Valproate and the risk for congenital malformations: Is formulation and dosage
E. Mawhinneya, J. Campbella, J. Craiga,*, A. Russellb, W. Smithsonc, L. Parsonsd, I. Robertsone,
B. Irwina, P. Morrisonf, B. Liggang, N. Delantyh, S. Hunta, J. Morrowa
aDepartment of Neurology, Belfast Health and Social Care Trust, Grosvenor Road, Belfast BT12 6BA, United Kingdom
bDept of Clinical Neurophysiology, Institute of Neurological sciences, Southern General Hospital, 1345 Govan Road, Glasgow, G51 4TF, United Kingdom
cAcademic Unit of Primary Medical Care, Samuel Fox House, University of Sheffield, Northern General Hospital, Herries Road, Sheffield, S5 7AU, United Kingdom
dNeurology Department, Luton & Dunstable Hospitals NHS Trust, Lewsley Road, Luton, LU4 0DZ, United Kingdom
eObstetric & Gynaecology Department, Sharoe Green Unit, Royal Preston Hospital, Presto, United Kingdom
fDept of Medical Genetics, Belfast City Hospital, Lisburn Road, Belfast, BT9 7AB, United Kingdom
gRCSI, Beaumont Hospital, Dublin 9, Ireland
hDept of Neurology, Beaumont Hospital, Dublin 9, Ireland
Prenatal exposure to antiepileptic drugs (AEDs) increases the
risk of major congenital malformations (MCMs) from the
background risk of 1–2% to between 4% and 9%.1–4Previous
studies, including those from the UK Epilepsy and Pregnancy
Register, have shown that use of valproate, both as monotherapy
and as part of a polytherapy regimen, is associated with a higher
risk of MCMs than for other AEDs such as carbamazepine and
lamotrigine. Several studies have also shown a trend towards a
positive dose response for MCMs and valproate, particularly for
doses over 1000 mg daily.4–8
Animal studies have suggested that the increased rate of MCMs
with valproate may be related to peak plasma concentration, as well
as the total daily dose.9Using a controlled release formulation of
valproate results in reduced peak plasma concentration and smaller
diurnal fluctuations than standard valproate.10Therefore, using a
three or four times daily regimen, or use of a controlled release
formulation of valproate, to minimise high plasma concentrations
has been proposed as possibly being of benefit and has been included
in a number of guidelines.11–14There have been no published clinical
studies investigating the validity of this approach. If confirmed this
could have a significant impact on the treatment options available to
women of childbearing age with epilepsy, in particular those with
Seizure 21 (2012) 215–218
A R T I C L E
I N F O
Received 21 September 2011
Received in revised form 16 January 2012
Accepted 18 January 2012
Major congenital malformations
A B S T R A C T
Background: Use of valproate in pregnancy, especially in doses over 1000 mg a day, is known to be
associated with a higher risk for major congenital malformations compared with other antiepileptic
drugs. We sought to investigate whether the increased risk could be minimised by using controlled
release or divided daily doses of valproate.
Methods: The UK Epilepsy and Pregnancy Register is a prospective, observational and follow up study set
up to determine the risks of major congenital malformations for infants exposed to antiepileptic drugs in
utero. In this study we have extracted data for those pregnancies exposed to valproate in monotherapy.
We have calculated malformation rates and relative risks as a function of valproate exposure.
Results: Outcome data were available for 1109 pregnancies exposed to valproate in monotherapy.
Exposure to 1000 mg a day or more of valproate was associated with almost double the risk of major
congenital malformation compared with daily valproate doses below 1000 mg daily (8.86% vs 4.88%, RR:
1.7; 95% CI: 1.1–2.9). There were no differences in the risks for malformations between standard release
valproate and controlled release valproate preparations (RR: 1.11; 95% CI: 0.67–1.83) or for those
exposed to single or multiple daily administrations (RR: 0.99, 95% CI: 0.58–1.70).
Conclusion: Prescribing controlled release valproate or multiple daily administrations in pregnancy did
not reduce the risk for malformations. Higher malformation rates observed with in utero exposure to
valproate are more likely related to total daily dose, rather than peak serum levels.
? 2012 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +44 2890 634590; fax: +44 2890 235258.
E-mail address: firstname.lastname@example.org (J. Craig).
Contents lists available at SciVerse ScienceDirect
jou r nal h o mep age: w ww.els evier .co m/lo c ate/ys eiz
1059-1311/$ – see front matter ? 2012 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.
generalised epilepsy syndromes, where valproate is more effective
than other AEDs.15
The UK Epilepsy and Pregnancy Register is a prospective,
observational, registration and follow-up study that was set up to
determine the relative safety of all AEDs taken in pregnancy. In this
report we have focussed on all pregnancies that were exposed
solely to valproate during the first trimester from December 1996
through April 2010.
Registration and follow up forms as well as the UK Epilepsy and
Pregnancy Register computer database were reviewed for all
women having registered a pregnancy on valproate monotherapy.
Suitable cases were women with epilepsy who became pregnant
whilst taking valproate (either standard release valproate or
controlled release valproate) in monotherapy and who were
referred before the outcome of the pregnancy was known. The
main outcome measure was the MCM rate. The rate of minor
congenital malformations (mCM) was also recorded. Full method-
ological details have been published previously.4Recording of
serum levels of AEDs, including valproate, is not part of the
methodology of the UK Epilepsy and Pregnancy Register and were
therefore not available for this study. Valproate preparation and
dose is recorded at registration only in the majority of cases.
A MCM was defined as an abnormality of an essential
embryonic structure requiring significant treatment and present
at birth or discovered in the first 6 weeks of life. Disorders
not conforming to this definition were assigned as mCMs based
on the definitions and lists of disorders in the EUROCAT
MCM and mCM rates were calculated for standard release
valproate and controlled release valproate at any dose and at doses
below and above 1000 mg daily. We also attempted to estimate the
potential influence of peak plasma valproate concentrations by
comparing malformation rates in women taking standard release
valproate with those taking controlled release valproate or divided
doses of standard valproate.
2.1. Statistical analysis
Malformation rates were calculated as [(total number of live
births with a malformation) + (total number of pregnancy losses
with a malformation)]/[(total number of live births) + (total
number of pregnancy losses with a malformation)]. For each
malformation rate, 95% confidence intervals (95% CI) were
calculated using the traditional method.17Relative risks with
95% CI were used to compare groups. Student’s t-test, ANOVA and
Chi-squared tests were used in comparison of characteristics
between groups. Significance was determined at p < 0.05. Data for
drug dose and administration schedule were not recorded for some
patients. These patients were excluded from the groups comparing
drug dosage and frequency of administration, irrespective of the
outcome of the pregnancy.
Patient characteristics and pregnancy outcome data.
All valproate exposures
(n = 1109)
Standard release valproate
exposures (n = 814)
Controlled release valproate
exposures (n = 295)
Mean gestational age (weeks)
Preconceptual folic acid
Family history MCM
Major congenital malformation (MCM) and minor congenital malformation (mCM) rates.
MCM rate (95% CI)
Minor CM rate (95% CI)
Mean dose mg (SD)
Total valproate exposures (n = 1085)
<1000 mg total daily dose (n = 584)
?1000 mg total daily dose (n = 501)
Total standard release exposures (n = 791)
<1000 mg total daily dose (n = 429)
?1000 mg total daily dose (n = 362)
Standard release valproate (single daily dose) (n = 251)
<1000 mg total daily dose (n = 158)
?1000 mg total daily dose (n = 94)
Standard release valproate (divided daily dose) (n = 538)
<1000 mg total daily dose (n = 271)
?1000 mg total daily dose (n = 268)
Valproate controlled release (n = 294)
<1000 mg total daily dose (n = 154)
?1000 mg total daily dose (n = 140)
Valproate controlled release (single daily dose) (n = 95)
<1000 mg total daily dose (n = 58)
?1000 mg total daily dose (n = 37)
Valproate controlled release (multiple daily doses) (n = 198)
<1000 mg total daily dose (n = 96)
?1000 mg total daily dose (n = 103)
E. Mawhinney et al. / Seizure 21 (2012) 215–218
Through April 2010, outcome data were available for 1109
pregnancies exposed to valproate as monotherapy. Of these, 1044
pregnancies (94.1%) resulted in a live birth and 65 (5.9%) a
pregnancy loss. Seven additional pregnancies were lost to follow
up and are not included in analysis. Valproate dose and
administration schedule was not recorded in 24. Patient char-
acteristics and pregnancy outcome data are outlined in Table 1.
Eight hundred and fourteen patients (73.4%) were exposed to
standard release valproate and 295 (26.6%) to controlled release
valproate. There were no statistical differences between the groups
in terms of mean gestational age, preconceptual folic acid intake
and family history of MCMs (Table 1).
The total MCM rate for all valproate exposures was 6.7% (95% CI:
5.2–8.2%) and for mCMs 7.7% (95% CI: 6.1–9.3%) (Table 2). Forty-six
point two percent were exposed to a total daily dose of 1000 mg or
more and 53.8% to less than 1000 mg a day. The risk of MCM
increased with total daily dose (Fig. 1) with the risk in the high dose
group being almost double that of the low dose group (8.9% vs
4.9%; RR 1.8 (95% CI: 1.1–2.9)). This pattern was also seen for mCMs
(9.1% with total daily doses at or above 1000 mg vs 5.8% for total
daily doses less than 1000 mg; RR 1.2 (95% CI: 1.0–2.5)).
The risk for MCMs and mCMs did not vary depending on
whether valproate was prescribed as standard or controlled
release. The relative risk (RR) for MCMs for those exposed to
controlled release valproate compared with standard release
valproate was 1.1 (95% CI: 0.7–1.8). The proportion of those
exposed to a total daily dose of 1000 mg or more was similar for the
standard and controlled release groups. No significant difference in
the risk of MCMs was noted between these groups (RR 1.2 (95% CI:
Two hundred and fifty one patients took standard release
valproate once daily with 832 either taking controlled release
valproate or standard release valproate, administered twice daily
or more (divided dose group). MCM rates were no different for
those pregnancies exposed to once daily standard valproate
compared with those taking either controlled release valproate
or standard release valproate in divided doses (RR 1.0 (95% CI: 0.6–
1.7)). This was also the case when pregnancies exposed to less than
1000 mg a day or to 1000 mg a day or more were considered (RR
0.6 (95% CI: 0.3–1.1)).
Overall there were no significant differences between the types
of MCMs seen in those exposed to standard release and controlled
release valproate (p = 0.99) (Table 3). Clefting abnormalities,
however, were almost five times more common in those exposed
to standard release valproate, with hypospadias and other
disorders of the genitourinary tract being almost twice as likely
in those exposed to controlled release valproate.
The MCM rate observed for valproate exposed pregnancies in
this study at 6.7% was lower than that reported by other groups,
where rates ranging between 8.3%9and 16.8%6have been recorded.
A trend towards increasing MCM risk with higher total daily
valproate dose has been previously observed in studies by the UK
Epilepsy and Pregnancy Register, EURAP, the North American AED
Pregnancy Registry and the Australian Epilepsy and Pregnancy
Registry.4–8In keeping with this, we found a statistically significant
positive dose response with almost a doubling of risk for those
exposed to more than 1000 mg per day compared with those on
less than 1000 mg per day (8.9% vs 4.9%). Use of controlled release
valproate did not significantly alter the risk for MCMs, irrespective
of total daily dose. Likewise, the risk for MCMs and mCMs was not
affected by the number of daily administrations of valproate. This
was demonstrated by the similar rates of MCMs, irrespective of
total daily dose, for pregnancies exposed either to a single dose of
valproate and those exposed to divided daily doses of controlled
release valproate. The types of MCMs were similar for those
Fig. 1. MCM rate according to valproate total daily dose (mg).
Types of MCMs for all valproate, standard release valproate and controlled release valproate exposures.
Type of MCM
All valproate monotherapy
exposures (n = 1109)
Standard release valproate
monotherapy exposures (n = 814)
Valproate controlled release
monotherapy exposures (n = 295)
p-Value standard release vs
Neural tube defect
E. Mawhinney et al. / Seizure 21 (2012) 215–218
exposed to standard and controlled release valproate. Clefting Download full-text
abnormalities, however, were almost five times more likely in
those exposed to standard release valproate and hypospadias and
other disorders of the genito-urinary tract almost twice as
common in those exposed to controlled release valproate. These
findings were not statistically significant and demonstrate the very
large number of pregnancies that would be required to detect such
differences, if they exist.
One obvious criticism of our study is that we did not record
serum valproate levels. None of the national or international
epilepsy and pregnancy registries record this information.18
Conducting such a study where valproate levels were recorded
would be difficult since it would require repeated trough and peak
serum valproate levels to be recorded throughout the pregnancy.
Considering MCMs only and taking into account the different
periods that are important for organogenesis would clearly require
very detailed study in the first trimester.
That peak serum levels of valproate may be an important
determinant for the risk of MCMs has been shown in a mouse model
which compared the effects of a conventional injection regime of
valproate and a steady drug infusion delivered by implanted mini-
pumps. In this study, a 10 times higher dose was required in the
infusion regimen to produce neural tube defect rates at a rate similar
to that observed in the injection regimen (up to 60% of foetuses).
Higher rates of embryolethality and foetal weight retardation were
noted however with the implanted minipumps.9
Prescribing valproate as a continuous infusion is clearly not
feasible in humans were the only practical means available to
minimise peak valproate concentrations are to reduce the dose,
split it into multiple daily administrations or prescribe controlled
release valproate. Our results, in contrast to the study in mice,9
showed that attempting to minimise peak serum concentrations
has no significant effect on MCM risk and that it is only by reducing
the total daily dose of valproate that the risk for MCMs and mCMs
can be lowered. We have therefore not been able to substantiate
the advice that for women planning pregnancy, or who become
pregnant, while taking valproate, that the daily dose should be split
and controlled release valproate prescribed.11–14In the absence of
any definite harm we would however still advocate this approach.
Clearly in some women increasing the frequency of drug
administration can reduce adherence, and discussion of the risks
and benefits is vital before recommending multiple daily admin-
istrations. This is the first report to study the effects of formulation
and dosage administration schedule for valproate in pregnancy
and there will be an opportunity for the other epilepsy and
pregnancy registers to do the same. Furthermore, with increasing
evidence that valproate can have an even more significant effect on
cognitive development and behaviour,19–26which also appears to
be dose related, further work is clearly needed to determine the
importance of drug formulation and dosage regime.
1. Olafsson E, Hallgrimsson JT, Hauser WA, Ludvigsson P, Gudmundsson G.
Pregnancies of women with epilepsy: a population-based study in Iceland.
2. Holmes LB, Harvey EA, Coull BA, Huntington KB, Khoshbin S, Hayes AM, et al.
The teratogenicity of anticonvulsant drugs. N Engl J Med 2001;344:1132–8.
3. Kaneko S, Battino D, Andermann E, Wada K, Kan R, Takeda A, et al. Congenital
malformations due to anti-epileptic drugs. Epilepsy Res 1999;33:145–58.
4. Morrow J, Russell A, Guthrie E, Parsons L, Robertson I, Waddell R, et al.
Malformation risks of antiepileptic drugs in pregnancy: a prospective study
from the UK Epilepsy and Pregnancy Register. J Neurol Neurosurg Psychiatry
5. Vajda FJE, Hitchcock A, Graham J, Solinas C, O’Brien TJ, Lander CM, et al. Foetal
malformations and seizure control: 52 months data of the Australian Pregnancy
Registry. Eur J Neurol 2006;13:645–54.
6. Vajda FJ. The Australian pregnancy register of antiepileptic drugs: 10 years of
progress. J Clin Neurosci 2010;17(December (12)):1485–8.
7. Diav-Citrin O, Shechtman S, Bar-Oz B, Cantrell D, Amon J, Omoy A. Pregnancy
outcome after in utero exposure to valproate: evidence of dose relationship in
teratogenic effect. CNS Drugs 2008;22(4):325–34.
8. Tomson T, Battino D, Bonizzoni E, Craig J, Lindhout D, Sabers A, et al. Dose
dependent risk of malformations with antiepileptic drugs: an analysis of data
from the EURAP epilepsy pregnancy registry. Lancet Neurol 2011;10:609–17.
9. Nau H. Teratogenic valproic acid concentrations: infusion by implanted mimi-
pumps vs conventional injection regimen in the mouse. Toxicol Appl Pharmacol
10. Kondo T, Tokinaga N, Suzuki A, Ono S, Yabe H, Kaneko S, et al. Altered
Pharmacokinetics and metabolism of valproate after replacement with the
slow release formulation in epileptic patients. Pharmacol Toxixol 2002;90:130–
11. Delgado-Escuera AV, Janz D. Consensus guidelines preconception counselling,
management and care of the pregnant woman with epilepsy. Neurology
12. Scottish Obstetric Guideline and Audit Project. The management of pregnancy in
women with epilepsy, No. 5. SPCERH Publications; 1997.
13. Crawford P, Appleton R, Betts T, Duncan J, Guthrie E, Morrow J. The women with
epilepsy guidelines development group. Best practice guidelines for the man-
agement of women with epilepsy. Seizure 1999;8:201–17.
14. Crawford P. Best practice guidelines for the management of women with
epilepsy. Epilepsia 2005;46(Suppl. 9):117–24.
15. Marson AG, Al-Kharusi AM, Alwaidh M, Appleton R, Baker GA, Chadwick DW,
et al., for the SANAD Study group. The SANAD study of effectiveness of
valproate, lamotrigine, or topiramate for generalised and unclassifiable epilep-
sy: an unblinded randomised controlled trial. Lancet 2007;369:1016–26.
16. De Wals P, Mastroiacovo P, Weatherall JAC, Lechat M, editors. EUROCAT guide
for the registration of congenital anomalies. Brussels: European Union; 1984.
17. Wilson EB. Probable inference, the law of succession, and statistical inference. J
Am Stat Assoc 1927;22:209–12.
18. Tomson T, Battino D, Craig J, Hernandez-Diaz S, Holmes LB, Lindhout D, et al.
Pregnancy registries: differences, similarities, and possible harmonization.
19. Adab N, Kini U, Vinten J, Ayres J, Baker G, Clayton-Smith J, et al. The longer term
outcome of children born to mothers with epilepsy. J Neurol Neurosurg Psychia-
20. Adab N, Jacoby A, Smith D, Chadwick D. Additional educational needs in
children born to mothers with epilepsy. J Neurol Neurosurg Psychiatry
21. Meador KJ, Baker GA, Browning N, Clayton-Smith J, Combs-Cantrell DT, Cohen
M, et al. Cognitive function at 3 years of age after fetal exposure to antiepileptic
drugs. N Engl J Med 2009;360:1597–605.
22. Meador KJ, Baker GA, Browning N, Cohen MJ, Clayton-Smith J, Kalayjian LA,
et al., for the NEAD Study Group. Foetal antiepileptic drug exposure and verbal
versus non-verbal abilities at three years of age. Brain 2011;134:396–404.
23. McVearry KM, Gaillard WD, Van Meter J, Meador KJ. A prospective study of
cognitive fluency and originality in children exposed in utero to carbamaze-
pine, lamotrigine, or valproate monotherapy. Epilepsy Behav 2009;16:609–16.
24. Bromley RL, Mawer G, Love J, Kelly J, Purdy L, McEwan L, et al. Early cognitive
development in children born to women with epilepsy: a prospective report.
25. Nadebaum C, Anderson VA, Vajda F, Reutens DC, Barton S, Wood AG. Language
skills of school-aged children prenatally exposed to antiepileptic drugs. Neu-
26. Cummings C, Stewart M, Stevenson M, Morrow J, Nelson J. Neurodevelopment
of children exposed in utero to lamotrigine, sodium valproate and carbamaze-
pine. Arch Dis Child 2011;96(7):643–7.
E. Mawhinney et al. / Seizure 21 (2012) 215–218