Epilepsy & Behavior 54 (2016) 150–157

Contents lists available at ScienceDirect

Epilepsy & Behavior

journal homepage: www.elsevier.com/locate/yebeh

Clinical Research

Specific adverse effects of antiepileptic drugs — A true-to-life

monotherapy study
Alexander B. Kowski ⁎,1, Florian Weissinger 1, Verena Gaus, Pawel Fidzinski, Florian Losch, Martin Holtkamp
Epilepsy-Center Berlin-Brandenburg, Department of Neurology, Charité — Universitätsmedizin Berlin, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Background: In patients taking antiepileptic drugs (AEDs) for epilepsy, adverse effects (AEs) often lead to unfa-
Received 4 September 2015 vorable quality of life, impaired adherence, and, eventually, discontinuation of pharmacological treatment. In a
Revised 20 October 2015 true-to-life sample of subjects from our academic epilepsy outpatient clinic, we aimed to identify predictors
Accepted 8 November 2015 for overall high AE burden and for specific AEs focusing on patients on monotherapy.
Available online 18 December 2015
Methods: All patients ≥16 years of age with epilepsy for ≥12 months were routinely asked to complete the Liver-
pool Adverse Event Profile (LAEP) just before their appointment. Demographic, epilepsy, and treatment variables
Keywords:
Adherence
were derived from our comprehensive outpatient database.
Epileptic seizure Results: Out of 841 patients, 438 (61% female, mean age: 44.7 ± 17.1 years) on monotherapy were included in
Liverpool Adverse Event Profile this study. Levetiracetam (n = 151), lamotrigine (n = 167), valproic acid (n = 73), or controlled-release carba-
Partial epilepsy mazepine (n = 47) were the most commonly used antiepileptic drugs (AEDs). Independent predictors for gen-
Seizure freedom eral high AE burden (LAEP score ≥ 45) were duration of epilepsy, lack of 12-month seizure freedom, and partial
Antiepileptic drug epilepsy, but none of the four individual AEDs. The most frequent LAEP-defined specific AEs were sleepiness, dif-
ficulty concentrating, tiredness, and memory problems. The three most frequent independent predictors for each
of the 19 AEs were lack of 12-month seizure freedom (13/19 AEs), individual AED (7/19 AEs), and partial epilepsy
(6/19 AEs). Levetiracetam was independently associated with anger/aggression, nervousness/agitation, upset
stomach, depression, and sleep disturbance; lamotrigine with nervousness/agitation, upset stomach, and difficul-
ty concentrating; and valproic acid with upset stomach and shaky hands.
Conclusion: Individual AEDs independently predicted some specific AEs, but not overall high AE burden. Our
findings may help to characterize patients with epilepsy who are at high risk for specific AEs. Dose reduction
or change to another AED may reduce LAEP score and potential nonadherence.
© 2015 Elsevier Inc. All rights reserved.

1. Introduction depression have been found to be predictors for abundant occurrence

Pharmacological treatment of epilepsy is often accompanied by
adverse effects (AEs) [1] which are one of the main reasons for not
achieving the dose required for adequate seizure control and which
exert a significantly negative impact on quality of life (QoL) [2]. Thus,
AEs may result in impaired treatment adherence and, in a substantial
proportion of patients, eventually, in discontinuation of antiepileptic
drugs (AEDs) [3,4]. Early identification of patient populations at risk
for high AE burden is critical to counter potential treatment failures.
Female sex, younger age at epilepsy onset, older age at investigation,
symptomatic epilepsy, higher seizure frequency, AED polytherapy, and
⁎ Corresponding author at: Epilepsy-Center Berlin-Brandenburg, Department of
Neurology, Charité — Universitätsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin,
Germany. Tel.: +49 30 450 660 266; fax: +49 30 450 560 938.
E-mail address: alexander.kowski@charite.de (A.B. Kowski).
1
These authors contributed equally to the manuscript.
of AEs [5–8]. So far, the role of individual AEDs to predict overall high
AE load and specific AE is not well determined. Due to potentially addi-
tive or even supra-additive pharmacodynamic effects in polytherapy
regimen [9], possible AED-specific AEs can only be assessed in patients
treated with AED monotherapy. Studies on patients treated with mono-
therapy have not yet able to demonstrate any such differences on AE
occurrence between commonly administered AEDs [5,10,11]. In these
studies, however, the respective number of patients treated with indi-
vidual AEDs was supposedly too small to detect clear-cut effects.
The current study aimed to identify independent predictors for an
overall high AE burden and for specific AEs in a large heterogeneous
population of consecutive patients with epilepsy treated with AED
monotherapy at an academic outpatient clinic. Particular emphasis
was put on the contribution of commonly administered individual
AEDs for occurrence of specific AEs. For standardized coverage and
quantification of AEs, we employed the well-validated 19-item Liver-
pool Adverse Event Profile (LAEP) [12]. Identification of specific AEs
caused by individual AEDs may allow for significant dose reduction or

http://dx.doi.org/10.1016/j.yebeh.2015.11.009
1525-5050/© 2015 Elsevier Inc. All rights reserved.
 A.B. Kowski et al. / Epilepsy & Behavior 54 (2016) 150–157 151

rather change to another substance in order to increase QoL, to reduce
nonadherence, and to improve overall treatment success.
2. Methods
2.1. Subjects and data collection
Since the year 2011, all patients attending our tertiary adult epilepsy
outpatient clinic (Charité — Universitätsmedizin Berlin, Germany) rou-
tinely have completed the LAEP questionnaire. This 19-item self-report
instrument quantifies occurrence of common AEs during the previous
4 weeks by a 4-point Likert scale (from 1 = “never” to 4 = “always or
often”) [12]. The total score ranges from 19 to 76, with a supposed cutoff
≥45 for “high toxicity” [2]. We complemented the LAEP by two 11-score
numerical rating scales (NRS, from 0 = “worst” to 10 = “best”) for
overall health status and QoL during the last 4 weeks, respectively.
The LAEP and the two NRS were filled in by the patients without any
comments by medical staff and prior to contact with their physician
on that particular day. The questionnaire can be completed within
5 min, which is feasible in the routine clinical setting.
For every patient, only the last LAEP questionnaire filled in between
July 01, 2011 and July 31, 2014 was considered for retrospective analy-
sis. Patients aged 16 years or older had to have epilepsy according
to criteria defined by the International League Against Epilepsy (ILAE)
[13] for at least 1 year, were excluded from the study if they had obvious
mental retardation, and had to have sufficient knowledge of German
language (Fig. 1). Data on patients' age at completion of the question-
naire, sex, epilepsy duration, epilepsy syndrome, 12-month terminal
seizure remission, current AED treatment, and drug resistance accord-
ing to ILAE criteria [14] were obtained from our comprehensive

Fig. 1. Stratification of study population. A total of 1033 patients filled in the Liverpool Adverse Event Profile and the two additional numerical rating scales on overall health status and
quality of life. One hundred and ninety-two patients were excluded as they did not fulfill the inclusion criteria. Out of 841 included patients, 507 patients were on monotherapy and 325 on
polytherapy; nine did not take antiepileptic drugs (AEDs) during the previous 4 weeks. Monotherapy with an AED administered to more than 5% of the monotherapy patients included
levetiracetam (LEV, n = 151), lamotrigine (LTG, n = 167), valproic acid (VPA, n = 73), and controlled-release carbamazepine (CR CBZ, n = 47). CLB — clobazam, CLZ — clonazepam, ESL —
eslicarbazepine acetate, ESM — ethosuximide, GBP — gabapentin, LCM — lacosamide, OXC — oxcarbazepine, PGB — pregabalin, PRM — primidone, STM — sulthiame, TPM — topiramate,
ZNS — zonisamide.
152 A.B. Kowski et al. / Epilepsy & Behavior 54 (2016) 150–157

outpatient database. Patients were considered for further subgroup 3. Results

analysis if they were treated with an AED in monotherapy which was
administered to at least 5% of all monotherapy patients.
In order to validate the German version of the LAEP, a subset of 120
patients additionally completed the “Patient Health Questionnaire
for Depression” (PHQ-D, German version [15]) and the “Quality of Life
Inventory 31” (QOLIE-31, German version [16]). Details on validation
of the LAEP are given in the supplemental material.
This cross-sectional study was approved by our institutional review
board (Charité — Universitätsmedizin Berlin).
2.2. Statistical analyses
All analyses were performed using SPSS Statistics 20 (IBM, Armonk,
NY, U.S.A.). Continuous data are given as mean ± standard deviation
or median where appropriate. For group comparisons, the Mann–
Whitney U-test was employed for continuous nonparametric data and
the Chi-square test for categorical variables. The LAEP items initially
quantified by the 4-point Likert scale were dichotomized into
infrequent (1 = “never” and 2 = “rarely”) and frequent occurrences
(3 = “sometimes” and 4 = “always or often”); the latter is simply re-
ferred to as AE occurrence. For identification of independent variables
predicting high LAEP scores, occurrence of specific AEs, and low scores
for overall health status or QoL, linear regression analyses (inclusion
method: stepwise backward, p b 0.1 [p in], p b 0.05 [p out], iteration
20, cutoff set 0.26, constant incorporated) including sex, age at comple-
tion of the questionnaire, duration of epilepsy, 12-month terminal
seizure remission, AED therapy (monotherapy vs. polytherapy or no
therapy, individual substances), drug resistance, and epilepsy syndrome
were performed to estimate odds ratios (OR) with 95% confidence inter-
vals (CI). P values less than 0.050 were considered statistically signifi-
cant. For binary logistic regression of specific AEs, the AED in
monotherapy exhibiting the lowest frequency of that particular AE
was used as reference and set to 1. To test for a linear relationship be-
tween two sets of data, the Pearson correlation coefficient was calculat-
ed and expressed as r.
The individual AED dosage per day was standardized for the com-
parison of drug usage between different substances. Standardized
daily dosage (SDD) was calculated with the formula “individual AED-
dosage per day” divided by “defined daily dose” (DDD; levetiracetam
1500 mg, lamotrigine 300 mg, valproic acid 1500 mg, and controlled-
release (CR) carbamazepine 1000 mg) [17].
3.1. Total study population
A total of 1033 patients had completed at least one questionnaire,
841 of which met the inclusion criteria for this study. Reasons for exclu-
sion are indicated in Fig. 1.
Median LAEP score of included subjects was 37 (range, 19–76),
with 24.4% of patients scoring ≥45 indicating high overall AE burden.
Independent predictors for LAEP scores ≥ 45 were female sex, lack
of 12-month terminal seizure remission, drug resistance, and partial
epilepsy (Table S1), and predictors for LAEP score ≥ 37 (median) were
female sex, lack of 12-month terminal seizure remission, and drug resis-
tance (data not shown).
Regarding antiepileptic treatment regimen, 507 patients were on
monotherapy, 325 on polytherapy, and nine were untreated. The four
most common AEDs in monotherapy, each administered to more than
5% of monotherapy patients, comprised levetiracetam (n = 151),
lamotrigine (n = 167), valproic acid (n = 73), and CR carbamazepine
(n = 47) (for details on demographic and epilepsy variables in these
four monotherapy groups, please refer to Table S4). These 438 patients
were considered for further analysis.
Comparison of this subgroup to the remaining 69 monotherapy pa-
tients taking an AED other than one of the four most common sub-
stances did not demonstrate significant differences in regard to age
at investigation, duration of epilepsy, lack of 12-month terminal seizure
remission, and epilepsy syndrome (Table S4). In the 69 patients on AED
monotherapy not considered for further analysis, male sex (55.1%) and
drug resistance (24.6%) were significantly more common compared
with the 438 patients eventually included (39.0%, p = 0.017; and
14.4%, p = 0.049).
3.2. AED monotherapy
3.2.1. Predictors for adverse effects
Detailed data on demographic, epilepsy, and treatment variables of
the subset of patients treated with one of the four most common AEDs
in monotherapy are given in Table 1. Binary logistic regression identi-
fied duration of epilepsy, lack of 12-month terminal seizure remission,
and partial epilepsy as independent predictors for LAEP scores ≥ 45
(Table 1). Predictors for LAEP scores ≥ 34 (median) were female sex
(OR 2.045), lack of 12-month terminal seizure remission (OR 2.170),

Table 1
Predictors for high burden of adverse effects in patients treated with antiepileptic monotherapy.

Variables All LAEP b 45 LAEP ≥ 45 Binary logistic regression

(n = 438) (n = 347) (n = 91) Exp(B) [95% CI]

Sex
Female, n (%) 267 (61.0) 206 (59.4) 61 (67.0) n.s.
Male, n (%) 171 (39.0) 141 (40.6) 30 (33.0) 1.000
Age, years mean ± SD 44.7 ± 17.1 43.9 ± 17.1 47.5 ± 16.6 n.s.
Duration of epilepsy, years mean ± SD 17.2 ± 15.6 16.1 ± 14.8 21.5 ± 17.4 1.025 [1.010–1.040], p = 0.001
12-month term. seizure remission
Yes, n (%) 190 (43.4) 167 (48.1) 23 (25.3) 1.000
No, n (%) 248 (56.6) 180 (51.9) 68 (74.7) 2.788 [1.645–4.727], p b 0.001
Specific AED
Levetiracetam, n (%) 151 (34.5) 116 (33.4) 35 (38.5) n.s.
Lamotrigine, n (%) 167 (38.1) 129 (37.2) 38 (41.8) n.s.
Valproic acid, n (%) 73 (16.7) 64 (18.4) 9 (9.9) 1.000
CR carbamazepine, n (%) 47 (10.7) 38 (11.0) 9 (9.9) n.s.
Drug resistance
Yes, n (%) 63 (14.4) 42 (12.1) 21 (23.1) n.s.
No, n (%) 375 (85.6) 305 (87.9) 70 (76.9) 1.000
Epilepsy syndrome
Generalized, n (%) 112 (25.6) 98 (28.2) 14 (15.4) 1.000
Partial, n (%) 296 (67.6) 225 (64.8) 71 (78.0) 2.434 [1.284–4.614], p = 0.006
Unclassified, n (%) 30 (6.8) 24 (6.9) 6 (6.86) 2.316 [0.780–6.873], p = 0.130

LAEP — Liverpool Adverse Event Profile, n — number, SD — standard deviation, AED — antiepileptic drug, term. — terminal, CR — controlled-release, Exp(B) — odds ratio, CI — confidence interval,
n.s. — not significant (variable was removed before logistic regression reached the last step). P values less than 0.05 (bold emphasized) were considered statistically significant.
 A.B. Kowski et al. / Epilepsy & Behavior 54 (2016) 150–157 153

Fig. 2. Occurrence of specific adverse effects with individual antiepileptic drugs. Frequency (%) of occurrence of the 19 specific adverse effects (AEs) as defined by the Liverpool Adverse Event
Profile (from “always or often” as indicated by the darkest part of each horizontal bar on the left to “never” as indicated by the brightest part on the right). For each of the 19 AEs, frequencies
are given for patients treated with levetiracetam (LEV, n = 151), lamotrigine (LTG, n = 167), valproic acid (VPA, n = 73), or controlled-release carbamazepine (CR CBZ, n = 47).
154 A.B. Kowski et al. / Epilepsy & Behavior 54 (2016) 150–157

and partial epilepsy (OR 1.777) (detailed data not shown). None of the

Disturbed
four individual AEDs was independently associated with LAEP scores

1.026

1.626

2.372

1.000
sleep
≥ 45 or ≥ 34. Overall, the most frequent LAEP-defined AE reported by
more than one-third of the patients were sleepiness (47.7%), difficulty

problems
Sleepiness Depression Memory
concentrating (41.3%), tiredness (39.0%), and memory problems

1.014

2.418

1.918
1.751
1.000
(37.9%). Frequencies of specific AEs reported with the four most com-

Figures indicate odd ratios; for each adverse effect (AE), the antiepileptic drug (AED) exhibiting the lowest frequency of that AE was set to 1.000, the three other AEDs were compared with that AED; CR — controlled-release.
mon individual AEDs are demonstrated in Fig. 2.
In the next step, analysis for AE predictors was performed for each of

2.310

3.021

1.000
the 19 distinct AEs defined by the LAEP. The specific AEs were mainly
predicted by lack of 12-month terminal seizure remission (13 out of
19 AEs), treatment with one of the individual AEDs (7 out of 19 AEs),

1.854

2.291

1.000
partial epilepsy (6 out of 19 AEs), age and duration of epilepsy (each
5 out of 19 AEs), female sex (4 out of 19 AEs), and drug resistance

Dizziness
(1 out of 19 AEs) (Table 2). To assess the independent contribution of

1.019

1.894

3.169
1.000
individual AEDs to frequent occurrence of specific AEs, logistic regres-
sion analysis was performed. The analysis demonstrated that levetirac-

Trouble Shaky Weight

1.779

1.000
etam was independently associated most often with specific LAEP items

hands gain
(5 out of 19 AEs), comprising feelings of anger/aggression (OR 7.303),
nervousness/agitation (OR 5.364), upset stomach (OR 5.406), depres-

1.784

4.678

2.178
1.000
sion (OR 3.021), and disturbed sleep (OR 2.372). Lamotrigine (3 out
of 19 AEs) was independently associated with nervousness/agitation

mouth

1.000
with

gum
(OR 4.368), upset stomach (OR 5.434), and difficulty concentrating

and
(OR 2.610), and valproic acid (2 out of 19 AEs) with upset stomach

concentrating
(OR 5.909) and shaky hands (OR 4.678). Controlled-release carbamaze-

Difficulty
pine had no significant association with any of the 19 AEs. Further de-

1.017

3.204

2.610
1.000
tails are given in Table S2.

3.2.2. Influence of antiepileptic drug doses stomach

Upset

5.406
5.434
5.909
1.000
Patients taking levetiracetam were on a mean daily dose of 1530 ±
795 mg [SDD 1.03 ± 0.52]. Corresponding doses for lamotrigine were
Problems Double/blurred

293 ± 182 mg [SDD 0.98 ± 0.60], for valproic acid 1020 ± 450 mg
[SDD 0.86 ± 0.30], and for CR carbamazepine 700 ± 349 mg [SDD
with skin vision

1.998

2.570
1.000
0.72 ± 0.34].
Levetiracetam mean daily doses as well as SDD were significantly
higher in patients reporting frequent occurrence of feelings of anger/
Independent predictors for occurrence of specific adverse effects in patients treated with antiepileptic monotherapy.

1.000
aggression (p = 0.002, p = 0.002), of nervousness/agitation (p =
0.023, p = 0.029), of depression (p = 0.037, p = 0.044), and of dis-
0.982
1.018

1.000
Headache Hair

turbed sleep (p = 0.022, p = 0.028) compared with patients with

loss

infrequent occurrence of these specific AEs. At most, a trend difference

in mean daily dose and a significant difference in SDD were seen in pa-
2.606
0.986

1.641

1.770
2.898
1.000

tients with lamotrigine regarding occurrence of difficulty with concen-

tration (p = 0.054, p = 0.039). For more details see Table 3.
Restlessness Feelings of Nervousness/
agitation

3.2.3. Overall health status and quality of life

5.364
4.368

1.000

Median score for both overall health status and QoL on an 11-score
scale was 7. There was a significant negative correlation between
aggression

total LAEP score and scores for overall health status (r = −0.553, p b
anger/

7.303

1.000

0.001) and QoL (r = −0.537, p b 0.001) (Fig. 3).

Compared with 12-month terminal seizure remission (=1.000).

Binary logistic regression identified age at interview, lack of 12-

month terminal seizure remission, and LAEP scores ≥45 as independent
predictors for poor overall health status and low QoL (Table S3).
2.001

1.000

Compared with generalized epilepsy (=1.000).

4. Discussion
Compared with no drug resistance (=1.000).
Unsteadiness Tiredness

1.982
1.015

1.989

1.000

In our cohort of consecutive outpatients with epilepsy, we systemat-

Compared with male sex (=1.000).

ically screened as part of the clinical routine for AEs using the LAEP, a
well-established self-report assessment instrument. Our German trans-
lation of the LAEP was validated in a sufficiently large subset of 120 of
2.118
1.031
1.032

2.141

3.555
4.922
1.000

the current patients and proved to be as reliable as the English original

[12] and as previous translations to various other languages [8,18,19].
seizure remissionb

In the total study population of 841 patients, we identified female

CR carbmazepine
Lack of 12-month

Partial epilepsyd
Levetiracetam

Drug resistancec

sex, lack of 12-month terminal seizure remission, drug resistance, and

Valproic acid
Lamotrigine
Antiepileptic

Unclassified
Female sexa

partial epilepsy to independently predict LAEP scores of 45 and more

epilepsyd
Duration of
epilepsy

indicating high AE burden and, thus, potential “toxicity” of AEDs [2].

drugs
Table 2

These findings are consistent with previous studies [5–8] indicating

Age

d
b
a

the representativeness of our patient sample.

 A.B. Kowski et al. / Epilepsy & Behavior 54 (2016) 150–157 155

Table 3
Antiepileptic drug doses and specific adverse effects.

Feelings of anger/aggression Nervousness/agitation Upset stomach

Frequent Infrequent ANOVA — p Frequent Infrequent ANOVA — p Frequent Infrequent ANOVA — p

LEV
mg 1882 ± 908 1416 ± 722 0.002 1735 ± 894 1426 ± 722 0.023 1535 ± 945 1509 ± 764 0.471
SDD 1.25 ± 0.61 0.95 ± 0.48 0.002 1.17 ± 0.62 0.96 ± 0.47 0.029 1.09 ± 0.63 1.01 ± 0.50 0.511
LTG
mg 285 ± 68 297 ± 188 0.692 255 ± 142 301 ± 189 0.216
SDD 0.95 ± 0.56 1.00 ± 0.62 0.630 0.85 ± 0.47 1.01 ± 0.63 0.193
VPA
mg 979 ± 272 1030 ± 485 0.703
SDD 0.65 ± 0.18 0.67 ± 0.32 0.703

LEV — levetiracetam, LTG — lamotrigine, VPA — valproic acid, SDD — standardized daily dosage. P values less than 0.05 (bold emphasized) were considered statistically significant.

The main aim of the current study was to determine the contribution
of individual AEDs to overall AE burden and to specific AEs defined by
the LAEP. To achieve this, we focused on the large subgroup of patients
with AED monotherapy allowing for an in-depth analysis of specific pre-
dictors down to the level of distinct AEDs. Previous studies including
patients receiving AED monotherapy treatment for either new-onset
epilepsy [5], controlled epilepsy [11], or epilepsy of different degrees
of severity [10] did not report differences of individual AEDs in regard
to overall or specific AE burden. This finding was confirmed by the cur-
rent total study population and by the subset of the current monother-
apy patients.
For the first time, we were able to demonstrate that distinct com-
monly administered AEDs in monotherapy are associated with specific
AEs, independently from important potentially confounding factors
such as age, sex, epilepsy duration and syndrome, and response to anti-
convulsant treatment.
Levetiracetam was significantly associated with five AEs out of
which three — anger/aggression, nervousness/agitation, and depres-
sion — were confirmative [20–24]. Sleep disturbances so far have been
reported only occasionally [21,25] and probably are not in the focus of
physicians' attention. In the regulatory monotherapy trial, however, in-
somnia was reported by patients treated with levetiracetam more than
twice as often as by patients on CR carbamazepine [26]. Specific sleep
assessments using polysomnography before and after treatment onset
with levetiracetam demonstrated worsening in patients with epilepsy
[27] contrasted with improvement in healthy controls [28].
The most surprising finding in the current study was the significant
and independent association between the use of lamotrigine and
difficulties concentrating. This AE is rarely, if at all, reported in random-
ized controlled monotherapy trials including lamotrigine [29,30]. The
U.K. trials on standard and new antiepileptic drugs (SANAD) in general-
ized [31] and partial epilepsies [32] reported confusion/difficulties
thinking/disorientation, which is a broader and probably more severe
AE compared with concentration difficulties, in 1.3% and 2.1% of patients
taking lamotrigine, respectively, without differences compared with
valproic acid (1.3%) and carbamazepine (2.4%). However, the current
unequivocal finding should urge physicians to specifically ask patients
taking lamotrigine about problems with concentration. Agitation/ner-
vousness with use of lamotrigine has not been indicated in randomized
controlled trials [29,30], and the SANAD data reporting the broader AE
category anxiety/agitation/nervousness are ambiguous in regard to
this AED [31,32].
The significant impact of valproic acid on frequent occurrence of
hand tremor (32% of our patients) is plausible. In randomized controlled
trials including valproic acid as treatment for patients with newly diag-
nosed epilepsy, tremor was reported spontaneously by patients most
often with this AED [24,33].
Controlled-release carbamazepine was not associated with any of
the 19 LAEP-defined specific AEs. Patients on carbamazepine had a sig-
nificantly longer duration of epilepsy (23.9 ± 16.6 years) than patients
on levetiracetam (13.0 ± 13.3 years, p b 0.001) and on lamotrigine
(16.1 ± 20.8 years, p = 0.003). One might hypothesize that those pa-
tients that had developed clinically relevant AEs towards carbamaze-
pine had stopped the drug earlier in the course of their disease. This
may also explain why rather unspecific AEs such as abdominal distur-
bances are the least likely associated with carbamazepine.
Interestingly, most of the current specific AEs predicted by an AED
did not occur in a dose-dependent manner which, however, is what
we would expect for type A adverse effects [34]. This most common
type of AE is related to the known drug's mechanism of action; manifes-
tation is acute, predictable, dependent on dose or serum concentration,
and reversible [32]. The lack of demonstrated dose-dependency may be
due to the still too small number of patients treated with each of the
four AEDs in a wide range of doses.
Our 11-score rating scales on overall health status and QoL unequiv-
ocally demonstrated that higher AE burden significantly correlates with
less favorable self-assessed health status and QoL. This finding strongly
supports previous data based on the LAEP and QOLIE-89 [2]. That
study had also demonstrated that reduction of overall LAEP score by
optimization of AED treatment results in significant improvement in
QoL. Though our data were not able to demonstrate general dose-
dependency of AEDs for occurrence of specific AE due to too small

Difficulty concentrating Shaky hand Depression Disturbed sleep

Frequent Infrequent ANOVA — p Frequent Infrequent ANOVA — p Frequent Infrequent ANOVA — p Frequent Infrequent ANOVA — p

LEV
mg 1771 ± 928 1455 ± 736 0.037 1718 ± 887 1413 ± 711 0.022
SDD 1.18 ± 0.60 0.98 ± 0.48 0.044 1.14 ± 0.59 0.95 ± 0.47 0.028
LTG
mg 319 ± 204 265 ± 151 0.054
SDD 1.08 ± 0.67 0.88 ± 0.50 0.039
VPA
mg 1118 ± 418 977 ± 461 0.223
SDD 0.75 ± 0.28 0.65 ± 0.31 0.223
156 A.B. Kowski et al. / Epilepsy & Behavior 54 (2016) 150–157
Fig. 3. Correlation between LAEP score and overall health status or quality of life. As indicated by the linear regression curves (thick lines), the Liverpool Adverse Event Profile (LAEP) scores
negatively correlate with (A) self-reported overall health status (Pearson correlation r = −0.553, p b 0.001) and (B) self-reported quality of life (Pearson correlation r = −0.537,
p b 0.001). The thin lines illustrate the 95% confidence interval. The number of patients at an individual score for health status and quality of life, respectively, and the accordant LAEP score
are indicated by the size of the circle.
numbers of patients, we speculate that targeting a specific AE by AED
dose reduction (if possible in regard to seizure control) or, even better,
by AED change to another substance presumably not associated with
that AE would result in improved QoL and, thus, better adherence.
This study has several limitations; the first and main of which is the
nonrandomized study approach that bears the risk of a selection bias, as
patients with distinct characteristics may not have been treated with
specific AEDs. However, half of the reported AEs are confirmative for
specific AEDs arguing against a general selection bias. Second,
the monocentric assessment in a tertiary epilepsy outpatient clinic,
which is specialized on more complex cases, may limit generalizability
of our results to the broad spectrum of patients with epilepsy in the
community. Third, the study lacks follow-up data, which would allow
for evaluating the effect of interventions on specific AEs, e.g., dose re-
duction or change to another AED. Fourth, though we included a large
group of patients with AED monotherapy, we were only able to focus
on the four most common AEDs administered in our cohort. Already
for the fifth most frequently administered AED, oxcarbazepine (24 out
of 507 patients, 4.7%), the number of patients was too small to allow
for proper statistical analysis. Fifth, although depression has been
shown previously to be a major predictor of high AE burden [5] and to
specifically increase reporting incidence of almost all AEs in the LAEP
[35], we were not able to consider depression as a further clinical vari-
able. This is due to the lack of reliable data as we used a cross-
sectional study approach and retrieved information from documenta-
tion of clinical routine procedures. Finally, in clinical practice, we advise
our patients on common AEs and focus on those which are likely to be
expected with individual AEDs. We cannot exclude that this in part con-
tributes to the unequivocal confirmation of specific AEs with distinct
AEDs. On the other hand, this practice would further strengthen our
finding that lamotrigine is independently associated with concentration
difficulties as we commonly advise our patients that this particular AED
does not significantly impair cognitive function.
In patients who require pharmacological treatment for epilepsy,
systematic screening for AEs seems to allow for identification of patients
at risk for general high AE burden and for occurrence of specific AEs
which are due to individual AEDs and independent from other common
confounding clinical variables. In these patients, optimization of AED
treatment may enable clinicians to decrease the general burden of AEs
and occurrence of specific AEs, to increase QoL, and to decrease the
risk of nonadherence and premature treatment failure. Future analyses
should focus on patients with AED polytherapy, which are expected to
bear an even higher burden of AEs.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.yebeh.2015.11.009.
Acknowledgment
We thank Henriette Kiep for collecting data for the Patient Health
Questionnaire for Depression and the Quality-of-Life Inventory in a sub-
set of the current patients for validation of the German version of the
Liverpool Adverse Event Profile.
Disclosure of conflicts of interest
We confirm that we have read the Journal's position on issues in-
volved in ethical publication and affirm that this report is consistent
with those guidelines.
Alexander B. Kowski was supported by a stipendium from the “Frie-
drich C. Luft” Clinical Scientist Pilot Program funded by Volkswagen
Foundation and Charité-Universitätsmedizin Berlin Foundation. Florian
Weissinger received speaker's honoraria and/or consultancy fees from
Eisai and UCB. Verena Gaus received speaker's honoraria from Eisai
and UCB. Pawel Fidzinski has no disclosures. Florian Losch received
speaker's honoraria and/or consultancy fees from Eisai. Martin
Holtkamp holds the “Friedrich-von-Bodelschwingh endowed Professor-
ship for Clinical and Experimental Epileptology” at the Charité —
Universitätsmedizin Berlin funded by von Bodelschwingh Foundation.
He received speaker's honoraria and/or consultancy fees from
Cyberonics, Desitin, Eisai, GlaxoSmithKline, Janssen-Cilag, UCB, and
Viropharma.
References
[1] Baker GA, Jacoby A, Buck D, Stalgis C, Monnet D. Quality of life of people with epilep-
sy: a European study. Epilepsia 1997;38:353–62.
[2] Gilliam FG, Fessler AJ, Baker G, Vahle V, Carter J, Attarian H. Systematic screening
allows reduction of adverse antiepileptic drug effects: a randomized trial. Neurology
2004;62:23–7.
[3] Perucca P, Carter J, Vahle V, Gilliam FG. Adverse antiepileptic drug effects: toward
a clinically and neurobiologically relevant taxonomy. Neurology 2009;72:1223–9.
[4] Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;
342:314–9.
[5] Perucca P, et al. Adverse antiepileptic drug effects in new-onset seizures: a case–
control study. Neurology 2011;76:273–9.
[6] Andrew T, Milinis K, Baker G, Wieshmann U. Self reported adverse effects of mono
and polytherapy for epilepsy. Seizure 2012;21:610–3.
[7] Canevini MP, et al. Relationship between adverse effects of antiepileptic drugs, num-
ber of coprescribed drugs, and drug load in a large cohort of consecutive patients
with drug-refractory epilepsy. Epilepsia 2010;51:797–804.
[8] Martins HH, et al. Are adverse effects of antiepileptic drugs different in symptomatic
partial and idiopathic generalized epilepsies? The Portuguese–Brazilian validation of
the Liverpool Adverse Events Profile. Epilepsy Behav 2011;22:511–7.
[9] Sake JK, et al. A pooled analysis of lacosamide clinical trial data grouped by mecha-
nism of action of concomitant antiepileptic drugs. CNS Drugs 2010;24:1055–68.
 A.B. Kowski et al. / Epilepsy & Behavior 54 (2016) 150–157
[10] Wieshmann UC, Tan GM, Baker G. Self-reported symptoms in patients on antiepilep-
tic drugs in monotherapy. Acta Neurol Scand 2011;124:355–8.
[11] Kwon OY, Park SP. What is the role of depressive symptoms among other predictors
of quality of life in people with well-controlled epilepsy on monotherapy? Epilepsy
Behav 2011;20:528–32.
[12] Baker GA, Jacoby A, Smith DF, Dewey ME, Chadwick DW. Development of a novel
scale to assess life fulfillment as part of the further refinement of a quality-of-life
model for epilepsy. Epilepsia 1994;35:591–6.
[13] Fisher RS, et al. ILAE official report: a practical clinical definition of epilepsy. Epilepsia
2014;55:475–82.
[14] Kwan P, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc
Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51:
1069–77.
[15] Gräfe K, Zipfel S, Herzog W, Löwe B. Screening psychischer Störungen mit dem
“Gesundheitsfragebogen für Patienten (PHQ-D)”. Diagnostica 2004;50:171–81.
[16] May TW, Pfäfflin M, Cramer JA. Psychometric properties of the German translation of
the QOLIE-31. Epilepsy Behav 2001;2:106–14.
[17] Wissenschaftlichen Institut der AOK (WIdO) des AOK-Bundesverbandes GbR,
Deutsches Institut für Medizinische Dokumentation und Information. Anatomisch-
therapeutisch-chemische Klassifikation mit Tagesdosen. https://www.dimdi.de/
dynamic/de/klassi/downloadcenter/atcddd/version2015/. [Accessed 26-Jul-2015].
[18] Carreno M, et al. Validation of the Spanish version of the Liverpool Adverse Events
Profile in patients with epilepsy. Epilepsy Behav 2009;15:154–9.
[19] Chen HF, Tsai YF, Shih MS, Chen JC. Validation of the Chinese version of the
Liverpool Adverse Events Profile in patients with epilepsy. Epilepsy Res 2011;94:
45–52.
[20] Mula M, Trimble MR, Yuen A, Liu RS, Sander JW. Psychiatric adverse events during
levetiracetam therapy. Neurology 2003;61:704–6.
[21] Helmstaedter C, Witt JA. The effects of levetiracetam on cognition: a non-
interventional surveillance study. Epilepsy Behav 2008;13:642–9.
[22] Lee JJ, Song HS, Hwang YH, Lee HW, Suh CK, Park SP. Psychiatric symptoms and
quality of life in patients with drug-refractory epilepsy receiving adjunctive leveti-
racetam therapy. J Clin Neurol 2011;7:128–36.
[23] Wieshmann UC, Baker GA. Self-reported feelings of anger and aggression towards
others in patients on levetiracetam: data from the UK antiepileptic drug register.
BMJ Open 2013;3.
157
[24] Trinka E, et al. KOMET: an unblinded, randomised, two parallel-group, stratified trial
comparing the effectiveness of levetiracetam with controlled-release carbamaze-
pine and extended-release sodium valproate as monotherapy in patients with
newly diagnosed epilepsy. J Neurol Neurosurg Psychiatry 2013;84:1138–47.
[25] Mbizvo GK, Dixon P, Hutton JL, Marson AG. The adverse effects profile of levetirace-
tam in epilepsy: a more detailed look. Int J Neurosci 2014;124:627–34.
[26] Brodie MJ, Perucca E, Ryvlin P, Ben-Menachem E, Meencke HJ. Comparison of
levetiracetam and controlled-release carbamazepine in newly diagnosed epilepsy.
Neurology 2007;68:402–8.
[27] Zhou JY, Tang XD, Huang LL, Zhong ZQ, Lei F, Zhou D. The acute effects of levetirac-
etam on nocturnal sleep and daytime sleepiness in patients with partial epilepsy.
J Clin Neurosci 2012;19:956–60.
[28] Cicolin A, Magliola U, Giordano A, Terreni A, Bucca C, Mutani R. Effects of levetirac-
etam on nocturnal sleep and daytime vigilance in healthy volunteers. Epilepsia
2006;47:82–5.
[29] Kwan P, Brodie MJ, Kalviainen R, Yurkewicz L, Weaver J, Knapp LE. Efficacy and
safety of pregabalin versus lamotrigine in patients with newly diagnosed partial
seizures: a phase 3, double-blind, randomised, parallel-group trial. Lancet Neurol
2011;10:881–90.
[30] Rosenow F, et al. The LaLiMo trial: lamotrigine compared with levetiracetam in the
initial 26 weeks of monotherapy for focal and generalised epilepsy — an open-label,
prospective, randomised controlled multicenter study. J Neurol Neurosurg
Psychiatry 2012;83:1093–8.
[31] Marson AG, 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.
[32] Marson AG, et al. The SANAD study of effectiveness of carbamazepine, gabapentin,
lamotrigine, oxcarbazepine, or topiramate for treatment of partial epilepsy: an un-
blinded randomised controlled trial. Lancet 2007;369:1000–15.
[33] Steinhoff BJ, Ueberall MA, Siemes H, Kurlemann G, Schmitz B, Bergmann L. 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.
[34] Perucca P, Gilliam FG. Adverse effects of antiepileptic drugs. Lancet Neurol 2012;11:
792–802.
[35] Panelli RJ, Kilpatrick C, Moore SM, Matkovic Z, D'Souza WJ, O'Brien TJ. The Liverpool
Adverse Events Profile: relation to AED use and mood. Epilepsia 2007;48:456–63.