The Effects of Lamotrigine on Sleep in Patients with Epilepsy

Epilepsia, 42(12):1569 1573, 2001 Blackwell Science, Inc. International League Against Epilepsy The Effects of Lamotrigine on Sleep in Patients with Epilepsy *Nancy Foldvary, *Michael Perry, Julia Lee, *Dudley Dinner, and *Harold H. Morris Departments of *Neurology and Epidemiology and Biostatistics, The Cleveland Clinic Foundation, Cleveland, Ohio, U.S.A. Summary: Purpose: The older antiepileptic drugs (AEDs) have a variety of effects on sleep, including marked reduction in REM, slow-wave sleep (SWS) and sleep latency, and increased percentage of light sleep. The effects of the newer AEDs on sleep are unknown. Our purpose was to study the effect of lamotrigine (LTG) on sleep. Methods: Ten adults with focal epilepsy, in whom the decision was made to add LTG to either phenytoin (PHT) or carbamazepine (CBZ) for control of seizures, were the subjects of this study. Patients underwent pre- and posttreatment polysomnography (PSG) and completed sleep questionnaires. Polygraphic variables and Epworth Sleepiness Scale (ESS) scores, a subjective measure of sleep propensity, were compared by using the Wilcoxon sign rank test. Results: Seven patients were taking CBZ, and three were treated with PHT. All subjects were titrated to an LTG dose of 400 mg/day. Treatment with LTG produced a significant decrease in SWS and an increase in stage 2 sleep percentage. No significant difference in ESS or any of the other polygraphic variables was observed. However, LTG treatment was associated with a reduction in arousals and stage shifts and an increase in REM periods. No subjects reported insomnia with treatment. Conclusions: LTG appears to be less disruptive to sleep than some of the older AEDs. Key Words: Lamotrigine Sleep Sleep architecture Insomnia Polysomnography. Individuals with epilepsy commonly report excessive daytime sleepiness (EDS) and fatigue that are typically attributed to the direct effects of seizures or antiepileptic drugs (AEDs). In patients with epilepsy, sleep organization is disrupted by frequent arousals, awakenings, and stage shifts, even in the absence of nocturnal seizures and AEDs (1). Daytime and nocturnal seizures fragment sleep, reducing the percentage of rapid-eye-movement (REM) and slow-wave sleep (SWS; 1 3). The older AEDs have a variety of effects on sleep (4 10). Relatively little is known of the effects of the newer AEDs on sleep. The purpose of this study was to investigate the effect of lamotrigine (LTG) on sleep perception, sleep architecture, and subjective daytime alertness. METHODS Patient selection Ten patients between the ages of 18 and 45 years with focal epilepsy in whom the decision was made to add LTG to either phenytoin (PHT) or carbamazepine (CBZ) for control of seizures were the subjects of this study. In all cases, epilepsy was documented by clinical history, Revision accepted Address correspondence and reprint requests to Dr. N. Foldvary at The Cleveland Clinic Foundation, 9500 Euclid Ave, S-51, Cleveland, OH 44195, U.S.A. E-mail: foldvan.ccf.org EEG, and neuroimaging abnormalities. Exclusion criteria included (a) use of barbiturates, benzodiazepines, ethanol, antihistamines, recreational drugs, and prescription and nonprescription sleep aids within 30 days of enrollment; (b) more than four partial or secondarily generalized seizures per month during the 3 months before enrollment; (c) inability to quantify seizures; (d) evidence of a sleep disorder based on clinical history or polysomnography (PSG), excluding primary snoring; and (e) documented psychiatric diagnosis and/or Beck Depression Inventory (BDI) score of 15 (11). Baseline phase After giving informed consent, subjects were interviewed and examined by a board-certified sleep medicine physician (N.F.). Subjects were instructed on the use of sleep logs and seizure calendars, which were maintained for the duration of the study. Trough free and total AED levels were obtained. Subjects completed three questionnaires including a 17-item sleep questionnaire designed to ascertain sleep patterns and sleep disorder symptoms, the Epworth Sleepiness Scale (ESS), and the BDI. The sleep questionnaire included the following items: average number of hours of sleep per night, sleep latency, number of awakenings and their cause(s), frequency of naps, daily caffeine ingestion, the presence of snoring, witnessed apnea, difficulty initiating or maintaining sleep, restless- 1569

1570 N. FOLDVARY ET AL. ness, and symptoms of restless leg syndrome (RLS; uncomfortable sensations in the legs in bed or in the evening relieved by movement). The ESS is a widelyused eight-item survey designed to ascertain sleep propensity during activities of daily living, and is a validated measure of subjective daytime sleepiness (12). Subjects rate the chance of dozing in each of eight activities of daily living from 0 (never) to 3 (high). The scores for the eight activities are tallied, producing a total score ranging from 0 to 24, with 24 indicating severe daytime sleepiness. The BDI was administered to screen for depression, a common cause of sleep disruption (11). Two consecutive nights of ambulatory PSG using Digitrace SleepScan were obtained: the first night for adaptation and the second for data collection. This system incorporates four EEG channels (C3, C4, O1, O2), two channels of electrooculogram (EOG; right and left outer canthus), chin and anterior tibialis electromyogram (EMG), electrocardiogram (ECG), airflow, respiratory effort (thoracic and abdominal), oxygen saturation, body position, and snoring. After each PSG, subjects were asked to rate their sleep as the same, worse, or better compared with a typical night. Subjects were entirely seizure free (including auras) for 48 h before and during each PSG. Titration phase After completion of baseline, subjects were treated with LTG as follows: 50 mg daily in weeks 1 and 2; 50 mg twice daily in weeks 3 and 4; 100 mg twice daily in weeks 5 and 6; and 200 mg twice daily in weeks 7 through 10. Dose reduction of concomitant AEDs was permitted in subjects experiencing adverse effects. Posttreatment phase Subjects underwent a single PSG during week 10 of treatment. We elected not to perform two consecutive studies because a significant first-night effect has not been found when recording in the home by other investigators (13,14). This was supported by preliminary findings from this study demonstrating similar sleep-stage percentages on baseline nights 1 and 2, which will be published elsewhere. Subjects completed the ESS and a posttreatment questionnaire designed to ascertain subjective changes in sleep quality or daytime alertness during the treatment period. Serum concentrations of AEDs were obtained the morning after the PSG. Analysis PSGs were scored by a board-certified R.PSGT. (M.P.) and interpreted by the first author with standard scoring procedures (15). Pre- and posttreatment polygraphic variables were compared by using the Wilcoxon sign rank test. A p value of <0.05 was considered to be statistically significant. PSG variables included total sleep time (TST; time occupied by stages 1 4, and REM, in minutes), sleep latency (SL; time from lights out to sleep onset, defined as the first of three consecutive epochs of stage 1 sleep or one epoch of any other stage, in minutes), sleep efficiency (SE; TST divided by time in bed, expressed as a percentage), REM latency (time from sleep onset to the first epoch of REM sleep, in minutes), percentage TST spent in non-rem stages 1, 2, and SWS (stages 3 and 4 combined) and REM sleep, number of stage shifts, number of REM periods, arousal index (number of arousals per hour of sleep), apnea hypopnea index, and periodic limb movement index. Mean sleep time per night was determined at baseline and after treatment by averaging the number of hours of sleep per night recorded in sleep logs for the week before baseline and posttreatment PSGs. Subjective reports of alterations in sleep quality after treatment were noted. RESULTS The mean age of the cohort was 33.9 years (21 54). Seven of the 10 subjects were women. Seven subjects were taking CBZ, and three were taking PHT. For nine patients with complex partial and secondarily generalized seizures, the baseline seizure frequency was 2.4 (0.15 4) per month. Subject 10 had an average of two auras per month. Overall, the percentage seizure reduction after treatment was 30%. Six patients had no appreciable change in seizure frequency, including the patient with auras only. The other four patients experienced a mean seizure reduction of 75% (50 100%). All subjects completed the study and achieved the LTG target dose of 400 mg. Two patients developed dizziness during the LTG titration, necessitating a reduction of CBZ by 200 mg per day. Mean pre- and posttreatment serum AED concentrations were 10.4 and 10.2 g/ml, respectively, for CBZ; and 14.6 and 15.0 g/ml, respectively, for PHT. Mean posttreatment LTG concentration was 2.7 g/ml. Mean baseline and posttreatment ESS scores were 7.7 and 7.1, respectively (p 0.56). Mean BDI was 4 (0 7). Mean sleep time per night over 1 week at baseline and after treatment was 8.3 and 8.1 h, respectively (p 0.25). Polygraphic variables are shown in Table 1. A significant increase in stage 2 and reduction in SWS after treatment were found. Although not reaching statistical significance, a decrease in the arousal index and stage shifts and an increase in REM periods were observed. At the completion of the treatment phase, six subjects reported no change in sleep quality or daytime alertness, whereas two subjects described a reduction in sleep latency and/or nocturnal awakenings compared with baseline. More frequent daytime napping and an increase in dreaming were reported by one subject each. Treatment with LTG did not produce an increase in snoring, witnessed apnea, restlessness, or symptoms of RLS.

EFFECT OF LAMOTRIGINE ON SLEEP 1571 TABLE 1. Polysomnographic variables Baseline Treatment % change/100 a p value Total recording time (min) 435.0 445.0 0.04 0.70 Total sleep time (min) 388.0 382.2 0.02 0.92 Sleep efficiency (%) 88.6 86.6 0.02 0.77 Sleep latency (min) 14.6 25.5 1.02 0.20 REM latency (min) 62.3 54.7 0.09 0.22 % stage 1 7.0 5.8 0.10 0.34 % stage 2 50.9 57.5 0.14 0.03 % stage 3 4 20.2 13.0 0.31 0.03 % REM 22.4 23.8 0.18 0.77 REM periods 3.5 3.9 0.15 0.36 Stage shifts 64.9 53.6 0.04 0.32 Arousal index 6.3 5.5 0.48 >0.99 Apnea-hypopnea index 0.3 0.6 2.47 0.44 PLM index 0.6 0.0 1.0 0.50 REM, rapid eye movement; PLM, periodic limb movement. a % change/100 posttreatment baseline/baseline. DISCUSSION Sleep is reportedly disrupted in patients with epilepsy, even in the absence of seizures and AEDs. Nocturnal seizures produce an increase in wake time, a reduction in REM and SWS, prolonged REM latency, and an increase in sleep fragmentation (1,2). Daytime complex partial seizures prolong REM latency and decrease REM percentage on nights after the seizure (3). To control for the effects of seizures on sleep, we excluded subjects who were unable to quantify seizures and those with more than four partial or secondarily generalized seizures per month, and performed all PSGs after a minimal seizurefree interval of 48 h. The interpretation of prior studies addressing the effects of AEDs on sleep is difficult because of methodologic variations, including wide differences in patient population, duration of treatment and drug dosages, and failure to control consistently for seizures and first-night effect. Phenobarbital (PB) has been shown to produce a shortened sleep latency, an increase in efficiency, a reduction of REM sleep, and an increase in stage 2 (4). The effects of PHT and CBZ appear to vary with treatment duration (5,6). Short-term PHT therapy (time to steady state) produced a reduction of sleep latency and stage 1 and increases in SWS and arousals, which reversed after several months of treatment (4). Similarly, a single dose of controlled-release CBZ produced a reduction in REM and increase in REM fragmentation that was no longer observed after 1 month of treatment (6). Patients with temporal lobe epilepsy treated for a prolonged period with CBZ were found to have reduced REM percentage compared with unmedicated patients, when controlling for seizures during video-eeg monitoring (7). Yet another study found no significant long-term effects on sleep architecture in patients taking CBZ (8). The effects of valproate (VPA) therapy on sleep ranged from a significant reduction in REM and increase in SWS to none (9,10). A variety of polysomnographic variables (PSG) can be used to ascertain sleep fragmentation. Sleep disruption is common on the first night in a series of overnight PSGs performed in the sleep laboratory, a phenomenon known as the first-night effect (13,14). These changes include reduced efficiency of sleep, increased light sleep (stage 1), prolonged latency to sleep, prolonged REM latency, increased REM periods, stage shifts, arousals and awakenings, and reduced SWS and REM sleep percentage. The first-night effect is either absent or significantly reduced when PSG is performed in the home (13). For this reason, in addition to the reduced cost and improved patient access, we used ambulatory PSG instead of attended laboratory studies, the current gold standard used for the diagnosis of sleep disorders. Although the number of channels is similar in the ambulatory and standard laboratory PSGs, a major limitation of home recordings is the risk of data loss because of artifacts or electrode/ sensor dysfunction. However, in the experience of our laboratory and others, data loss due to equipment failure has not occurred (16). Our study demonstrates that sleep may be less disrupted in patients treated with LTG as compared with the older AEDs. The only significant effects of treatment included an increase in stage 2 (light sleep) and a decrease in SWS (deep sleep). The significance of these changes is unclear. The hypothesis that SWS or REM sleep serves a specific restorative role has not been firmly proven in human subjects (17,18). Conversely, studies have shown a correlation between arousals during sleep and daytime alertness in normal subjects and patients with sleep disorders (19,20). The upper airway resistance syndrome (UARS) is one such example. In subjects with UARS, sleep is fragmented by recurrent, brief arousals due to abnormal increases in respiratory effort as measured by esophageal pressure monitoring (20). Excessive daytime sleepiness occurs in the absence of frank sleep apnea. Although it did not reach statistical

1572 N. FOLDVARY ET AL. significance, treatment with LTG was associated with a slight reduction in arousals and stage shifts and an increase in the number of REM periods without affecting sleep efficiency, suggesting a tendency for sleep to be less disrupted. These findings are in contrast to many previous studies in which the older AEDs appeared to produce a shift toward more stage 1 sleep, reduced sleep efficiency, prolonged sleep latency, and an increase in arousals. The effect of LTG on sleep has been addressed in two previous reports. Placidi et al. (21) performed PSGs in 13 subjects with drug-resistant epilepsy (11 partial, two Lennox Gastaut syndrome) treated with LTG (21). All but one subject were taking two to four concomitant AEDs, including benzodiazepines, barbiturates, and vigabatrin (VGB), a drug that has been shown to prolong REM latency in a small cohort (22). Compared with baseline, adjunctive therapy with LTG, 300 mg daily, produced a significant increase in REM sleep (8.5 vs. 13.3%), a reduction in the number of entries into REM sleep, and a significant decrease in the number of stage shifts and SWS percentage. Seizure reduction was observed in 84.6% of patients, and 50% were completely free of generalized tonic clonic seizures. In another series, 11 patients with temporal lobe epilepsy taking LTG (six in monotherapy) had REM sleep percentages similar to those of unmedicated patients when studied in the epilepsy monitoring unit (23). Several methodologic differences may explain the varying effects of LTG on REM sleep between ours and the study of Placidi et al. The majority of subjects in the prior study were taking multiple AEDs including drugs known to suppress REM. This is the most likely explanation for the low baseline REM percentage. Although baseline seizure frequency was not reported, presumably these patients had more severe epilepsy. Subjects in the current study had normal baseline REM percentages and no more than four seizures per month with a single AED not known to produce significant REM suppression. Differences in serum LTG concentrations, not provided in the prior study, also may have affected these results. Finally, subjects in the prior study experienced a more significant seizure reduction with treatment, which would be expected to have the effect of enhancing REM sleep. Although our series is small, no patients reported insomnia, and only one subject reported more frequent naps during treatment. Seizure frequency was unchanged in six of 10 subjects, resulting in an overall mean monthly reduction of only 30%. Therefore, we do not believe our findings are due to an improvement in seizure frequency, although this may be a contributing factor. It must be noted that the posttreatment mean LTG level of 2.7 g/ml is low, probably because of the presence of enzyme-inducing AEDs. Whether higher LTG concentrations would produce similar results is unknown. In placebo-controlled add-on trials, sleep disorders were observed in 1.4% of subjects treated with LTG versus 0.5% of those treated with placebo (24). Somnolence and insomnia were more common in subjects treated with LTG (14.2 and 6%, respectively) than controls (6.9 and 2.1%, respectively) (24). The average daily LTG dose ranged from 150 to 500 mg. In a recent series of 109 subjects treated with LTG, insomnia of sufficient severity to require discontinuation or dose reduction was experienced by seven (6.4%) subjects (25). Difficulty initiating and maintaining sleep developed shortly after LTG was introduced, increased with dose escalation, and resolved quickly with discontinuation or dose reduction. Symptoms developed at a mean daily dose of 286 mg (100 500). Based on these data, we believe that LTG may be less disruptive to sleep than the older AEDs. Because sleep fragmentation reduces seizure threshold in some individuals, these changes may contribute to the anticonvulsant effects of the drug. Acknowledgment: This study was supported by Glaxo Wellcome. We thank Digitrace for the use of their ambulatory PSG recorder. REFERENCES 1. Touchon J, Baldy-Moulinier M, Billiard M, et al. Sleep organization and epilepsy. In: Degen R, Rodin EA, eds. Epilepsy, sleep and sleep deprivation. 2nd ed. Amsterdam: Elsevier Science, 1991:73 81. 2. Castro LHM, Bazil CW, Walczak TS. Nocturnal seizures disrupt sleep architecture and decrease sleep efficiency. Epilepsia 1997; 38:49. 3. Bazil CW, Castro LHM, Walczak TS. Daytime seizures increase REM latency and decrease total REM. Epilepsia 1997;38:176. 4. Wolf P, Röder-Wanner UU, Brede M. Influence of therapeutic phenobarbital and phenytoin medication on the polygraphic sleep of patients with epilepsy. Epilepsia 1984;25:467 75. 5. Röder-Wanner UU, Noachtar S, Wolf P. Response of polygraphic sleep to phenytoin treatment for epilepsy: a longitudinal study of immediate, short, and long-term effects. Acta Neurol Scand 1987; 76:157 67. 6. Gigli GL, Placidi F, Diomedi M, et al. Nocturnal sleep and daytime somnolence in untreated patients with temporal lobe epilepsy: changes after treatment with controlled release carbamazepine. Epilepsia 1997;38:696 701. 7. Bazil CW, Walczak TS. Anticonvulsants and seizures independently reduce REM sleep in patients with temporal lobe epilepsy. Epilepsia 1998;39:69. 8. Declerck AC, Wauquier A. Influence of antiepileptic drugs on sleep patterns. In: Degen R, Rodin EA, eds. Epilepsy, sleep and sleep deprivation. 2nd ed. Amsterdam: Elsevier Science, 1991:153 63. 9. Röder UU, Wolf P. Effects of treatment with dipropylacetate and ethosuximide on sleep organization in epileptic patients. In: Dam M, Gram L, Penry JK, eds. Advances in epileptology: XIIth Epilepsy International Symposium. New York: Raven Press, 1981:145 57. 10. Harding GFA, Alford CA, Powell TE. The effect of sodium valproate on sleep, reaction times, and visual evoked potential in normal subjects. Epilepsia 1985;26:597 601.

EFFECT OF LAMOTRIGINE ON SLEEP 1573 11. Beck AT. Beck Depression Inventory. San Antonio, TX: The Psychological Corporation, 1987. 12. Johns MW. Reliability and factor analysis of the Epworth Sleepiness Scale. Sleep 1992;15:376 81. 13. Edinger JD, Fins AI, Sullivan RJ. Sleep in the laboratory and sleep at home: comparisons of older insomniacs and normal sleepers. Sleep 1997;20:1119 26. 14. Sharpley AL, Solomon A, Cowen PJ. Evaluation of first night effect using ambulatory monitoring and automatic sleep stage analysis. Sleep 1988;11:273 6. 15. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Washington, DC: U.S. Government Printing Office, 1968. 16. Fry JM, DiPhillipo MA, Curran K, et al. Full polysomnography in the home. Sleep 1998;21:635 42. 17. Johnson LC, Naitoh P, Moses JM, et al. Interaction of REM deprivation and stage 4 deprivation with total sleep loss: experiment 2. Psychophysiology 1974;11:147 59. 18. Lubin A, Moses JM, Johnson LC, et al. The recuperative effects of REM sleep and stage 4 sleep on human performance after complete sleep loss: experiment I. Psychophysiology 1974;11:133 46. 19. Bonnet MH. Performance and sleepiness as a function of frequency and placement of sleep disruption. Psychophysiology 1986; 23:263 71. 20. Guilleminault C, Stoohs R, Clerk A, et al. A cause of excessive daytime sleepiness: the upper airway resistance syndrome. Chest 1993;104:781 7. 21. Placidi F, Diomedi M, Scalise A, et al. Effect of anticonvulsants on nocturnal sleep in epilepsy. Neurology 2000;54(suppl 1):S25 32. 22. Siegel HE, Hunter K, Vega-Bermudez F, et al. The effects of vigabatrin on the REM sleep of patients with complex partial seizures. Epilepsia 1998;39:199. 23. Lamotrigine: data on file. Glaxo Wellcome, 1995. 24. Sadler M. Lamotrigine associated with insomnia. Epilepsia 1999; 40:322 5.