Epilepsia, 46(4):588–589, 2005

Blackwell Publishing, Inc.

Original Letters

Hypersomnia in an Epilepsy Patient Treated with Levetiracetam

Ramin Khatami, Adrian M. Siegel, and Claudio L. Bassetti

Neurology Department, University Hospital, Zurich, Switzerland

Summary: We report a patient with focal epilepsy in whom of subjective excessive daytime sleepiness (EDS) during an add-
increased sleep needs (hypersomnia) developed in the absence on treatment with levetiracetam (LEV).

A 36-year old woman with simple partial and secon- min to non–rapid eye movement (NREM) sleep stage 2],
darily generalized seizures due to a fibrillary astrocytoma large amounts of slow-wave sleep (32% of total sleep
(WHO II) in the right limbic area (inferior gyrus, insu- time), and a sleep efficiency of 82%. The arousal in-
lar cortex, gyri orbitales, temporal white matter, tempo- dex (<5/h), apnea–hypopnea index (3/h), and the peri-
ral pole) became seizure free after surgery for 1.5 years odic limb-movement index (<5/h) were normal. The mean
with sodium valproate (VPA). Seizures reappeared, up to sleep latency on multiple sleep latency test (MSLT) the
six focal seizures a day due to a slight tumor progres- next day was 2 min in the absence of sleep-onset REM pe-
sion, as documented in a follow-up magnetic resonance riods. No seizures and no EEG epileptic discharges were
imaging (MRI). Seizures became refractory to monother- recorded during a 24-h EEG monitoring, and seizure fre-
apy with VPA, 1,200 mg/day (serum level of 761 µM) quency remained stable over the following months. In De-
and to monotherapy with carbamazepine (CBZ), 1,350 cember 2001, an enhanced [18 F]fluorodeoxyglucose up-
mg/day (serum level of 53 µM). When LEV, 1,000 mg/day, take on positron emission tomography (FDG-PET) was
was added to CBZ in December 2000, seizure frequency consistent with mild tumor progression. Under chemother-
(two seizures/month) and severity (seizures confined to apy with temozolomide (one cycle at a dose of 75 mg/m2
the first 2 h of sleep without generalization) were ini- daily for 5 days, followed by continuous low-dose ther-
tially improved. Two month later when dosage of LEV apy), tumor extension remained stable while the tracer
was increased to 2,000 mg/day, the patient reported a dis- uptake on FDG-PET decreased. In March 2002, treatment
abling daytime fatigue/somnolence with an increase in with lamotrigine (LTG) was started and increased to 150
sleep needs (12 h of sleep/day). Her Epworth sleepiness mg/day within the following 4 months. EDS and hyper-
score (ESS) was, however, within normal limits (score at somnia were unchanged until LEV was tapered off in June
7/24). A repeated MRI (March 2001) revealed no tumor 2002. The patient’s sleep needs have normalized (7.5–8.5
progression, and her neurologic examination remained h of sleep/day), and ESS was 4/24, despite unchanged
normal. MRI findings (last control, January 2004). During a sec-
The CBZ serum levels (40–50 µM) were similar to ond 2-week actigraphy, the patient was “asleep” 40% of
those obtained when the patient did not complain of the recording time. On a second MSLT, the mean sleep
EDS/hypersomnia. During a 2-week actigraphy (motion- latency had normalized (13 min). The patient still has one
logger device including light sensor data; Ambulatory to two seizures per month with treatment with CBZ, 600
Monitoring Inc., Ardsley, NY, U.S.A.), the patient was mg/day, and LTG, 150 mg/day, at her visit in December
“asleep” 46% of the recording time. An overnight 2003.
polysomnography documented a short sleep latency [6 The observation of subjective and objective increase in
sleep needs but normal ESS after add-on treatment with
Accepted November 29, 2004. LEV is of potential clinical relevance for three reasons.
Address correspondance and reprint requests to Dr. C. Bassetti at
Neurologische Poliklinik Universitätsspital, Frauenklinikstrasse 26, CH- First, LEV has been reported to cause severe “somno-
8004 Zürich, Switzerland. E-mail: [email protected] lence, tiredness, and asthenia” in ∼4–15% of the patients

588
 LEVETIRACETAM AND HYPERSOMNIA 589

in clinical mono- and polytherapy trials (1–3) and led to the absence of subjective sleepiness implies that LEV
drug discontinuation in ≤10% of patients (2,4). However, may affect specific brain structures. Together with the
objective data on the character and/or severity of “somno- observed increase in slow-wave sleep, a disinhibition of
lence” and “drowsiness” associated with LEV are missing NREM sleep–generating structures could be suggested.
in the literature. We show, for the first time to our best Similar effects have been described in epileptics with
knowledge, that these symptoms may be the expression gabapentin (7) and in healthy subjects with brief expo-
of an objective and severe increase in sleep propensity, as sure to CBZ (8). By contrast, a decrease of both subjec-
indicated by our actigraphic and MSLT results. tive daytime alertness and slow-wave sleep has been re-
Second, the identification of the etiology of EDS in ported after a single dose of 1,000 mg LEV in patients
patients treated for epilepsy can be difficult, because with focal epilepsy receiving stable CBZ therapy (9). Our
several potential factors such as primary sleep disorders, report documents the existence of a severe but potentially
poor seizure control, and side effects of antiepileptic drugs reversible hypersomnia with LEV as add-on treatment
(AEDs) are involved (5,6). This point is well exemplified in some epilepsy patients, and stresses the importance
by our observation. We assume a causative link between of considering the possibility of objective sleep–wake
EDS/hypersomnia and LEV on the basis of the follow- tests in epilepsy patients complaining of daytime/fatigue/
ing four arguments: (a) the temporal coincidence between somnolence.
appearance of EDS/hypersomnia and start of LEV treat-
ment, and between resolution of EDS/hypersomnia and
LEV discontinuation; (b) stable seizure control and ab- REFERENCES
sence of epileptic discharges on routine and long-term
EEG recording; (c) no detectable MRI tumor progression 1. Alsaadi TM, Thieman C. Levetiracetam monotherapy for newly di-
agnosed epilepsy patients. Seizure 2003;12:154–6.
when EDS/hypersomnia started and, conversely, resolu- 2. Ben-Menachem E, Gilland E. Efficacy and tolerability of levetirac-
tion of EDS/hypersomnia sleepiness despite unchanged etam during 1-year follow-up in patients with refractory epilepsy.
MRI findings; and (d) exclusion of primary sleep disor- Seizure 2003;12:131–5.
3. Harden C. Safety profile of levetiracetam. Epilepsia 2001;42(suppl
ders as a cause of EDS by polysomnography and acti- 4):36–9.
graphy. Although we considered various potential EDS- 4. Bird JM, Joseph ZA. Levetiracetam in clinical use: a prospective
related factors in our patient, we cannot rule out that an observational study. Seizure 2003;12:613–6.
5. Bazil CW. Sleep and epilepsy. Curr Opin Neurol 2000;13:171–5.
interaction between LEV, CBZ, epilepsy, and brain tumor 6. Malow BA, Bowes RJ, Lin X. Predictors of sleepiness in epilepsy
contributed to the reported symptoms. Thus LEV may in- patients. Sleep 1997;20:1105–10.
duce hypersomnia only when acting in combination with 7. Legros B, Bazil CW. Effects of antiepileptic drugs on sleep architec-
ture: a pilot study. Sleep Med 2003;4:51–5.
other factors. Among them, exogenous (e.g., other AEDs) 8. Yang JD, Elphick M, Sharpley AL, et al. Effects of carbamazepine
and endogenous (e.g., genetic predisposition) factors may on sleep in healthy volunteers. Biol Psychiatry 1989;26:324–8.
be relevant. 9. Bell C, Vanderlinden H, Hiersemenzel R, et al. The effects of leve-
tiracetam on objective and subjective sleep parameters in healthy vol-
Third, the mechanisms by which LEV may cause hy- unteers and patients with partial epilepsy. J Sleep Res 2002;11:255–
persomnia remain unclear. An increase of sleep needs in 63.

Epilepsia, Vol. 46, No. 4, 2005