¬ ication reported. Since these were the first epileptic patients to be treated directly with the carbamazepine-10,ll-epoxide, study was associated with the limita¬ tions of a pilot study: a small patient sample receiving combination ther¬ apy, a nonrandomized design, and only a single-blind control. The present de¬ sign, however, yields more reliable re¬ sults than previous attempts to assess the effects of the epoxide metabolite during carbamazepine therapy. The following conclusions may be drawn. (1) It is highly unlikely that the epoxide metabolite is a major contrib¬ utor to the neurotoxic side effects of
carbamazepine therapy. (2) Routine monitoring of the plasma concentra¬ tion of the epoxide metabolite during carbamazepine therapy is not justi¬ fied. This pilot study does not provide data to say if carbamazepine-10,11epoxide would be an advantageous al¬ ternative to carbamazepine therapy. The pharmacokinetic problems make
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
carbamazepine-10,ll-epoxide less at¬ tractive. The favorable results of the side effects evaluation, however, prompt further studies, for which an improved formulation of the epoxide is necessary. This study was supported by grants from the Karolinska Institute, Stockholm, Sweden, Stiftelsen Stora Sköndal, Karin and Sven Sander's Foundation, and the Swedish Medical Research Council (grant 3902).
Carbamazepine-10,ll-epoxide was kindly sup¬ plied by Ciba-Geigy, Basel, Switzerland. References
1. Mattson RH, Cramer JA, Collins JF, et al. Comparison of carbamazepine, phenobarbital, phenytoin and primidone in partial and secondarily generalized tonic-clonic seizures. N Engl J
Med. 1985;313:145-151. 2. Blom S. Tic douloureux treated with new anticonvulsant: experiences with G 32883. Arch Neurol. 1963;9:285-290. 3. Post RM, Uhde TW, Ballenger JC. The efficacy of carbamazepine in affective illness. In: Usdin E, \l=A%o\sbergM, Bertilsson L, Sj\l=o"\qvistF, eds. Frontiers in Biochemical and Pharmacological Research in Depression. New York, NY: Raven Press; 1984:421-437. 4. Bertilsson L, Tomson T. Clinical pharmacokinetics and pharmacological effects of carbamazepine and carbamazepine-10,11-epoxide. Clin Pharmacokinet. 1986;11:177-198. 5. Patsalos PN, Stephenson TJ, Krishna S, Elyas AA, Lascelle PT, Wiles CM. Side effects induced by carbamazepine-10,11-epoxide. Lancet.
1985;2:496.
6. McKee RJW, Larkin JG, Brodie MJ. Acute psychosis with carbamazepine and sodium val-
proate. Lancet. 1989;1:167. 7. Kerr M, Levy RH. Inhibition of epoxide hydrolase by anticonvulsants and risk of teratogenicity. Lancet. 1989;1:610-611.
8. Theodore WH, Narang PK, Holmes MD, Reeves P, Nice FJ. Carbamazepine and its epoxide: relation of plasma levels to toxicity and seizure control. Ann Neurol. 1989;25:194-196. 9. Lindhout D, H\l=o"\ppenerRJEA, Meinardi H. Teratogenicity of antiepileptic drug combinations with special emphasis on epoxidation (of carba-
mazepine). Epilepsia. 1984;25:77-83. 10. Kutt H, Solomon GE, Dhar AK, Resor SR Jr, Krall RL, Morselli PL. Effects of progabide on carbamazepine epoxide and carbamazepine concentrations in plasma. Epilepsia. 1984;25:674. 11. Meijer JWA, Binnie CD, Debets RMC, van Parys JAP, De Beer-Pawlikowski NKB. Possible hazard of valpromide-carbamazepine combination therapy in epilepsy. Lancet. 1984;1:802. 12. Rosenfeld WE, Holmes GB, Graves NM, Peterson AL. Carbamazepine-10,11-epoxide serum concentration as a predictor of toxicity. Epilepsia. 1987;28:580-581. 13. Dam M. Oxcarbazepine. In: Dam M, Johannessen SI, Nilsson B, Sillanp\l=a"\\l=a"\M, eds. Epilepsy: Progress in Treatment. New York, NY: John Wiley & Sons Inc; 1987:257-261. 14. Levy RH, Kerr BM. Clinical pharmacokinetics of carbamazepine. J Clin Psychiatry. 1988;
49:58-62. 15. Gillham RA, Williams N, Wiedmann K, Butler E, Larkin JG, Brodie MJ. Concentration\x=req-\ effect relationships with carbamazepine and its epoxide on psychomotor and cognitive function in epileptic patients. J Neurol Neurosurg Psychia-
try. 1988;51:929-933.
16. Schoeman JF, Elyas AA, Brett EM, Lascelles PT. Correlation between plasma carbamazepine-10,11-epoxide concentration and drug side-effects in children with epilepsy. Dev Med Child Neurol. 1984;26:756-764. 17. Tomson T, Bertilsson L. Potent therapeutic effect of carbamazepine-10,11-epoxide in trigeminal neuralgia. Arch Neurol. 1984;41:598-601. 18. Tomson T, Svensson JO, Hilton-Brown P. Intraindividual dose to plasma concentration relationship of carbamazepine: indication of dose\x=req-\ dependent induction of metabolism. Ther Drug Monit. In press. 19. Spina E, Tomson T, Svensson JO, Faigle
JW, Bertilsson L. Single-dose kinetics of an enteric-coated formulation of carbamazepine-10,11\x=req-\ epoxide, an active metabolite of carbamazepine. Ther Drug Monit. 1988;10:382-385. 20. Novelly RA, Schwartz MM, Mattsson RH,
toxicity associated with antiepileptic drugs: concepts and methods of assessment. Epilepsia. 1986;27:331-340. 21. Andrewes DG, Bullen JG, Tomlinson L, Elwes RDC, Reynolds EH. A comparative study of the cognitive effects of phenytoin and carbamazepine in new referrals with epilepsy. EpilepCramer JA. Behavioral
sia. 1986;27:128-134. 22. McPhee GJA, Goldie C, Roulston D, et al. Effect of carbamazepine on psychomotor performance in naive subjects. Eur J Clin Pharmacol.
1988;30:37-42.
23. Gamberale F, Iregren A, Kjellberg A. Computerized performance testing in neurotoxicology. Workshop on Behavioral Aspects of Neurotoxicity. Canberra, Australia: In press. 24. Iregren A, Gamberale F, Kjellberg A. A microcomputer based behavioral testing system. In: Neurobehavioral Methods in Occupational and Environmental Health. Geneva, Switzerland: World Health Organization; 1985;75-80. 25. Sternberg S. High-speed scanning in human memory. Science. 1966;153:652-654. 26. Wechsler D. Wechsler Adult Intelligence Scale-Revised. Cleveland, Ohio: Psychological Corp; 1981. 27. Baddeley AD. A 3-minute reasoning test based on grammatical transformation. Psychonomic Sci. 1968;10:341-342. 28. Tomson T, Tybring G, Bertilsson L. Single dose kinetics and metabolism of carbamazepine\x=req-\ 10,11-epoxide. Clin Pharmacol Ther. 1983;33:58\x=req-\
65. 29. Pisani F, Fazio A, Oteri G, Spina E, Perucca E, Bertilsson L. Effect of valpromide on the pharmacokinetics of carbamazepine-10,11-epoxide. Br J Clin Pharmacol. 1988;25:611-613. 30. Perucca E, Garratt A, Hebdrige S, Richens A. Water intoxication in epileptic patients receiving carbamazepine. J Neurol Neurosurg Psychia-
try. 1978;41:713-718.
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
Study
Torbj\l=o"\rnTomson, MD; Ove Almkvist, BSc; Bengt Y. Nilsson, MD; Jan-Olof Svensson, Engr; Leif Bertilsson, PhD \s=b\ The effects of carbamazepine-10,11\x=req-\ epoxide, an active metabolite of carbamazepine, were evaluated in seven outpatients with frequent epileptic seizures. The study included an initial 4-week period with the carbamazepine dose optimized for each patient. Patients were then crossed over, dose by dose, to carbamazepine-10,11-epoxide and followed up for another 4 weeks. Dosing was single
blind. The evaluation of the anticonvulsant effect was hampered by marked fluctuations in plasma levels during treatment with carbamazepine-10,11-epoxide. There was, however, no significant change in seizure control. During epoxide treatment, no subjective side effects were
reported despite epoxide plasma
concen-
trations up to 57 \g=m\mol/L. Neuropsychological assessment revealed a significant improvement in finger motor speed and logical reasoning during the carbamazepine-10,11-epoxide period. Subnormal serum sodium levels in two patients were normalized after switching from carbamazepine to the epoxide. Continued investigations with this active metabolite of carbamazepine in epilepsy are therefore
justified. (Arch Neurol. 1990;47:888-892)
Accepted for publication November 3, 1989. From the Departments of Neurology (Dr Tomson) and Clinical Neurophysiology (Dr Nilsson) S\l=o"\dersjukhuset;Department of Psychiatry, St G\l=o"\ran'sHospital (Dr Almkvist); and Department of Clinical Pharmacology, Huddinge University Hospital, Karolinska Institute (Dr Bertilsson and Mr Svensson), Stockholm, Sweden. Reprint requests to Department of Neurology, S\l=o"\dersjukhuset,S-100 64 Stockholm, Sweden (Dr
Tomson).
is Carbamazepine the treatment
first-line drug in of epilepsy1 and trigeminal neuralgia2 and is now also frequently used in bipolar depression.3 Carbamazepine is metabolized to carbamazepine-10,ll-epoxide, which is an active metabolite that appears in plasma at concentrations of 10% to 40% of the parent compound." The possible contribution of this metabo¬ lite to the effects of carbamazepine therapy has recently been debated intensely.58 It is a common opinion that carbamazepine-10,ll-epoxide to a large extent contributes to the side ef¬ fects of carbamazepine therapy.67·915 Schoeman et al16 have suggested that neurotoxic side effects are a rule in pa¬ tients treated with carbamaze¬ pine when carbamazepine-10,ll-epoxide levels exceed 9 µ /L. Rosenfeld et al12 claimed that all patients with epoxide concentrations of 15 /¿mol/L or a
during carbamazepine treat¬ ment were either clinically intoxicated or unable to complain owing to mental more
retardation. These and other authors recommend routine monitoring of carbamazepine-10,ll-epoxide plasma lev¬ els to control for side effects of car¬ bamazepine therapy. The high epoxide concentrations in these studies oc¬ curred almost exclusively in patients treated with carbamazepine in combi¬ nation with other drugs, eg, phenytoin, which induce the epoxidation of car¬ bamazepine. Hence, it was not possible to determine if the observed side ef¬ fects were caused directly by the drug given in addition to carbamazepine or by the high epoxide levels. A reliable assessment of the effects
of carbamazepine-10,ll-epoxide re¬ quires direct administration of the ep¬ oxide to patients. In the only previous study of carbamazepine-10,ll-epoxide given as such to patients, we found a potent pain-relieving effect in trigeminal neuralgia, and no side effects were observed.17 We now report the results of a pilot study of carbamazepine10,11-epoxide in the treatment of epi¬
lepsy.
PATIENTS AND METHODS Patients
The study included seven adult outpa¬ tients (aged 21 to 56 years) with complex partial seizures with or without secondary generalization. All patients had at least five seizures per month before the study despite optimal treatment with carbamazepine given in combination with one or two other antiepileptic drugs (Table 1). The carba¬ mazepine dosage was unchanged for at least 6 weeks before the study. The duration of epilepsy ranged from 6 to 41 years. Patients with brain tumors and progressive brain disorders were excluded.
Study Design The study had a single-blind crossover design. Patients were followed up during a first 4-week period with their original opti¬ mal carbamazepine treatment dosed four times a day (7 am, noon, 5 pm, and 10 pm; pe¬ riod A). During the subsequent week they were switched over, dose by dose, from carbamazepine to carbamazepine-10,11epoxide therapy four times a day. This was followed by a 1-week washout period of possible carbamazepine effects and there¬ after by a 4-week observation period with carbamazepine-10,ll-epoxide (period B). The study design is illustrated in Pig 1. Pa¬ tients were kept unaware of changes in the
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
Table
1.—Dosage and Plasma Concentrations of Carbamazepine and Carbamazepine-10,11-Epoxide During Two Treatment Periods in Seven Epileptic Patients* Plasma Concentration, umol L
Patient No.
Treatment Period
Active
Drug Therapy, mg/d Carbamazepine, 1200 Carbamazepine-10,11-epoxide, Carbamazepine, 1200 Carbamazepine-10,11-epoxide, Carbamazepine, 800 Carbamazepine-10,11-epoxide, Carbamazepine, 1400 Carbamazepine-10,11 -epoxide, Carbamazepine, 1200 Carbamazepine-10,11 -epoxide, Carbamazepine, 1200 Carbamazepine-10,11 -epoxide, Carbamazepine, 600 Carbamazepine-10,11-epoxide, observations before morning dose at 7 am.
Concomitant Therapy
Carbamazepine-
Carbamazepine
10,11-epoxide
35
6.2
1200 1200 800
32 1400
24 1200
31
6.9
1200
29 600
8.5
22t
Plasma
Drug, mg/d Phenytoin, 200 Phenytoin, 200 Phenytoin, 50 Phenytoin, 50 Phénobarbital, 50 Phénobarbital, 50 Phénobarbital, 25 Phénobarbital, 25 Phenytoin, 200 Phenytoin, 200
Clonazepam, 1 Clonazepam, 1 Valproate sodium, Valproate, 3000
Concentration,
µ
/L
10
45
0.024 3000
*Mean of all tSingle observation.
therapy by use of carbamazepine-10,ll-epoxide-matched placebo tablets during pe¬ riod A and carbamazepine-matched pla¬ cebo tablets during period B. Concomitant
therapy (Table 1) was unchanged through¬ study, which altogether lasted 10
out the
weeks. Patients were seen at weekly intervals in the outpatient clinic and in addition hospi¬ talized during 2 days at the end of each 4week period for registration of clinical effects. Drugs were given to the patients in
specially designed dispensers containing
the separate dosages for 1 week. Compli¬ ance was tested by checking the dispensers at each visit. Blood samples for drug analysis were ob¬ tained immediately before the morning dose at each weekly visit, and in addition every second hour (7am to 7 pm) during the last day of the two 4-week periods. The plasma concentrations of carbamazepine and carbamazepine-10,ll-epoxide were de¬ termined with high-performance liquid
chromatography.18
Drugs
Carbamazepine was given as plain Tegretol tablets (Ciba-Geigy Corp). Carbamazepine-10,ll-epoxide in unstable is gas¬ tric juice. Enteric-coated tablets provided by Ciba-Geigy Corp" were therefore used. Evaluation of Effects
Seizures were recorded by the patient or, when appropriate, relatives on a calendar that was checked at every visit. All sponta¬ neous reports of side effects were recorded, and in addition patients were actively ques¬ tioned at each visit about neurotoxic symp¬
following a special questionnaire (available on request). The following tests were performed during the last 2 days of the two treatment periods with the patient in hospital. The different tests were made at the same time of day on the two occasions. Biochemical Tests.—Hemoglobin, hematocrit, red and white blood cell counts, diftoms
as illustrated by patient 1. Carbamazepine and carbamazepine-10,11-epoxide were administered four times daily. Trough morning plasma levels of carbamazepine (open squares) and carbamazepine-10,11-epoxide (solid squares) are indicated in top panel.
Fig 1.—Design and results of study
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
ferential and platelet counts formed. Electrolytes, albumin,
were
per¬
creatinine, aminotransferases, -glutamyltransf erase, alkaline phosphatase, folate, thyrotropin, and immunoglobulins were all measured in serum.
Psychological Assessment.—Based on re¬ concerning behavioral effects associated with antiepileptic drugs20 and especially carbamazepine,2122 six computercent studies
administered tests of load23-24 were selected to
varying assess
status.
mental
cognitive
Finger Motor Speed.—The subject was instructed to tap a button as fast as possi¬ ble in five trials of three conditions: the in¬ dex finger of the dominant hand, the index finger of the nondominant hand, and alter¬ nating the right and left index fingers. The mean number of taps for five 10-second pe¬ riods was used to measure performance. Simple Reaction Time.—The subject had to hit a button when a time-randomized signal appeared on the screen. The mean latency during a 5-minute testing time was used to measure performance. Complex Reaction Time.—The subject was instructed to respond as fast as possi¬ ble by pushing a button when the word meaning (one of four color words) and the print color (one of four colors) corresponded to each other, and otherwise not to respond. The mean reaction time during a 10-minute testing period was used to measure perfor¬ mance.
Short-term Memory Scanning.—Fol¬ lowing a paradigm developed by Sternberg,25 the subject had to push as fast as possible one of two buttons as to whether a random probe digit matched (positive re¬ sponses) one of the digits of a preestablished set of digits or not (negative respons¬ es). The mean latency for positive responses was, timed for three different set sizes (1,3, and 5 digits in the set). Symbol Digit Substitution.—In this test, a relative of the Digit Symbol Test of the Wechsler Adult Intelligence Scale-Re¬ vised,26 the subject had to match digits to symbols according to an existing code scheme as fast as possible. The mean la¬
tency of five trials
was
measured after
a
training period. Logical Reasoning.—The subject had to decide whether a verbal premise, eg, A pre¬ cedes B, was false or true compared with a picture of the letters, eg, A. The number of correct judgments of 64 items and the mean latency were registered. This test is a computerized version of the Baddeley ver¬ bal reasoning task.27 Patients were also subjected to a neurophysiological examination including elec¬ troencephalogram, visual evoked poten¬ tials, and visual contrast sensitivity. De¬ tails of this evaluation will be reported elsewhere. Informed consent from all patients.
was
obtained
RESULTS
The steady-state plasma concentra¬ tions of carbamazepine and carbamazepine-10,ll-epoxide during the two treatment periods are given in Table 1.
Fig 2.—Total number of seizures of seven patients treated with carbamazepine (left) and carba¬ mazepine- 10,11 -epoxide (right). Note that seizures during epoxide treatment of patient 7 refer to the crossover and washout period, after which the patient dropped out of the study. No significant difference in seizure frequency was found (Wilcoxon Signed-Rank Test).
For the whole patient group, mean steady-state concentrations were 32 µ /L of carbamazepine and 7.3 µ /L of carbamazepine-10,ll-epoxide during carbamazepine treatment. The mean carbamazepine-10,ll-epoxide concentration was 25 µ /L dur¬ ing epoxide therapy. Morning plas¬ ma concentrations were stable during carbamazepine treatment, whereas there were marked day-to-day varia¬ tions in the steady-state concentra¬ tions of carbamazepine-10,ll-epoxide during treatment with carbamazepine-10,ll-epoxide (Fig 1). Morning steady-state carbamazepine-10,11epoxide levels above 30 µ /L oc¬
curred
on
at
least
one
occasion in five
patients (patients 2 through 6). High¬ est steady-state concentrations of carbamazepine-10,ll-epoxide were 37 µ /L (patient 4), 57 µ /L (patient 5), and 49 µ /L (patient 6). Diurnal plasma concentrations (7 am to 7 pm) of both carbamazepine and carbamazepine-10,ll-epoxide were stable during carbamazepine treatment. In contrast,
diurnal fluctuations in plasma levels were variable and often pronounced when the patients were treated with
carbamazepine-10,ll-epoxide. In some patients the time of peak concentra¬ tion appeared to be unrelated to the time of dosing. The peak concentration sometimes occurred immediately be¬ fore the morning dose and was fol¬ lowed by a steady decline in the plasma
concentration
over the day despite re¬ peated epoxide dosing. Steady-state
concentrations of the concomitant an-
tiepileptic drug therapy were stable throughout the study. On his own request, patient 7 dis¬ continued the trial prematurely, be¬ cause of four generalized seizures in the washout period after changing to carbamazepine-10,ll-epoxide. Side ef¬ fects were, however, evaluated accord¬ ing to the protocol before discontinua¬ tion immediately after the washout period and are included in the results.
The number of seizures for each pa¬ tient during the two treatment periods is shown in Fig 2. There was no signif¬ icant difference in seizure control be¬ tween the treatment periods (Wilcoxon Signed-Rank Test). None of the patients experienced subjective side effects at any time of the study. There was a trend toward improved perfor¬ mance in 9 of the 10 neuropsychological tests during treatment with car-
bamazepine-10,ll-epoxide compared carbamazepine (Table 2). The improvement of finger motor speed with dominant hand and logical rea¬ soning was significant (P < .05, paired t test). There was also a significant in¬ crease (P < .05 Wilcoxon Signed-Rank Test) in serum sodium levels when patients were changed to carbamazepine-10,ll-epoxide therapy (Fig 3). There was no significant difference with
in results of other biochemical tests.
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
Table
Carbamazepine-
Test
Finger
motor
speed, taps/10
2.—Neuropsychological Testing* No.
Carbamazepine
10,11-epoxide
s
Dominant hand
7
Nondominant hand
7
55.8 ± 3.4 51.3 ± 7.3
58.0 ±
3.0t
7 72.1 ±20.5 77.9 ± 12.3 Alternating hand Simple reaction time, latency, ms 707 ± 87.7 696 ± 95.6 Complex reaction time, latency, ms 6 Short-term memory scanning, latency; ms 513 ± 92.8 1 digit 5* 3 digits 696 ± 125.6 640 ± 141.8 5 713 ± 153.3 5 digits 779 ± 141.8 5 Symbol digit substitution, 7 38.9 ± 14.2 34.8 ± 10.5 latency, (s) 7.76 ± 2.13 6.22 ± 2.07f Logical Reasoning, latency, s 5 * Values are mean ± SD. Best performance was with carbamazepine-10,11-epoxide, except for short-term memory scanning, one digit.
tP< .05. ÍPatients excluded because of partial seizure at the time of testing.
COMMENT
Plasma clearance of carbamazepine10,11-epoxide was about fourfold higher than the clearance of carba¬ mazepine in a previous single-dose study.28 We therefore expected carbamazepine-10,ll-epoxide plasma lev¬
els to be much lower than those we ac¬ tually found in this study since the epoxide and carbamazepine were given in the same dosage. A saturable firstpass elimination of carbamazepine10,11-epoxide may be a partial expla¬ nation for the unexpectedly high epox¬ ide concentrations. Higher epoxide doses were used in this study than in
previous pharmacokinetic investiga¬ tions, and studies of carbamazepine10,11-epoxide and valpromide interac¬
tions indicate that a saturable firstpass elimination may exist.29 Another obvious factor contributing to the high levels was the formulation of carbamazepine-10,ll-epoxide. The absorption from the tablet appeared to be erratic and unpredictable. This was probably an effect of the enteric coat¬ ing allowing the epoxide to be absorbed only after the tablet had passed into the small intestine. This, together with a short plasma half-life of carbamazepine-10,ll-epoxide,28 resulted in the pronounced diurnal fluctuations in epoxide plasma concentrations. These fluctuations and the day-to-day varia¬ tions in carbamazepine-10,ll-epoxide concentrations may well be seizure provoking. The pharmacokinetic pro¬ file of the enteric-coated formulation was therefore considered unacceptable for an extended study of patients with epilepsy. The assessment of the antiepileptic effect of carbamazepine-
10,11-epoxide was clearly complicated by the shortcomings of the enteric-
coated
—
formulation.
Nevertheless, patient who dropped out after the washout period, there was no tendency toward deterioration of seizure control during treatment with carbamazepine-epoxide. apart from
Fig 3. Serum sodium of seven epileptic pa¬ tients during treatment with carbamazepine and carbamazepine-10,11-epoxide. Serum sodium level was significantly higher during treatment with epoxide (P < .05, Wilcoxon Signed-Rank Test). Shaded area shows nor¬ mal range of serum sodium levels.
one
If evaluation of the anticonvulsant effect was hampered, the unexpect¬
edly high carbamazepine-10,ll-epox-
ide concentrations on the other hand offered an excellent opportunity for direct evaluation of the neurotoxic po¬ tential of this compound. The plasma levels reached during carbamazepine10,11-epoxide therapy were several times higher than those generally found in patients receiving carba¬
mazepine therapy.4 Higher plasma
concentrations have rarely been re¬ ported." The plasma levels were also considerably higher than those pro¬ posed to be toxic during carba¬ mazepine therapy.1216 It is therefore remarkable that none of the patients reported side effects. These subjective assessments were supported by the re¬ sults of the neuropsychological tests. In fact, a trend toward improvement was consistently found when patients switched to carbamazepinewere 10,11-epoxide therapy. The differences were statistically significant for two of the tests despite the small patient number. Fatigue and learning effects were minimal in response-speed vari¬ ables of these neuropsychological tests.23 It is therefore unlikely that the nonrandomized treatment sequence, carbamazepine followed by carbamazepine-10,ll-epoxide, influenced the results. It may be argued that we studied patients with an extremely high toler¬ ance to side effects. However, the car¬ bamazepine dose that patients re¬ ceived when entering the study was the
highest tolerable for each individual. All patients had previously experi¬ enced neurotoxic, dose-dependent side effects with higher carbamazepine dosage. Another interesting observation that low serum sodium levels were normalized when patients were switched from carbamazepine to car-
was
bamazepine-10,ll-epoxide treatment. Hyponatremia is a not-infrequent dose-dependent side effect of carba¬ mazepine therapy.30 This indicates that the parent drug and not the epox¬ ide is responsible for the water intox¬ ication reported. Since these were the first epileptic patients to be treated directly with the carbamazepine-10,ll-epoxide, study was associated with the limita¬ tions of a pilot study: a small patient sample receiving combination ther¬ apy, a nonrandomized design, and only a single-blind control. The present de¬ sign, however, yields more reliable re¬ sults than previous attempts to assess the effects of the epoxide metabolite during carbamazepine therapy. The following conclusions may be drawn. (1) It is highly unlikely that the epoxide metabolite is a major contrib¬ utor to the neurotoxic side effects of
carbamazepine therapy. (2) Routine monitoring of the plasma concentra¬ tion of the epoxide metabolite during carbamazepine therapy is not justi¬ fied. This pilot study does not provide data to say if carbamazepine-10,11epoxide would be an advantageous al¬ ternative to carbamazepine therapy. The pharmacokinetic problems make
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
carbamazepine-10,ll-epoxide less at¬ tractive. The favorable results of the side effects evaluation, however, prompt further studies, for which an improved formulation of the epoxide is necessary. This study was supported by grants from the Karolinska Institute, Stockholm, Sweden, Stiftelsen Stora Sköndal, Karin and Sven Sander's Foundation, and the Swedish Medical Research Council (grant 3902).
Carbamazepine-10,ll-epoxide was kindly sup¬ plied by Ciba-Geigy, Basel, Switzerland. References
1. Mattson RH, Cramer JA, Collins JF, et al. Comparison of carbamazepine, phenobarbital, phenytoin and primidone in partial and secondarily generalized tonic-clonic seizures. N Engl J
Med. 1985;313:145-151. 2. Blom S. Tic douloureux treated with new anticonvulsant: experiences with G 32883. Arch Neurol. 1963;9:285-290. 3. Post RM, Uhde TW, Ballenger JC. The efficacy of carbamazepine in affective illness. In: Usdin E, \l=A%o\sbergM, Bertilsson L, Sj\l=o"\qvistF, eds. Frontiers in Biochemical and Pharmacological Research in Depression. New York, NY: Raven Press; 1984:421-437. 4. Bertilsson L, Tomson T. Clinical pharmacokinetics and pharmacological effects of carbamazepine and carbamazepine-10,11-epoxide. Clin Pharmacokinet. 1986;11:177-198. 5. Patsalos PN, Stephenson TJ, Krishna S, Elyas AA, Lascelle PT, Wiles CM. Side effects induced by carbamazepine-10,11-epoxide. Lancet.
1985;2:496.
6. McKee RJW, Larkin JG, Brodie MJ. Acute psychosis with carbamazepine and sodium val-
proate. Lancet. 1989;1:167. 7. Kerr M, Levy RH. Inhibition of epoxide hydrolase by anticonvulsants and risk of teratogenicity. Lancet. 1989;1:610-611.
8. Theodore WH, Narang PK, Holmes MD, Reeves P, Nice FJ. Carbamazepine and its epoxide: relation of plasma levels to toxicity and seizure control. Ann Neurol. 1989;25:194-196. 9. Lindhout D, H\l=o"\ppenerRJEA, Meinardi H. Teratogenicity of antiepileptic drug combinations with special emphasis on epoxidation (of carba-
mazepine). Epilepsia. 1984;25:77-83. 10. Kutt H, Solomon GE, Dhar AK, Resor SR Jr, Krall RL, Morselli PL. Effects of progabide on carbamazepine epoxide and carbamazepine concentrations in plasma. Epilepsia. 1984;25:674. 11. Meijer JWA, Binnie CD, Debets RMC, van Parys JAP, De Beer-Pawlikowski NKB. Possible hazard of valpromide-carbamazepine combination therapy in epilepsy. Lancet. 1984;1:802. 12. Rosenfeld WE, Holmes GB, Graves NM, Peterson AL. Carbamazepine-10,11-epoxide serum concentration as a predictor of toxicity. Epilepsia. 1987;28:580-581. 13. Dam M. Oxcarbazepine. In: Dam M, Johannessen SI, Nilsson B, Sillanp\l=a"\\l=a"\M, eds. Epilepsy: Progress in Treatment. New York, NY: John Wiley & Sons Inc; 1987:257-261. 14. Levy RH, Kerr BM. Clinical pharmacokinetics of carbamazepine. J Clin Psychiatry. 1988;
49:58-62. 15. Gillham RA, Williams N, Wiedmann K, Butler E, Larkin JG, Brodie MJ. Concentration\x=req-\ effect relationships with carbamazepine and its epoxide on psychomotor and cognitive function in epileptic patients. J Neurol Neurosurg Psychia-
try. 1988;51:929-933.
16. Schoeman JF, Elyas AA, Brett EM, Lascelles PT. Correlation between plasma carbamazepine-10,11-epoxide concentration and drug side-effects in children with epilepsy. Dev Med Child Neurol. 1984;26:756-764. 17. Tomson T, Bertilsson L. Potent therapeutic effect of carbamazepine-10,11-epoxide in trigeminal neuralgia. Arch Neurol. 1984;41:598-601. 18. Tomson T, Svensson JO, Hilton-Brown P. Intraindividual dose to plasma concentration relationship of carbamazepine: indication of dose\x=req-\ dependent induction of metabolism. Ther Drug Monit. In press. 19. Spina E, Tomson T, Svensson JO, Faigle
JW, Bertilsson L. Single-dose kinetics of an enteric-coated formulation of carbamazepine-10,11\x=req-\ epoxide, an active metabolite of carbamazepine. Ther Drug Monit. 1988;10:382-385. 20. Novelly RA, Schwartz MM, Mattsson RH,
toxicity associated with antiepileptic drugs: concepts and methods of assessment. Epilepsia. 1986;27:331-340. 21. Andrewes DG, Bullen JG, Tomlinson L, Elwes RDC, Reynolds EH. A comparative study of the cognitive effects of phenytoin and carbamazepine in new referrals with epilepsy. EpilepCramer JA. Behavioral
sia. 1986;27:128-134. 22. McPhee GJA, Goldie C, Roulston D, et al. Effect of carbamazepine on psychomotor performance in naive subjects. Eur J Clin Pharmacol.
1988;30:37-42.
23. Gamberale F, Iregren A, Kjellberg A. Computerized performance testing in neurotoxicology. Workshop on Behavioral Aspects of Neurotoxicity. Canberra, Australia: In press. 24. Iregren A, Gamberale F, Kjellberg A. A microcomputer based behavioral testing system. In: Neurobehavioral Methods in Occupational and Environmental Health. Geneva, Switzerland: World Health Organization; 1985;75-80. 25. Sternberg S. High-speed scanning in human memory. Science. 1966;153:652-654. 26. Wechsler D. Wechsler Adult Intelligence Scale-Revised. Cleveland, Ohio: Psychological Corp; 1981. 27. Baddeley AD. A 3-minute reasoning test based on grammatical transformation. Psychonomic Sci. 1968;10:341-342. 28. Tomson T, Tybring G, Bertilsson L. Single dose kinetics and metabolism of carbamazepine\x=req-\ 10,11-epoxide. Clin Pharmacol Ther. 1983;33:58\x=req-\
65. 29. Pisani F, Fazio A, Oteri G, Spina E, Perucca E, Bertilsson L. Effect of valpromide on the pharmacokinetics of carbamazepine-10,11-epoxide. Br J Clin Pharmacol. 1988;25:611-613. 30. Perucca E, Garratt A, Hebdrige S, Richens A. Water intoxication in epileptic patients receiving carbamazepine. J Neurol Neurosurg Psychia-
try. 1978;41:713-718.
Downloaded From: http://archneur.jamanetwork.com/ by a Karolinska Institutet University Library User on 05/24/2015
