Seizure 51 (2017) 95–101

Contents lists available at ScienceDirect

Seizure
journal homepage: www.elsevier.com/locate/yseiz

Regular Article

Lamotrigine versus valproic acid monotherapy for generalised

epilepsy: A meta-analysis of comparative studies
Linjun Tanga,* , Linbo Gea , Weijun Wua , Xuguang Yanga , Pinhe Ruia , Yong Wua , Wan Yub ,
Xi Wangc
a
Department of Neurosurgery, Tongling Municipal Hospital, Tongling, Anhui, China
b
Department of Neurosurgery, The Affiliated Hospital of Integrated Traditional Chinese and Western Medicine, Nanjing University of Chinese Medicine,
Nanjing, Jiangsu, China
c
Department of Neurosurgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu, China

A R T I C L E I N F O A B S T R A C T

Article history: Purpose: The standard for generalized epilepsies (GE) monotherapy in treatment is valproic acid (VPA)
Received 10 January 2017 and lamotrigine (LTG) has been proposed as an alternative to VPA. This study aimed to evaluate the safety
Received in revised form 19 July 2017 and efficacy of LTG on GE seizure in comparison with VPA.
Accepted 3 August 2017
Method: A search was conducted based on the databases from Pubmed, Embase and the Cochran database
up to February 2017. The relative risk odds ratios (ORs) and the relevant 95% confidence intervals (CI)
Keywords: were determined.
Valproic acid
Results: Five randomized controlled trials and four observational cohort studies involving 1732 cases
Lamotrigine
Generalized epilepsies
were included. The results indicated that VPA was significantly superior to LTG for the outcome rate to
Monotherapy treatment withdrawal for any reason and seizure freedom. The ORs and 95% CI of VPA versus LTG for
Meta-analysis withdrawal after 12- and 24-month treatment were 0.39(0.27, 0.56) and 0.50(0.14, 1.75), respectively, and
were 3.51(2.68, 4.59) and 8.58(5.40, 13.63)for 12- and 24- month seizure free intervals, respectively.
Moreover, the risk of adverse effects (OR (95%CI); 1.11(0.61–2.01)) was not significantly different between
the two groups. However, the treatment withdrawal due to lack of seizure control were in the LTG group
(OR (95%CI); 0.15(0.10–0.23)), while the treatment withdrawal due to intolerable side effects were in the
VPA group (OR (95%CI); (1.75(1.10–2.80)).
Conclusions: The meta-analysis suggests that VPA appears to be a better choice in controlling seizure
following GE. However, therapy should be switched to alternative monotherapy if an adequate trial of
VPA monotherapy is not effective and intolerable, especially in young women.
© 2017 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.

1. Introduction be divided into clonic, tonic, absence, atonic, tonic-clonic and

Epilepsy influences 65 million people worldwide and entails a
major burden in seizure-related disability, stigma, mortality, and
costs [1,2]. Around 30–40% of patients have seizures that are
generalized at onset. In general, generalized epilepsies (GE) are
determined and affect otherwise normal people of both sexes and
races [3]. Besides, GE was characterized by widespread involve-
ment of bilateral cortical regions at the onset. They are usually
accompanied by impairment of consciousness, which can further
* Corresponding author at: Department of Neurosurgery, Tongling Municipal
Hospital, 2999 Changjiang West Road, Tongling 244000, Anhui, China.
E-mail address: doctortlj@163.com (L. Tang).
myoclonic seizure types [4].
The goal in the first line pharmacologic management of
epilepsy is monotherapy due to it is effective, well tolerated and
associated with low costs, higher quality of life as well as better
patient compliance. A long-term (up to 6 years) un-blinded study
was designed by The Standard and New Antiepileptic Drugs
(SANAD) trial, which declared that valproic acid (VPA)was
identified as a first-line treatment for patients diagnosed with
generalised-onset seizures [5]. VPA is a very effective anticonvul-
sant drug for GE, yet it carries some risks connected with its side
effects profile [6,7]. Particularly for women of childbearing age,
VPA was concerned about higher rates of teratogenicity and
delayed cognitive development in children in utero. Taking it into
consideration, lamotrigine(LTG) has been suggested as an alterna-
tive to VPA [8]. LTG has been proposed as first line new AED in

http://dx.doi.org/10.1016/j.seizure.2017.08.001
1059-1311/© 2017 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.
96 L. Tang et al. / Seizure 51 (2017) 95–101

2. Methods

2.1. Search strategy

The following electronic databases were searched till February

2017 such as Pubmed, Embase and the Cochrane database. The
electronic search strategy included the terms, respectively,
epilepsy; seizure; myoclonic epilepsy; absence epilepsy; tonic-
clonic epilepsy; clonic epilepsy; tonic epilepsy; atonic epilepsy;
generalised epilepsies; lamotrigine; valproate and valproic acid.
Studies only in English were retrieved. Two authors reviewed the
titles and abstracts of articles obtained from electronic databases
respectively. If the abstract was relevant to this study, we read the
full text and decided which articles were eligible for full-text
review and disagreement was settled by mutual discussion.

2.2. Inclusion and exclusion criteria

Fig. 1. The flow diagram shows the selection of studies for the meta-analysis.
treating childhood absence, juvenile absence, juvenile myoclonic
and generalized tonic–clonic epilepsy according to National
Institute for Health and Care Excellence (NICE) guidelines [9].
LTG is a phenyltriazine derivative which acts through inhibition of
voltage-activated sodium channels and possibly calcium channels,
preventing the release of glutamate [10]. Besides, LTG is also
effective in controlling absence seizures and generalized tonic–
clonic seizures [11,12]. Nevertheless, there are some reports of
myoclonic seizure exacerbation [13].
Although conventional anti-epileptic drug VPA and the modern
LTG are identified as optimal first line or second-line monotherapy
for GE, effectiveness and course of treatment vary between the
patients and still remain a matter of discussion [7,9,14]. In 2007,
Tudur Smith et al. [15] performed a meta-analysis to compare AEDs
for different types of epilepsy. However, they included only one
study to compare VPA with LTG. Their result might be not robust.
Therefore, we conducted a meta-analysis of published trials
through comparing VPA with LTG to evaluate the effect on total
withdrawal rate, the seizure-freedom rate, and adverse events in
patients with GE.
Inclusion criteria were: (i) comparative study(randomized
controlled trial (RCT), cohorts, case-controls and observational
studies), (ii) investigated GE patients, (iii) the study compared VPA
to LTG(iv) reported the number of outcome events in different
interventions. The aim was to include only RCTs in the analysis.
However, due to their paucity, other intervention studies and
observational studies were included among which RCTs and
observational studies were analyzed separately in subgroup for the
reason that direct comparison between the estimates of observa-
tional studies and RCTs might be misleading. Exclusion criteria
were shown as follows(i) review articles, meta-analysis, and
guidelines, (ii) unavailability of a medical treatment comparison
group, (iii) studies with no LTG arm, (iv) seizure data was not
reported.
2.3. Data extraction
Two reviewers independently extracted the data. In case of
disagreement between the two reviewers, a third reviewer
extracted the data. The following information was extracted from
the trails including the name of first author; country of origin;
patients’ characteristics (mean age, gender) as well as operational
definitions and outcomes. For dichotomous outcomes, the number

Table 1
Characteristics of studies included in the meta-analyses.

First Author, Country Study type Population Type Analyzed Mean Age % Males Dosage/day Seizures assessed at
Year (year)

VPA LTG VPA LTG VPA LTG VPA LTG

Nicolson, 2004 UK Cohort IGE 549 156 12.2 14.1 46 26.2 1286 mg 324 mg 12 months, 24 months & 60
[10] (mean) (mean) months
Coppola, 2004 Italy Randomized Typical Absence Seizures 19 19 7.5 7.5 52.6 36.8 20– 2–12 mg/ 1 month, 3 months & 12
[20] Controlled Trial 30 mg/kg kg months
Steinhoff, 2005 Germany Randomized GE 30 33 23.3 22.3 46.7 39.4 9.0– 1.5–7.4 17 weeks and 24 weeks
[24] Controlled Trial 24.4 mg/ mg/kg
kg
Mazurkiewicz, Poland Cohort IGE 132 82 9.5 8.2 46.9 32.9 20– 5–13 mg/ 12 months & 24 months
2010 [27] 32 mg/kg kg
Hwang, 2012 Korea Cohort Childhood Absence 59 21 6.4 6.6 33.9 23.8 15– 3–6 mg/ 3 months, 6 months, 12
[26] Epilepsy 45 mg/kg kg months or 24 months
Machado, 2013 Cuba Randomized Juvenile Myoclonic 31 41 15.3 16.3 32.3 36.6 900– 150– 3, 6 or 24 months
[23] Controlled Trial Epilepsy 2700 mg 400 mg/
Glauser, 2013 America Randomized Childhood Absence 146 146 7.5 7.5 48 38 10– 0.3– 16, 20 weeks & 12 months
[22] Controlled Trial Epilepsy 60 mg/kg 12 mg/kg
Chowdhury, UK Cohort Juvenile Myoclonic 142 66 16 16 44 21 NP NP over 12 months
2016 [25] Epilepsy
Giri, 2016 [21] India Randomized Idiopathic Generalized 30 30 18– 18– 66.7 56.7 10– 1–12 mg/ 3 months, 6 months & 12
Controlled Trial Tonic Clonic Seizures 70 70 30 mg/kg kg months

UK, united kingdom; IGE, idiopathic generalised epilepsies; GE, generalised epilepsies; VPA, valproic acid; LTG, lamotrigine; mg, milligram; kg, kilogram; NP, not provided.
 L. Tang et al. / Seizure 51 (2017) 95–101 97

the Cochrane Non-Randomized Studies Methods Working Group.

When analyzing case-control trials, NOS addresses three areas
including selection, comparability and exposure, while it includes
selection, comparability and outcome in cohort studies[16]. A
quality score of 0–9 points was allocated to each nonrandomized
study. RCTs with low risk of bias and nonrandomized studies
achieving 7 points were considered to be of high quality. This
scale was developed for application in systematic reviews and
meta-analyses.

2.6. Statistical analysis

Dichotomous variables were respectively analyzed by odds

ratios. Odds ratios (ORs) for binary outcomes were chosen due to
be associated with less heterogeneity in meta-analysis than risk
differences or relative risks [17]. Cochran's Q-statistic test was
applied to access between-study heterogeneity and I2 were used to
test for heterogeneity between the included studies (P < 0.05 is
considered for significant heterogeneity). The random effects
model results were used if there was evidence for heterogeneity
between studies because it provides a more conservative effect
than the fixed-effects model [18]. If heterogeneity was found, we
performed sensitivity analysis through eliminating one study at a
time checking for resolution of heterogeneity. Publication bias was
assessed by the visual funnel plot [19]. Data were analyzed based
on the use of the Review Manager (RevMan version 5.3; Cochrane
Collaboration, Oxford, UK).

3. Result
Fig. 2. Quality assessment of the included RCTs using the Cochrane risk of bias
assessment. 3.1. Study selection

A total of 657 potential trials were identified through using the

Table 2
Quality assessment of studies included in the meta-analysis. first search strategy. A diagram summarizing the process of study
selection is shown in Fig. 1. After applying the selection criteria, five
Study Selection Comparability Outcome Total score
RCTs [20–24] and four cohort studies [10,25–27] were identified in
*** ** ***
Mazurkiewicz et al. [27]
*** * ***
8/9 the final analysis.
Chowdhury et al. [25] 7/9
*** ** ***
Hwang et al. [26] 8/9
Nicolson et al. [10] *** ** ***
8/9 3.2. Characteristic of trials
*
Equal to one point. Together, identified studies enrolled a total of 1732 patients and
**
Equal to two point.
***
Equal to three point.
included studies were published between 2003and 2016. Duration
of follow-up ranged from 1 to24 months. The detail information
was summarized in Table 1. The risk of bias of the RCTs included in
of participants experiencing the outcome and the number assessed this meta-analysis is summarized in Fig. 2. Although most studies
in each treatment group were pointed out. When information was had a low risk of bias, only one study did have an unclear risk.
missing, we attempted to contact the primary author via email to Among the studies, the risk of bias was assessed to be low. After the
seek clarification. use of NOS to evaluate the quality of nonrandomized studies, four
nonrandomized studies were considered being of high quality
2.4. Studied outcomes (Table 2).

The following outcomes relevant to the efficacy of VPA versus
LTG were collected: (1) Proportion of patients with treatment
withdrawal for any reason; (2) Proportion of patients achieving
seizure freedom; (3) Proportion of patients with adverse effects.
2.5. Quality assessment
To evaluate the risk of bias in the included RCTs, two authors
independently assessed the quality of each trial through using the
Cochrane Collaboration’s tool. The following domains were
assessed including selection bias, attrition bias, performance
and detection bias, reporting bias, and other bias. The Newcastle-
–Ottawa Scale (NOS) was used to evaluate the quality of
nonrandomized studies, discriminating between case-control
trials and cohort studies. The NOS was a scale recommended by
3.3. Treatment withdrawal for any reason
Data of six trials [20–22,24,26,27] was available on treatment
withdrawal for12 months including 747 patients. The pooled
estimated OR was 0.39 (95%CI: 0.27–0.56), favoring VPA (P < 0.01).
Moreover, there was no significant heterogeneity among the
identified studies evaluating mortality (I2 = 5%, P = 0.38). In the
subgroup analysis, the results were also robust in RCT and cohort
groups. (Fig. 3A)
As for treatment withdrawal for 24 months, all two trials
[26,27] were included. The estimates of pooled OR was 0.40 (95%
CI: 0.24-0.68), favoring VPA (P < 0.01), yet with significant
heterogeneity (I2 = 76%, P = 0.04). After using the random effects
model, the result presented no significant differences between the
VPA and LTG groups (OR, 0.50; 95% CI 0.14–1.75; P = 0.28). (Fig. 3B)
98 L. Tang et al. / Seizure 51 (2017) 95–101
Fig. 3. Forest plot of treatment withdrawal and patients who had discontinued treatment due to common reasons (A) treatment withdrawal for 12 months (B) treatment
withdrawal for 24 months (C) Discontinued therapy due to lack of seizure control(D) Discontinued therapy due to intolerable side effects. VPA, valproic acid; LTG, lamotrigine;
M-H, Random = Mantel-Haenszel, Random-effects model.
The most common reasons for treatment withdrawal were lack
of seizure control and intolerable side effects. We performed a
subgroup meta-analysis that was stratified by lack of seizure
control and intolerable side effects. Outcomes are shown in Fig. 3C
and D. Due to lack of seizure control, the majority of the patients
with treatment withdrawal were in the LTG group (OR, 0.15; 95% CI
0.10–0.23; P < 0.01).The largest percentage of the subjects
discontinuing due to intolerable side effects were in the VPA
group (OR, 1.75; 95% CI 1.10–2.80; P = 0.02).
3.4. Seizure freedom
Information on seizure-free intervals for 12 months was
available for 1383 patients in six trials [10,20–22,25,26]. The
outcomes demonstrated that VPA cure was superior to LTG (OR,
3.51; 95% CI 2.68 to 4.59, P < 0.01). There was no significant
heterogeneity (I2 = 29%, P = 0.22).In the subgroup analysis, the
results were also robust in RCT and cohort groups (Fig. 4A).
Three studies [10,23,27] including 991 participants, reported
seizure-freedom for 24 months. The outcomes showed that VPA
cure was still superior to LTG (OR, 8.58; 95% CI 5.40–13.63,
P < 0.01). There was no significant heterogeneity (I2 = 0%, P = 0.70).
Besides, in the subgroup analysis, the results were also robust in
RCT and cohort groups (Fig. 4B).
3.5. Adverse effects
Sedation, tremor, weight gain and hair loss were among the four
most common adverse effects in VPA group. Headache, skin rash,
nervousness and dizziness were most common recorded in
patients taking LTG. Eight trials [20–27] reported the incidence
of overall adverse events in the VPA group and LTG group. However,
there was no significant difference between VPA and LTG groups
(OR = 1.11; 95% CI = 0.61–2.01; p = 0.73).Besides, the heterogeneity
between trials was moderate (I2 = 63%, P = 0.008). In the subgroup
analysis, the results did not show difference in RCT and cohort
groups (Fig. 5).
3.6. Subgroup meta-analysis
Considering that the evidence acquired from these RCTs
showed high reliability, we performed a subgroup meta-analysis
that was stratified by RCTs. RCTs and non-randomized studies
presented the same results for the overall OR estimate for
treatment withdrawal, seizure freedom and adverse effects
(Table 3).
3.7. Sensitivity analysis and publication bias
The sensitivity analysis was performed by randomly excluding
one trial and interchanging fixed-effects and random-effects
models from pooled analysis, which confirmed the outcomes of
being stable.
Publication bias was assessed by funnel plots and Egger’s test.
According to Fig. 6, the shape of funnel plot did not reveal an
indication of obvious asymmetry. In addition, we used Egger’s test
to provide statistical evidence of funnel plot symmetry and the
results did not show any proof of publication bias.
4. Discussion
This meta-analysis of five RCTs and four nonrandomized cohort
studies, all of which included a total of 1732 patients through
comparing the efficacy of the VPA and the LTG as optimal first line
monotherapy for GE. Results demonstrated that VPA was
significantly superior to LTG for the outcome rate to treatment
withdrawal for any reason and seizure freedom. Moreover, there
was no difference between treatment groups regarding analysis
adverse effects. Sensitivity analysis indicated that the findings are
robust and not affected by publication bias.
For the treatment of the GE, VPA was the conventional
antiepileptic drug of choice. It is widely considered as the
treatment of choice for generalized epilepsies or difficult-to-
classify seizures [5]. In SANAD, VPA was more efficacious than LTG.
The guidelines were equal to our findings. Based on our analysis,
 L. Tang et al. / Seizure 51 (2017) 95–101 99

Fig. 4. Forrest plot for patients of seizure-freedom (A) 12 months seizure-freedom (B) 24 months seizure-freedom. VPA, valproic acid; LTG, lamotrigine.

we concluded that VPA appeared to be more effective for 12-month
and 24-monthseizure freedom. Moreover, we found that VPA leads
to lower treatment withdrawal for 12-month and 24-monthin
comparison with LTG. Besides, the present study findings are
consistent with studies [5,24,28].
In addition, lack of seizure control and intolerable side effects
were the most common reasons for treatment withdrawal.
Principal reason of earlier discontinuation of therapy in LTG group
was referred to lower efficacy in comparison with VPA group. The
largest percentage of the subjects discontinuing due to intolerable
side effects were in the VPA group. As the adverse event profile of
the drug is often a major determinant in the choice of therapy,
long-term tolerability of AEDs in patients is important, [29]. VPA
had the widest range of side-effects, including weight gain, hair
loss, gynaecomastia and sedation and relatively fewer different
side-effects were reported for LTG. However, there was no
significant difference between treatment groups in terms of
analysis adverse effects. This result was inconsistent with meta-
analysis by Brigo et al. [30]. They demonstrated that more adverse
events occurred in the VPA group. However, their analysis included
only one RCT and absence seizures in children.
Although VPA is the AED of choice in most clinical practice
guidelines for GE, it is associated with teratogenic and adverse
esthetic effects (obesity, hirsutism, alopecia) that have an
especially negative impact on young women [31,32]. The
International League Against Epilepsy (ILAE) and European
Academy of Neurology (EAN) joint taskforce discouraging the
use of VPA in women and girls unless other treatments fail [7].
Besides, the ILAE-EAN taskforce also advocate considering a switch
away from VPA if a female patient established on treatment wished
to consider pregnancy. In addition, LTG was reported as a safe and
effective treatment option for GE from monotherapy trials [33,34].
However, Nicolson et al. [10] suggested that when VPA fails
because of lack of efficacy, switchover to LTG is unlikely to be
100 L. Tang et al. / Seizure 51 (2017) 95–101

Fig. 5. Forrest plot for patients of adverse effects; M-H, Random = Mantel-Haenszel, Random-effects model.

Table 3
Main results concerning the efficacy of valproic acid versus lamotrigine.

Outcomes Number of studies Patients(VPA/LTG) Effect size Model I2 (%) Ph

12 months treatment withdrawal 6 416/331 OR, 0.39; 95% CI 0.27, 0.56 FE 5% 0.38
RCT 4 225/228 OR, 0.37; 95% CI, 0.24, 0.59 FE 0% 0.59
cohort studies 2 191/103 OR, 0.43; 95% CI, 0.24, 0.76 FE 69% 0.07
24 months treatment withdrawal 2 191/103 OR, 0.50; 95% CI, 0.14, 1.75 RE 76% 0.04
RCT – – – – – –
cohort studies 2 191/103 OR, 0.50; 95% CI, 0.14, 1.75 RE 76% 0.04
12 months seizure-freedom 6 945/438 OR, 3.51; 95% CI, 2.68, 4.59 FE 29% 0.22
RCT 3 195/195 OR, 2.71; 95% CI, 1.75, 4.21 FE 0% 0.85
cohort studies 3 750/243 OR, 4.05; 95% CI, 2.88, 5.70 FE 60% 0.08
24 months seizure-freedom 3 712/279 OR, 8.58; 95% CI, 5.40, 13.63 FE 0% 0.7
RCT 1 31/41 OR, 8.40; 95% CI, 4.55, 15.50 FE - -
cohort studies 2 681/238 OR, 8.91; 95% CI, 4.50, 17.62 FE 0% 0.84
adverse effects 8 589/438 OR, 1.11; 95% CI, 0.61, 2.01 RE 63% 0.008
RCT 5 256/269 OR, 1.09; 95% CI, 0.46, 2.60 RE 71% 0.007
cohort studies 3 333/169 OR, 1.07; 95% CI, 0.62, 1.83 FE 53% 0.12

VPA, valproic acid; LTG, lamotrigine; RCT, random controlled trial; FE, fixed-effects model; RE, random-effects model; OR, odds ratio; CI, confidence interval; Ph: p value for
heterogeneity.

successful and a second antiepileptic drug should be added in this

situation. Moreover, two reviews also recommend LTG to female,
especially in pregnancy. Weston et al. [35] found that the children
exposed to VPA were at a higher risk of malformation compared
with children born to women without epilepsy and to women with
untreated epilepsy. There was no increased risk for major
malformation for LTG. Moreover, Bromley et al. [36] demonstrated
that the intelligence quotient (IQ) of children exposed to VPA was
lower than for children born to women without epilepsy. Also, IQ
was significantly lower for children exposed to VPA versus LTG. If,
however, VPA fails because of adverse effects, then it is appropriate
to switch to an alternative antiepileptic drug.
There are several potential limitations of our analysis. Firstly,
unpublished studies were not included because of the difficulty in
accessing their data even though no evidence of publication bias in
the results was obtained. Secondly, for the reason that most of the
studies reported that the doses used were the therapeutic dose
range of the AEDs, it was impossible to assess tolerability by
different doses. However, it is known that most of the adverse
Fig. 6. Funnel plot of publication bias.
 L. Tang et al. / Seizure 51 (2017) 95–101
effects are dose dependent. Thirdly, the result of adverse effects in
this study might be not accurate because there are few studies and
they employ different methodologies [10,23,25]. Methodologies
bias could exist, since dosing was left to the physician’s usual
practice and lower doses of VPA were used in females [23]. In
addition, women were more likely to receive LTG as first line
treatment, although it is unlikely to have affected the outcome
[10]. The analysis of the outcome of adding or switching over to LTG
from VPA is consistent with LTG being less effective [25].
Additionally, one included study was the lack of description of
the blinding of participants and the blinding of outcome assessors;
which were due to the nature of the intervention.
To conclude, this meta-analysis shows that VPA generally
appears to be a better choice in controlling seizure following GE.
Moreover, therapy should be switched to alternative monotherapy
if an adequate trial of VPA monotherapy is not effective and
intolerable. To make any practical recommendations, major
double-blind, RCTs for evaluating the safety, efficacy and quality
of life of patients following GE are required, especially in women
and girls.
Conflict of interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This study was supported by grants from Tongling Health and
Family Planning Commission Science Project (No. 201625, No.
201617).
References
[1] Moshe SL, Perucca E, Ryvlin P, Tomson T. Epilepsy: new advances. Lancet
2015;385:884–98.
[2] O’ Rourkea G, O’ Brien JJ. Identifying the barriers to antiepileptic drug
adherence among adults with epilepsy. Seizure 2016;45:160–8.
[3] Yacubian EM. Juvenile myoclonic epilepsy: Challenges on its 60th anniversary.
Seizure 2017;44:48–52.
[4] Gavvala JR, Schuele SU. New-onset seizure in adults and adolescents: a review.
JAMA 2016;316:2657–68.
[5] Marson AG, Al-Kharusi AM, Alwaidh M, Appleton R, Baker GA, Chadwick DW,
et al. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate
for generalised and unclassifiable epilepsy: an unblinded randomised
controlled trial. Lancet 2007;369:1016–26.
[6] Crespel A, Gelisse P, Reed RC, Ferlazzo E, Jerney J, Schmitz B, et al. Management
of juvenile myoclonic epilepsy. Epilepsy Behav 2013;28(Suppl. 1):S81–6.
[7] Tomson T, Marson A, Boon P, Canevini MP, Covanis A, Gaily E, et al. Valproate in
the treatment of epilepsy in girls and women of childbearing potential.
Epilepsia 2015;56:1006–19.
[8] 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
2006;77:193–8.
[9] Bergey GK. Evidence-based treatment of idiopathic generalized epilepsies
with new antiepileptic drugs. Epilepsia 2005;46(Suppl. 9):161–8.
[10] Nicolson A, Appleton RE, Chadwick DW, Smith DF. The relationship between
treatment with valproate, lamotrigine, and topiramate and the prognosis of
the idiopathic generalised epilepsies. J Neurol Neurosurg Psychiatry
2004;75:75–9.
[11] Christensen J, Sidenius P. Epidemiology of epilepsy in adults: implementing
the ILAE classification and terminology into population-based epidemiologic
studies. Epilepsia 2012;53(Suppl. 2):14–7.
101
[12] Posner EB, Mohamed K, Marson AG. Ethosuximide, sodium valproate or
lamotrigine for absence seizures in children and adolescents. Cochrane
Database Syst Rev 2005;CD003032.
[13] Carrazana EJ, Wheeler SD. Exacerbation of juvenile myoclonic epilepsy with
lamotrigine. Neurology 2001;56:1424–5.
[14] Mole TB, Appleton R, Marson A. Withholding the choice of sodium valproate to
young women with generalised epilepsy: are we causing more harm than
good? Seizure 2015;24:127–30.
[15] Tudur Smith C, Marson AG, Chadwick DW, Williamson PR. Multiple treatment
comparisons in epilepsy monotherapy trials. Trials 2007;8:34.
[16] Oremus M, Oremus C, Hall GB, McKinnon MC, ECT and Cognition Systematic
Review Team. Inter-rater and test-retest reliability of quality assessments by
novice student raters using the Jadad and Newcastle-Ottawa Scales. BMJ Open
2012;2.
[17] Deeks JJ. Issues in the selection of a summary statistic for meta-analysis of
clinical trials with binary outcomes. Stat Med 2002;21:1575–600.
[18] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat
Med 2002;21:1539–58.
[19] Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for
publication bias. Biometrics 1994;50:1088–101.
[20] Coppola G, Auricchio G, Federico R, Carotenuto M, Pascotto A. Lamotrigine
versus valproic acid as first-line monotherapy in newly diagnosed typical
absence seizures: an open-label, randomized, parallel-group study. Epilepsia
2004;45:1049–53.
[21] Giri VP, Giri OP, Khan FA, Kumar N, Kumar A, Haque A. Valproic acid versus
lamotrigine as first-line monotherapy in newly diagnosed idiopathic
generalized tonic -clonic seizures in adults – a randomized controlled trial.
J Clin Diagn Res 2016;10:FC01–4.
[22] Glauser TA, Cnaan A, Shinnar S, Hirtz DG, Dlugos D, Masur D, et al.
Ethosuximide, valproic acid, and lamotrigine in childhood absence epilepsy:
initial monotherapy outcomes at 12 months. Epilepsia 2013;54:141–55.
[23] Machado RA, Garcia VF, Astencio AG, Cuartas VB. Efficacy and tolerability of
lamotrigine in juvenile myoclonic epilepsy in adults: a prospective, unblinded
randomized controlled trial. Seizure 2013;22:846–55.
[24] Steinhoff BJ, Ueberall MA, Siemes H, Kurlemann G, Schmitz B, Bergmann L,
et al. The LAM-SAFE study: lamotrigine versus carbamazepine or valproic acid
in newly diagnosed focal and generalised epilepsies in adolescents and adults.
Seizure 2005;14:597–605.
[25] Chowdhury A, Brodie MJ. Pharmacological outcomes in juvenile myoclonic
epilepsy: support for sodium valproate. Epilepsy Res 2016;119:62–6.
[26] Hwang H, Kim H, Kim SH, Kim SH, Lim BC, Chae JH, et al. Long-term
effectiveness of ethosuximide, valproic acid, and lamotrigine in childhood
absence epilepsy. Brain Dev 2012;34:344–8.
[27] Mazurkiewicz-Beldzinska M, Szmuda M, Matheisel A. Long-term efficacy of
valproate versus lamotrigine in treatment of idiopathic generalized epilepsies
in children and adolescents. Seizure 2010;19:195–7.
[28] Stephen LJ, Sills GJ, Leach JP, Butler E, Parker P, Hitiris N, et al. Sodium valproate
versus lamotrigine: a randomised comparison of efficacy, tolerability and
effects on circulating androgenic hormones in newly diagnosed epilepsy.
Epilepsy Res 2007;75:122–9.
[29] Montouris G, Abou-Khalil B. The first line of therapy in a girl with juvenile
myoclonic epilepsy: should it be valproate or a new agent? Epilepsia 2009;50
(Suppl 8):16–20.
[30] Brigo F, Igwe SC. Ethosuximide, sodium valproate or lamotrigine for absence
seizures in children and adolescents. Cochrane Database Syst Rev 2017;2:
CD003032.
[31] Luef G. Female issues in epilepsy: a critical review. Epilepsy Behav
2009;15:78–82.
[32] Christensen J, Gronborg TK, Sorensen MJ, Schendel D, Parner ET, Pedersen LH,
et al. Prenatal valproate exposure and risk of autism spectrum disorders and
childhood autism. JAMA 2013;309:1696–703.
[33] Coppola G, Licciardi F, Sciscio N, Russo F, Carotenuto M, Pascotto A. Lamotrigine
as first-line drug in childhood absence epilepsy: a clinical and neurophysio-
logical study. Brain Dev 2004;26:26–9.
[34] Holmes GL, Frank LM, Sheth RD, Philbrook B, Wooten JD, Vuong A, et al.
Lamotrigine monotherapy for newly diagnosed typical absence seizures in
children. Epilepsy Res 2008;82:124–32.
[35] Weston J, Bromley R, Jackson CF, Adab N, Clayton-Smith J, Greenhalgh J, et al.
Monotherapy treatment of epilepsy in pregnancy: congenital malformation
outcomes in the child. Cochrane Database Syst Rev 2016;11:CD010224.
[36] Bromley R, Weston J, Adab N, Greenhalgh J, Sanniti A, McKay AJ, et al.
Treatment for epilepsy in pregnancy: neurodevelopmental outcomes in the
child. Cochrane Database Syst Rev 2014;CD010236.