If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author at: Mainz Comprehensive Epilepsy and Sleep Medicine Center, Department of Neurology, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstr.1, Mainz 55131, Germany.
Mainz Comprehensive Epilepsy and Sleep Medicine Center, Department of Neurology, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstr.1, Mainz 55131, GermanyDepartment of Neurology, Philipps-University Marburg, Baldingerstr, Marburg 35043, Germany
Department of Neurology, Focus Program Translational Neuroscience (FTN), Rhine Main Neuroscience Network, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstr.1, Mainz 55131, Germany
Mainz Comprehensive Epilepsy and Sleep Medicine Center, Department of Neurology, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstr.1, Mainz 55131, Germany
Department of Neurology, Focus Program Translational Neuroscience (FTN), Rhine Main Neuroscience Network, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstr.1, Mainz 55131, Germany
Specific antiseizure medications (ASM) would improve the outcome in post-stroke epilepsy (PSE). The aim of this multicenter observational study was to compare different antiseizure monotherapies in PSE.
Methods
We collected the data from 207 patients with PSE who did not change their initial antiseizure monotherapy during the period of 12 months. Efficacy was assessed by a standardized three month seizure frequency and seizure freedom. Safety was estimated by the reported side effects.
Results
The mean three month seizure frequency was 1.9 ± 3.1 on eslicarbazepine, 2.1 ± 3.2 on lacosamide, 3.4 ± 4.4 on levetiracetam, 4.3 ± 6.8 on lamotrigine, and 5.1 ± 7.3 on valproate (p < 0.05 for eslicarbazepine or lacosamide in comparison with levetiracetam, lamotrigine and valproate, respectively). The lowest seizure frequency and the highest seizure freedom was observed on ASMs acting via the slow inactivation of sodium channels in comparison to other mechanisms of action (0.7 ± 0.9 vs 2.2 ± 2.4, p < 0.01). Among side effects, the most frequently reported were vertigo (25%) and tiredness (15.9%). They were similar in all investigated groups of ASM. The independent factors increasing seizure frequency that were identified in multiple regression analyses were increased size of infarction, cortical involvement, hemorrhagic transformation, neurological deficits at admission and functional impairment. Administration of ASM with the mechanism of action via the slow inactivation of sodium channels was an independent factor decreasing the seizure frequency.
Conclusion
Our data show that antiseizure medications acting via the slow inactivation of sodium channels, such as lacosamide and eslicarbazepine, are well tolerated and might be associated with better seizure control in PSE.
1. Introduction
With advances in stroke treatment, patient mortality has dropped dramatically worldwide [
]. A highly important clinical issue that can occur in stroke survivors is epilepsy. Post-stroke epilepsy (PSE) has an incidence rate of around 7% as shown in longitudinal cohort studies with over 100,000 patients [
None of the currently available antiseizure medications (ASMs) were shown to have a preventive potential for epilepsy development after a stroke. Primary medical prevention is, therefore, not recommended [
]. However, treatment with ASM, as a secondary prevention, should be considered after a first unprovoked seizure in patients with a high risk of recurrence [
]. Despite the high incidence of post-stroke epilepsy and changing population dynamics, no recommendations for the choice of anticonvulsants for the treatment of post-stroke seizures have been published to date. Although cerebrovascular disease is one of the most common causes of acquired epilepsy in adults and the elderly, only a few randomised controlled trials have specifically examined ASM in post-stroke seizures. The trials did not find a preferable substance for the treatment [
]. In our study, we analysed the data of two large neurological registers to compare seizure control and safety of different ASMs in the treatment of PSE.
2. Methods
2.1 Study design
Data on all patients who initiated antiseizure monotherapy after the diagnosis of PSE were collected and analysed within the Mainz Epilepsy Register (MAINZ-EPIREG), and the Marburg Stroke Register (MARSTREG). MAINZ-EPIREG is a population-based register of patients with epilepsy who were treated in the Mainz Comprehensive Epilepsy and Sleep Medicine Center (reference area of 4.000.000 inhabitants) in cooperation with two specialized office-based epilepsy practices (Ingelheim am Rhein and Nierstein, Germany). MARSTREG is a population-based stroke register recruiting all patients with acute ischemic stroke, who were permanent residents in the district Marburg-Biedenkopf (Hessen, Germany, reference population 240.000 inhabitants). In both registers, patients received continuous follow-ups with intervals of six months. According to the practical definition of the International League Against Epilepsy (ILAE), we diagnosed structural epilepsy after stroke in patients with at least one late-unprovoked seizure [
]. The data were collected on case report forms and included demographics, seizure frequency, drug treatment, severity of stroke (NIH Stroke Scale, Barthel Index, modified Rankin Scale) and depression (Geriatric Depression Scale). The follow-ups were performed at the time points 6 and 12 months after the initiation of ASM. Seizure freedom was defined as the occurrence of no seizures between the follow-ups at 6 and 12 months, i.e. during the last six months of the study period. The study received approval from the local ethic committees (reference numbers: 135/11 from 03/05/2014 and 837.560.17/11,394 from 01/19/2018).
2.2 Statistical analysis
We used IBM SPSS Statistics Version 23.0 (IBM Corp., Armonk, NY, USA) for statistical analysis. The Kolmogorov-Smirnov test was used to prove for normal distributions. The comparisons of normally distributed variables were performed using the t-test. In the case of not normally distributed data, the Mann-Whitney U test (two independent groups), the Kruskal-Wallis test (more than two independent groups), or the Wilcoxon rank test (two dependent groups) were applied. The plan of statistical analysis was prespecified according to single ASMs and according to presented classification of mechanism of action.
We performed a multivariate regression analysis to identify independent factors determining seizure frequency. The variability accounted for by independent factors of seizure frequency was estimated using the R2 method.
3. Results
We registered 216 patients, in which antiseizure monotherapy was initiated because of PSE. At the 12-months follow-up, 3.5% of patients (n = 9) were on combination antiseizure therapy and therefore excluded from the analysis. Two hundred and seven patients who stayed on monotherapy did not change their initial ASM during the whole observation period of 12 months (retention rate: 82.8%). The retention rates of most commonly initially administered ASMs were 91.7% for lacosamide (LCM), 90.9% for eslicarbazepine (ESL), 82.0% for lamotrigine (LTG), 77.8% for levetiracetam (LEV), and 73.7% for others (p < 0.05 for comparisons LCM vs LEV or others and ESL vs LEV or others).
In order to homogenize the data, our further analysis focused on patients who did not change their initial monotherapy during the first 12 months since initiation of ASM. The clinical characteristics of these patients are provided in Table 1. In this cohort, 29.5% of patients (N = 61) were treated with LEV, 18.4% (N = 38) with ESL, 20.3% (N = 42) with LCM, 19.3% (N = 40) with LTG, 8.7% (N = 18) with valproate (VPA) (Fig. 1). The remaining 3.8% (N = 8) of patients received topiramate (TPM, N = 3), zonisamide (ZNS, N = 2), gabapentin (GBP, N = 2) and carbamazepine (CBZ, N = 1).
Table 1Clinical parameters of patients with post-stroke epilepsy.
Total (N = 207)
LEV (N = 61)
LCM (N = 42)
LTG (N = 40)
ESL (N = 38)
Others (N = 26)
Age, mean±SD
68±8.2
69±8.1
67±9.7
68±8.0
68±7.8
69±7.2
Gender, n (%)
male
121 (58.5)
37 (60.7)
25 (59.5)
24 (60.0)
21 (55.3)
14 (53.8)
female
86 (41.5)
24 (39.3)
17 (40.5)
16 (40.0)
17 (44.7)
12 (46.2)
Size of infarction (cm3), mean±SD
65.9 ± 29.0
65.8 ± 28.3
70.0 ± 31.1
63.8 ± 28.72
65.3 ± 26.9
64.6 ± 32.0
Cortical involvement, n (%)
164 (79.2)
47 (77.0)
34 (81.0)
32 (80.0)
31 (81.6)
20 (76.9)
Hemorrhagic transformation, n (%)
78 (37.7)
23 (37.7)
15 (35.7)
16 (40.0)
14 (36.8)
10 (38.5)
Acute symptomatic seizures, n (%)
7 (3.4)
2 (3.3)
2 (4.8)
1 (2.5)
1 (2.6)
1 (3.8)
Time from stroke to PSE*, mean±SD
19.1 ± 12.1
19.1 ± 11.6
16.8 ± 11.0
19.7 ± 9.4
19.4 ± 13.9
21.7 ± 15.8
Dosis (mg/d), mode (min-max)
n.a.
2000 (1000–4000)
200 (200–600)
400 (150–700)
800 (400–1600)
n.a.
Barthel Index, mean±SD
discharge
65.75±22.75
62.62±25.67
63.21±21.41
68.25±23.00
67.89±19.92
70.19±21.19
12 months
72.18±20.56
69.03±19.85
70.71±22.27
73.13±21.53
75.66±17.60
75.38±22.00
NIHSS, mean±SD
admission
8.1 ± 6.6
8.3 ± 6.6
8.9 ± 6.4
7.7 ± 6.8
7.9 ± 6.7
7.2 ± 6.3
12 months
5.0 ± 4.3
5.3 ± 3.7
5.4 ± 4.7
4.6 ± 4.4
4.6 ± 4.5
4.5 ± 4.7
No statistically significant differences in clinical parameters were found between different groups of ASMs.
Abbreviations: SD, standard deviation; PSE, post-stroke epilepsy; n.a., not applicable; LEV, levetiracetam; LCM, lacosamide; LTG, lamotrigine; ESL, eslicarbazepine.
*Time from the event of ischemic stroke till the development of PSE expressed in months.
Considering the most often administered ASMs, the mean three months seizure frequency at the 12-month follow-up was 0.7 ± 1.0 (standard deviation, SD) on LCM, 0.7 ± 0.9 on ESL, 1.5 ± 1.7 on LTG, 2.1 ± 2.4 on LEV, and 2.4 ± 2.7 on VPA (p < 0.01 for comparison LCM vs LTG, LCM vs LEV, LCM vs VPA, ESL vs LEV, ESL vs VPA and p < 0.05 for comparison, ESL vs LTG). In Fig. 2, the comparison of seizure frequency according to the mechanism of action of different ASMs is provided. The lowest seizure frequency was observed on ASMs with selectivity to the slow-inactivated state of sodium channels (0.7 ± 0.9 seizures during three months in comparison to other mechanisms of action, p < 0.01). Seizure freedom was achieved by LCM and ESL with 44% each and to a lesser extent by LTG (25%), LEV (23%), and VPA (11%) (p < 0.05).
Fig. 2Three months seizure frequency according to the mechanism of action of different ASMs
Safety profiles of ASMs are presented in Table 2. Among side effects, the most frequently were vertigo (25.1%), tiredness (15.9%) and headache (13.0%). Post-stroke depression was detected in patients on LEV, VPA and ZNS. Other ASMs did not show association with depression. Insomnia was reported by 17.5% of patients on LTG. Weight gain was observed in 44.4% of patients on VPA and in 50% of patients on gabapentin. Weight loss was reported in all patients on topiramate. Hyponatremia was detected in patients on eslicarbazepine, lamotrigine, valproate and levetiracetam. Hyponatremia was most frequent in patients on eslicarbazepine (10.5%); however, it did not fall below 129 mmol/L and was clinically asymptomatic.
Table 2Comparison of safety profiles of ASMs in post-stroke epilepsy.
Total n (%)
LEV n (%)
LCM n (%)
LTG n (%)
ESL n (%)
VPA n (%)
Others n (%)
Vertigo
52 (25.1)
14 (23)
11 (26.2)
10 (25)
9 (23.7)
6 (33.3)
2 (25)
Tiredness
33 (15.9)
9 (14.8)
7 (16.7)
6 (15)
6 (15.8)
4 (22.2)
1 (12.5)
Headache
27 (13.0)
8 (13.1)
5 (11.9)
5 (12.5)
5 (13.2)
3 (16.7)
1 (12.5)
Depression
26 (12.6)
17 (27.9)*
5 (11.9)
0 (0)
0 (0)
3 (16.7)
1 (12.5)
Irritability
10 (4.8)
10 (16.4)*
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
Insomnia
14 (6.8)
2 (3.3)
1 (2.4)
7 (17.5)
1 (2.6)
2 (11.1)
1 (12.5)
Nausea
11 (5.3)
1 (1.6)
2 (4.8)
3 (7.5)
3 (7.9)
1 (5.6)
1 (12.5)
Weight gain
10 (4.8)
0 (0)
0 (0)
0 (0)
0 (0)
8 (44.4)*
2 (25)*
Vomiting
9 (4.3)
1 (1.6)
2 (4.8)
3 (7.5)
2 (5.3)
1 (5.6)
0 (0)
Hyponatremia
8 (3.9)
1 (1.6)
0 (0)
2 (5)
4 (10.5)*
1 (5.6)
0 (0)
Diplopia
4 (1.9)
0 (0)
1 (2.4)
2 (5)
1 (2.6)
0 (0)
0 (0)
Weight loss
3 (1.4)
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
3 (37.5)*
*statistically significant (p<0.05) in comparison to other ASM-groups.
The results of the multivariate regression analyses of factors determining the seizure frequency are shown in Table 3. Increased size of infarction, cortical involvement, hemorrhagic transformation, neurological deficits at admission (measured by NIHSS) and functional impairment at discharge (measured by Barthel Index) were identified as independent factors increasing seizure frequency. In addition, administration of ASM with selectivity to the slow-inactivated state of sodium channels was identified as a factor influencing the seizure frequency in PSE. The above mentioned clinical variables could explain 51.3% of the variance in the seizure frequency (adjusted R2).
Table 3Independent determinants of seizure frequency in multiple regression analysis.
Seizure frequency
B
95% CI
p-value
Female gender
0.33
−0.72; 1.38
0.54
Age
−0.02
−0.61; 0.06
0.94
Size of infarction (cm3)
0.04
0.02; 0.64
0.02
Cortical involvement
2.22
0.85; 3.60
0.02
Hemorrhagic transformation
2.23
0.90; 3.57
0.01
Barthel-Index
−0.06
−0.96; −0.03
<0.01
NIHSS at admission
−0.93
−0.20; 0.01
0.78
SV2A-agonists
−5.52
−12.22; 1.18
0.11
Fast inactivation SCB
−2.73
−9.20; 3.74
0.41
Slow inactivation SCB
−6.17
−11.85; −0.49
0.03
GABA-ergic ASMs
−0.70
−6.08; 4.68
0.80
Calcium channel blockers
1.92
−2.02; 5.85
0.34
Adjusted R2 *
0.513
*Total adjusted R2 for each model.
Abbreviations: B, regression coefficient; CI, confidence interval; SCB, sodium channel blockers; ASM, antiseizure medication; NIHSS, National Institutes of Health Stroke Scale.
We performed an analysis of two large neurological registers (MAINZ-EPIREG and MARSTREG) in order to estimate the seizure control and safety of different ASM in monotherapy of PSE. The administered ASMs were well tolerated by patients with PSE. The safety profile was characteristic to the data published in elderly patients [
Prescription patterns differ between regions and countries. While LEV, LTG, LCM and ESL were the most common ASMs in monotherapy of PSE, a recent registry-based study from Sweden showed a more frequent prescription of LEV and CBZ for treatment of PSE in this country [
Relative importance of clinical outcomes and safety risks of antiseizure medication monotherapy for patients and physicians: discrete choice experiment eliciting preferences in real-world study "VOTE".
Our analyses depicted similar effects for seizure freedom as a clinical variable in PSE treated with LEV or LTG. Interestingly, LCM and ESL outperformed other ASM, when addressing this variable. In a previous study, ESL already reported a better responder rate and significantly higher seizure freedom in patients with PSE compared to patients with other epilepsy [
]. In addition, another study found a higher seizure-free rate with LCM compared to CBZ therapy as well as a more favourable side effect profile for LCM in patients with cerebrovascular epilepsy [
]. Existing data has not yet compared LCM and ESL with other ASM in PSE. Our patients showed not only increased seizure freedom but also a reduced seizure frequency on LCM and ESL compared to other ASMs. This finding also depicts a possible important pathophysiological link, since both substances share a similar mechanism of action. Thus, our data primary stress the importance of ASMs acting on voltage-dependent sodium channels, most specifically the ASMs acting on the selective binding to the slow-inactivated state of sodium channels. This is in contrast to drugs like carbamazepine or phenytoin, which bind tightly to fast-inactivated states, or to drugs showing a different mechanistic way of action in the treatment of PSE.
The cell death that occurs in brain ischemia is not just a consequence of energy loss. Multiple electrical and chemical events occur at the site of ischemic brain lesion and its surrounding tissue in the acute phase of stroke as well as during the phase of gliotic scar formation. In acute stroke, these events include glutamate release, inflammation, neurotransmitter imbalance and electrophysiological instability with peri‑infarct depolarizations. Hypoxia leads to sodium pump failure with Na+influx and depolarization [
]. Sodium channel blockers inhibit the Na+influx, cell depolarization and release of glutamate. In vitro some sodium channel blockers demonstrated also protective effects for brain tissue in cerebral ischemia [
]. The depolarization due to the Na+influx leads to the activation of calcium channels. The fast Ca2+ influx can result in abnormal neuronal discharges which can clinically manifest in seizures. Therefore, the application of sodium channel blockers can theoretically be beneficial for specific treatment of ischemia-related seizures.
The physiological function of sodium channels is based on a rapid, voltage-dependent activation, which results in opening of the channel, as well as inactivation, which closes the channel pore until the hyperpolarization of the cell and makes the cell refractory to the firing [
]. Inactivation can be classified in two distinguishable kinetic components: fast inactivation and slow inactivation. In fast inactivation, which occurs within milliseconds, an inactivating particle closes the pore by so-called “hinged lid” mechanism. Slow inactivation is a much slower process, which has a more complex mechanism unrelated to an inactivation particle. Rather, the mechanism is based on structural changes of the channel pore. In addition, it regulates the cells’ excitability by modulating burst discharges. The conventional sodium channel blockers, such as phenytoin or carbamazepine, exert their mechanism of action by blocking the fast inactivation state of gated sodium channels [
Effectiveness of eslicarbazepine acetate in dependency of baseline anticonvulsant therapy: results from a German prospective multicenter clinical practice study.
Eslicarbazepine and the enhancement of slow inactivation of voltage-gated sodium channels: a comparison with carbamazepine, oxcarbazepine and lacosamide.
]. Given the theoretical benefit of sodium channel blockers in the treatment of post-stroke seizures and given the results of our observational study, which showed the clinical advantages of the ASMs with this mechanism of action (LCM and ESL), future clinical studies are encouraged to investigate these ASMs in PSE in the setting of randomized clinical trials (RCT).
Our study has a number of limitations. It is an observational study and the findings having a lower level of evidence than RCT. However, our data would encourage future RCTs to investigate the efficacy of ASMs with slow inactivation of sodium channels in PSE. A selection bias towards more severe cases cannot be excluded. Although, the Mainz Comprehensive Epilepsy and Sleep Medicine Center is an epilepsy reference center for epilepsy in a region with four million inhabitants, there could have been less severe cases in our referral region, which did not require neurological management of their disease course. Only patients with epilepsy after ischemic stroke were evaluated, therefore the effects of ASM in post-hemorrhagic epilepsy cannot be judged. Finally, we cannot exclude residual confounding by unmeasured variables in multiple regression analysis.
To conclude, our data show that antiseizure medication acting through the slow inactivation of sodium channels, such as LCM and ESL, is well tolerated and may be associated with a better seizure control in PSE. These findings support future research to identify the most effective ASM for this indication.
Funding
An unrestricted grant was received from UCB Pharma for medical writing assistance and statistical work. UCB was not involved in the study design; the collection, analysis, and interpretation of the data gathered; nor the writing of the study report or the manuscript. Data collection in Marburg was supported by a Research Grant of the University Medical Center Giessen and Marburg.
Declaration of Competing Interest
YW reports honoraria for educational presentations and consultations from Arvelle Therapeutics, Bayer AG, BIAL, Eisai, LivaNova, Novartis and UCB Pharma. SG received compensation for professional services from Abbott, Abbvie, Bial, Medtronic, UCB and Zambon; research grants from Abbott, Boston Scientific, MagVenture, German Research Council and German Ministry of Education and Health. TU reports personal fees from Merck Serono and Pfizer, grants from Else Kröner‐Fresenius Stiftung and the German Research Council. SK has received honoraria for presentations and teaching courses from UCB Pharma and Novartis Pharma and has been invited to congresses and educational meetings by UCB Pharma and Eisai Pharma. KS and SSB declare no conflict of interest.
References
Bartholome L.
Winter Y.
Quality of life and resilience of patients with juvenile stroke: a systematic review.
Relative importance of clinical outcomes and safety risks of antiseizure medication monotherapy for patients and physicians: discrete choice experiment eliciting preferences in real-world study "VOTE".
Effectiveness of eslicarbazepine acetate in dependency of baseline anticonvulsant therapy: results from a German prospective multicenter clinical practice study.
Eslicarbazepine and the enhancement of slow inactivation of voltage-gated sodium channels: a comparison with carbamazepine, oxcarbazepine and lacosamide.
00156-X&id=gr1.jpg){kind=link}
00156-X&id=gr2.jpg){kind=link}