Short communication
Department of Medical Genetics, University Medical Centre Utrecht, PO Box 85090, 3508 AB Utrecht, The Netherlands
b Department of Obstetrics and Perinatology, University Medical Centre Utrecht, The Netherlands
c Department of Neurology and Epileptology, Epilepsy Institute of The Netherlands SEIN, Heemstede, The Netherlands
Received 26 October 2004; received in revised form 23 November 2004; accepted 4 December 2004
Available online 8 January 2005
Abstract
Prenatal exposure to levetiracetam (LEV) has been shown to cause skeletal abnormalities and growth retardation in animal studies, but the
teratogenicity of this new antiepileptic drug in humans is still unknown. We detected no malformations in a series of 11 pregnancies with
LEV exposure, although it was striking that three cases had a low birth weight. There may be an association between maternal LEV use and
reduced birth weight, but too few cases have been monitored so far. We recommend that the outcomes of all pregnancies exposed to LEV
should be carefully registered.
2004 Elsevier Inc. All rights reserved.
Keywords: Levetiracetam; Antiepileptic drugs; Epilepsy; Pregnancy; Malformation; Birth defect; Birth weight; Fetal exposure
1. Introduction
Maternal antiepileptic drug use is associated with an increased frequency of adverse pregnancy outcome including congenital malformations [1]. However, continuation
of medication during pregnancy is often necessary to prevent seizures which would be harmful to mother and fetus.
Recently, some new antiepileptic drugs (AEDs) have been
launched, promising better antiepileptic properties and fewer
side effects; however, the teratogenicity of this new generation of AEDs in humans is still unknown. The same holds
true for levetiracetam (LEV), which has been registered for
treatment of partial epilepsy and seems to be effective in idiopathic generalized epilepsy as well [2]. It is used mostly as
adjunctive therapy in doses of 10003000 mg/day. The drug
is chemically unrelated to existing AEDs and precisely how
it prevents seizures is unknown.
In animal studies, LEV showed developmental toxicity at doses similar to or greater than human therapeutic
doses [35]. In rats and rabbits, treatment with 350 and
0890-6238/$ see front matter 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.reprotox.2004.12.001
600 mg/kg/day, respectively, was associated with an increase in embryo-fetal mortality, minor skeletal abnormalities and growth retardation [35]. In a mouse model of
teratogenicity, no gross external malformations were observed in offspring prenatally exposed to LEV and its major
metabolite in humans, 2-pyrrolidinine N-butyric acid (PBA),
at doses up to 2000 mg/kg/day, except for cleft palate in
two fetuses exposed to LEV 1200 mg/kg/day [3]. In both
the LEV and PBA groups, at doses 600 mg/kg/day fetal
weights were reduced and the overall frequency of skeletal abnormalities, specifically of hypoplastic phalanges, was
increased. These skeletal abnormalities were not actual malformations but they are well-recognized indicators of perturbations in fetal growth and development; therefore, they
point towards LEV affecting prenatal growth. Furthermore,
LEV treatment at 600 mg/kg/day caused a 15% reduction
in maternal weight gain associated with the delayed fetal
growth.
Few data are available on the teratogenicity of LEV in
humans. Long [6] reported a normal pregnancy outcome
for three cases treated with monotherapy of 7503000 mg
LEV daily. Birth weights were within the normal range.
During safety trials, 23 women became pregnant while tak-
176
3. Discussion
In a series of 11 pregnancies with LEV exposure, we observed one spontaneous abortion and nine live births, whereas
one pregnancy was terminated because of maternal social
reasons. Among the nine live births no malformations were
found. However, two out of nine live-born children had a
Table 1
Maternal data and pregnancy outcome in 11 pregnancies with LEV treatment
Case no.
1
6
7
8
9-I
9-II
10
Seizure type
Generalized tonic-clonic
seizures
Cigarette smoking
(number/day)
Age (years)
Parity
Pregnancy duration
(weeks)
Sex
Abnormal outcome
Levetiracetam
1500a ,
carbamazepine 600,
valproate 2000
Levetiracetam 2000,
oxcarbazepine
2400b , lamotrigine
100c
Levetiracetam 2000,
carbamazepine 2400
Levetiracetam
1000d ,
carbamazepine 400e
Levetiracetam 3500,
carbamazepine 800,
clobazam 20
Levetiracetam 500f ,
valproate 1200g
Levetiracetam 1500
Levetiracetam 1500,
oxcarbazepine 1500,
clobazam 10
Levetiracetam 2000,
carbamazepine 400
Levetiracetam 1500,
carbamazepam 400
Levetiracetam 1000
Partial
22
G1P0
n.a
n.a.
Partial
Monthly
26
G1P0
38
2390 (P5)
Partial
28
G1P0
40
3300 (P50)
Partial
<1/Month
10
22
G2P0
39
3495 (P65)
Partial
Weekly
15
21
G1P0
n.a.
n.a.
Generalized
21
G1P0
39
3500 (P60)
Voluntary abortion
for maternal social
reasons at 8 weeks
Partial
Myoclonic
0
<1/Month
3h
0
33
23
G2P0
G1P0
41
32
F
M
3200 (P25)
930 (P2)
Partial
30
G1P0
39
3680 (P80)
Partial
32
G2P1
41
3825 (P60)
Myoclonic
10
40
G2P1
40
2800 (P5)
Medication during
pregnancy (mg/day)
177
178
low birth weight (<10th centile) and one a very low birth
weight (<2.3rd centile). The daily dose of LEV to which
the pregnancies had been exposed was within the generally
recommended therapeutic dose of 10003000 mg/day with
only two exceptions: 3500 mg/day in case 5 (pregnancy terminated) and 500 mg/day in case 6 (normal outcome). Other
factors than LEV treatment may, however, have been responsible for the reduced birth weights.
In case 2 pre-eclampsia was present, a well-known risk
factor for growth retardation. In case 8 abnormal placental
blood flow was found, suggesting that placental insufficiency
was the major cause of the severe prenatal growth retardation. However, no explanation for this insufficiency could be
found. It is not known whether LEV exposure influences prenatal growth by changing placental blood flow or by inducing
pre-eclampsia.
LEV was prescribed in combination with oxcarbazepine
and lamotrigine in case 2, and with oxcarbazepine and
clobazam in case 8. So far, prenatal exposure to these drugs
has not been associated with reduced birth weight in humans.
Co-medication with oxcarbazepine reduces LEV serum levels compared to LEV monotherapy, whereas lamotrigine has
no significant effect on LEV serum levels [8]. Until now, no
other important pharmacokinetic interactions between LEV
and other AEDs have been observed, so that an effect on birth
weight due to such interactions with co-prescribed drugs is
unlikely whereas pharmacodynamic interactions cannot be
excluded.
The reduced birth weight in case 10 cannot be explained by nicotine exposure [9], since the same mother
gave birth also to healthy twins with normal birth weights
while she smoked the same amount of cigarettes per day. In
that twin pregnancy she had used phenobarbital instead of
LEV.
In animal studies, delayed fetal growth was associated
with reduced maternal weight gain during LEV treatment [3].
In our cases, no abnormal maternal weight changes during
or prior to pregnancy were observed, although weight loss
has been associated with LEV treatment in both adults and
children [1012]. High concentrations of LEV have also been
reported in human breast milk but it is not known whether
this affects postnatal growth [13].
Thus far, experience with LEV treatment in pregnancy
does not point towards a distinct malformation pattern in humans; however, too few cases have been published to draw
conclusions about safety with respect to congenital malformations. Small case series, like this one, may nonetheless be
helpful in signaling adverse effects on continuous pregnancy
outcome parameters like birth weight.
Although animal studies have shown an association between prenatal LEV exposure and growth retardation, no
clear effect on weight in human pregnancies can be established. However, as low birth weight is associated with increased morbidity and mortality [14], our cases suggest we
should be aware of a possible association in humans. Largescale pregnancy registries are needed to find out whether pre-