Cerebrovascular disease is the leading cause of epilepsy in adults. Epilepsy secondary to acute ischemic stroke (AIS) is associated with worse outcomes. The prevalence varies in different studies. Risk factors for this condition have been described, some inconsistency.
ObjectiveTo determine the prevalence of Post-ischemic Stroke epilepsy (PSE) at a fourth-level hospital in Bogotá, Colombia.
MethodologyCross-sectional study included adults with a diagnosis of AIS by MRI or CT scan and epilepsy during 2014–2019. The prevalence of PSE (a seizure 7 days after AIS) was determined, and its characteristics were described. Statistical analysis was performed in Stata/SE 17.
ResultsA total of 465 patients were identified, of whom 343 did not present lost to follow-up. Eighteen (5.25%) developed epilepsy, and 83.3% were men. The most common seizure was focal to bilateral tonic–clonic component (53%). Patients who developed PSE had a higher average on the NIHSS scale (15.67 vs 6.25), a higher proportion of cortical involvement (88.89% vs 67.08%), cardioembolic etiology (38.89% vs 27.38%). The principal outcomes were hemorrhagic transformation (50% vs 9.54%), intravenous thrombolysis (44.44% vs 23.09%), and mechanical thrombectomy (22.22% vs 4.92%). The most common antiseizure drug was levetiracetam.
SignificanceThe prevalence of PSE was 5.25%. The patients who developed this complication were mostly men and presented a higher proportion of severity, cortical compromise, cardioembolic etiology, hemorrhagic transformation, and reperfusion treatments.
Epilepsy is a common neurological disease, with a significant proportion of disease burden worldwide, and ranks fourth after tension-type headache, migraine, and Alzheimer's disease.1,2 In addition, it accounts for 0.7% of disability-adjusted life years lost (DALYs) in men and women.1,3,4
The leading cause of epilepsy in adults is cerebrovascular disease; it is estimated that up to 50% of de novo epilepsy in adults 60 years or older have a history of stroke.5–8 In the context of stroke, seizures are classified as (i) provoked or acute symptomatic (ASS), which occur during the first 7 days, and (ii) unprovoked, which occur beyond the first week and condition a process of epileptogenesis, with a risk of recurrence >60% during the first 10 years.6,9 To diagnose PSE in these patients, the presence of an unprovoked seizure is sufficient.10
Epilepsy is a sequela of stroke associated with worse quality of life, a lower degree of functionality in the modified Rankin scale (mRankin), higher morbidity and mortality, and increased health costs.6,9,11–13 The prevalence of PSE has been challenging to estimate due to the different definitions among the studies; however, it is estimated in a range from 2.5% to 15%.14–18
Similarly, it has been challenging to establish the risk factors for epilepsy in patients who have suffered a stroke, due to highly variable and, in some cases, contradictory results.9,11,19,20 In the studies performed so far, a greater degree of severity in the National Institutes of Health Stroke Scale (NIHSS), a stroke of the cerebral cortex or middle cerebral artery (MCA), history of coronary artery disease (CAD), atrial fibrillation (AF), hypertension (HTN), and diabetes mellitus (DM) among others have been associated with a higher risk of epilepsy; however, no consensus has been established.6,9,12,14,21–23 In this cross-sectional study, the prevalence of PSE was calculated in a stroke center in Colombia between 2014 and 2019. Likewise, the associated risk factors were evaluated.
Methods and materialsDesign of the studyA retrospective cross-sectional study was conducted at a fourth-level hospital, in Bogotá, Colombia. Patients with a diagnosis of AIS admitted to the Stroke Clinical Care Center during the years 2014–2019 were considered. Patients over 18 years old with an AIS diagnosis by neuroimaging either computed tomography scan (CT) or magnetic resonance imaging (MRI) were included. We excluded those patients with a history of epilepsy or seizures before the ischemic stroke and patients with epileptogenic comorbidities such as neurological infections, history of neurosurgical interventions, severe cranioencephalic trauma, and venous sinus thrombosis. The protocol of this study was approved by the ethics committee of the corresponding institution.
Data collectionPatient medical records were reviewed to collect information on: demographic data; comorbidities; stroke characteristics including severity according to the NIHSS, stroke location (cortical involvement, laterality, and MCA) territory involvement, whether revascularization treatment was received and the type (endovenous thrombolysis and/or mechanical thrombectomy); whether there was the presence of early seizures (ASS) defined as those occurring in the first 7 days after the vascular event according to the International League Against Epilepsy (LICE), the number of seizures, the type of seizures (focal with or without awareness, focal onset to generalized tonic–clonic or status epilepticus), the treatment received and the follow-up time of each patient. The outcome evaluated was the development of PSE which, according to the LICE, is defined as the presence of at least one seizure after the first 7 days of the ischemic attack.
Statistical analysisThe demographic description of the population was carried out with measures of central tendency for continuous variables and frequency for categorical variables. The prevalence of PSE was calculated considering the described population. Statistical analysis was performed in Stata/SE 17 to evaluate the variables associated with the development of epilepsy in these patients, including those included in the SeLECT scale.
ResultsPopulationWe identified 465 patients with AIS diagnosis, of which 343 were included in the analysis due to insufficient follow-up of the other participants. The median observation time was 21.42 months.
Of the 343 patients, 18 (5.25%) patients presented PSE, of whom 83.3% were men. Among patients with epilepsy, 61.11% were overweight according to Body Mass Index (BMI), 27.78% had a normal weight and 11.11% were obese. Similarly, important comorbidities were as follows: HTN (61.11%), DM (22.2%), AF (22.2%), and CAD (11.1%). These percentages were like those found in patients who did not develop PSE, except for CAD (14.15%), AF (18.77%), and dementia (4.92%). The characteristics of these patients are listed in Table 1.
Population characteristics. Source: Authors.
| Total(n = 343) | With PSE(n = 18) | Without PSE(n = 325) | ||
|---|---|---|---|---|
| Population characteristics | ||||
| Average Age (SD) | 70.38(16.46) | 62.9 (16.51) | 70.8 (16.40) | |
| Sex | ||||
| n (%) | Men | 163 (47.52) | 15 (83.33) | 148 (45.53) |
| Women | 180 (52.47) | 3 (16.66) | 177 (54.46) | |
| Laterality | ||||
| n (%) | Right | 323 (94.16) | 18 (100) | 305 (93.8) |
| Left | 10 (2.91) | 0 | 10 (3.10) | |
| Ambidextrous | 0 | 0 | 0 | |
| Unknown | 10 (2.91) | 0 | 10 (3.10) | |
| BMI | ||||
| n (%) | Low weight | 3 (0.87) | 0 | 3 (0.92) |
| Normal | 169 (49.27) | 5 (27.78) | 164 (50.46) | |
| Overweight | 135 (39.35) | 11 (61.11) | 124 (38.15) | |
| Obesity | 36 (10.46) | 2 (11.11) | 34 (10.46) | |
| Comorbidities | ||||
| n (%) | Hypertension | 215 (62.68) | 11 (61.11) | 204 (62.77) |
| Diabetes | 69 (20.11) | 4 (22.22) | 65 (20) | |
| Coronary artery syndrome | 48 (13.99) | 2 (11.11) | 46 (14.15) | |
| Atrial fibrilation | 65 (18.95) | 4 (22.22) | 61 (18.77) | |
| Dementia | 16 (4.66) | 0 (0) | 16 (4.92) | |
NIHSS score was calculated in 326 patients who had sufficient information for the patients without epilepsy, the mean value of 6.77 (SD 6.55); the most frequent etiology was unknown/cryptogenic (39.06%), and the majority presented cortical involvement (68.22%) without hemorrhagic transformation (88.33%). On the other hand, in patients who developed PSE, the mean NIHSS was 15.67 points (SD 7.10), the most frequent etiology was cardioembolic (38.89%) with greater involvement of the right hemisphere (50%) and greater cortical involvement (88.9%) (Table 2).
Characteristics of acute ischemic stroke. Source: Authors.
| Characteristics of AIS | ||||
|---|---|---|---|---|
| Total(n = 343) | With PSE(n = 18) | Without PSE(n = 325) | ||
| NIHSS⁎ average (SD) | 6.77 (6.55) | 15.67 (7.10) | 6.25 (6.14) | |
| Etiology (TOAST) | ||||
| n (%) | Large-artery atherosclerosis | 40 (11.66) | 2 (11.11) | 38 (11.69) |
| Small-vessel occlusion | 32(9.32) | 0 (0.00) | 32 (9.85) | |
| Cardioembolism | 96 (27.98) | 7 (38.89) | 89 (27.38) | |
| Other determined etiology | 41 (11.95) | 5 (27.78) | 36 (11.08) | |
| Cryptogenic | 134 (39.06) | 4 (22.22) | 130 (40.00) | |
| Cerebral hemisphere | ||||
| n (%) | Left | 159 (46.35) | 6 (33.33) | 153 (47.08) |
| Right | 155 (45.18) | 9 (50.00) | 146 (44.92) | |
| Bilateral | 25 (7.28) | 3 (16.67) | 22 (6.77) | |
| Unknown | 4 (1.16) | 0 (0.00) | 4 (1.23) | |
| Cortical involvement | ||||
| n (%) | Yes | 234 (68.22) | 16 (88.89) | 218 (67.08) |
| No | 109 (31.77) | 2 (11.11) | 107 (32.92) | |
| Hemorrhagic transformation | ||||
| n (%) | Yes | 40 (11.66) | 9 (50.00) | 31 (9.54) |
| No | 303 (88.33) | 9 (50.00) | 294 (90.46) | |
Pharmacological thrombolysis as reperfusion treatment was performed in 24% of the cases, with 44.44% of patients with PSE and 23.7% in the group without epilepsy. Thrombectomy was performed only in 5% of the total number of patients, corresponding to a higher percentage of patients in the PSE group (22.2%) in contrast to the group that did not develop epilepsy (4.9%) (Table 3).
Treatment for AIS. Source: Authors.
| Treatment AIS | ||||
|---|---|---|---|---|
| With PSE(n = 18) | Without PSE(n = 325) | |||
| Thrombolysis | ||||
| n (%) | Yes | 83 (24.20) | 8 (44.44) | 75 (23.08) |
| Thrombectomy | ||||
| n (%) | Yes | 20 (5.83) | 4 (22.22) | 16 (4.92) |
| No | 323 (94.16) | 14 (77.78) | 309 (95.08) | |
PSE: Post-ischemic stroke epilepsy. AIS: Acute Ischemic Stroke.
The distribution in both groups was very similar in the different types of seizures. The most common was focal onset to bilateral tonic–clonic, in the case of ASS (52.94%) and PSE (50%). In the ASS, focal onset seizures corresponded to a total of 35.29%, with 23.53% focal awareness. In the PSE group, focal onset seizures accounted for approximately 50%, with 27.78% focal awareness. Regarding status epilepticus, only 2 patients (11.76%) in the ASS group presented this seizure as opposed to those with epilepsy who did not have this type of seizure (Table 4).
Seizure types. Source: Authors.
| Seizures | |||
|---|---|---|---|
| ASS(n = 17) | PSE (n = 18) | ||
| Seizures type | |||
| n (%) | Focal awareness | 4 (23.53) | 5 (27.78) |
| Focal impaired awareness | 2 (11.76) | 4 (22.22) | |
| Focal to bilateral tonic–clonic | 9 (52.94) | 9 (50.00) | |
| Status epilepticus | 2 (11.76) | 0 (0.00) | |
| Treatment | |||
| n (%) | Levetiracetam | 11 (64.71) | 9 (50.00) |
| Lacosamide | 1 (5.88) | 4 (22.22) | |
| Valproic acid | 1 (5.88) | 4 (22.22) | |
| Phenytoin | 1 (5.88) | 0 (0.00) | |
| Levetiracetam + valproic acid | 1 (5.88) | 1 (5.56) | |
| Benzodiazepines | 1 (5.88) | 0 (0.00) | |
| None | 1 (5.88) | 0 (0.00) | |
ASS: Acute Symptomatic Seizures, PSE: Post-Stroke Epilepsy.
Regarding pharmacological management for these seizures, levetiracetam was the most used drug in 64.71% of the patients with ASS and 50% in the group with PSE. In the group of patients with epilepsy, lacosamide and valproic acid were used as a treatment in 50% of the remaining patients. In contrast to the ASS group, these drugs and others such as benzodiazepines, phenytoin and combinations had an equal distribution with a percentage of 5.88% (Table 4).
DiscussionIn this cross-sectional study, the population of patients with AIS in a fourth-level university hospital was described, and the prevalence of PSE was calculated. The demographic characteristics of the patients, the mean age, and sex distribution, were like what has been described for patients with an AIS.24–27
The prevalence of PSE in this study was 5.25%, which corresponds to the reported in the literature which, although it varies significantly according to the definition of epilepsy used by the researchers, is between 2.5 and 15%.14,16 Similarly, it is estimated that approximately 6% of patients with ischemic stroke will develop PSE.14 The group of patients with PSE seems to be younger with an average age of 62.9 years compared to patients who did not develop epilepsy (70.8 years). It has been described that younger patients have a higher risk of PSE, with a mean of 65.5 years old, and our results align with this trend.20,28
Regarding comorbidities, most patients had a history of HTN, followed by DM as shown in our study. This is explained by evidence that these pathologies are the main risk factors for developing cerebrovascular disease.25,29 However, in those patients with PSE, the frequency of AF (22.2%) was higher compared to patients without PSE (18.77%). This may be related to the fact that patients with AF develop more cardioembolic cerebral infarcts involving cortical areas.29,30 Additionally, it has been described that AF and cardioembolic etiology are consistent risk factors for the development of PSE.29
Regarding the etiology of AIS according to the TOAST classification and the development of PSE, in this study, about 40% of these patients had cardioembolic etiology; it has been proposed that patients with cerebral infarction of cardioembolic origin have a higher proportion of cortical involvement and severity that increases the risk of having ASS, PSE, and recurrence of hospital visits after this diagnosis.29,31 However, the consistency of the direct association between cardioembolic etiology and the development of PSE is debated, and some studies have not shown that the risk is higher than that of other causes such as large vessel occlusion.19
Stroke severity according to NIHSS and hemorrhagic transformation had been consistent risk factors for the development of PSE.9,31–33 In this research, the mean NIHSS of patients with PSE was more than twice that of those without secondary epilepsy, and up to 50% of patients with PSE had a hemorrhagic transformation, which reinforces the importance as predictors in the development for this condition.
Similarly, cortical involvement is part of several risk predictor scales for both hemorrhagic stroke and PSE development as in the SeLECT scale.23,31,33 In this study, the proportion of cortical involvement was higher in patients with PSE compared to those who did not develop this condition (88.9% vs. 67.08%). This is probably because the pathophysiology of epileptogenesis following an ischemic event generates changes at the molecular and cellular level of cortical neurons that produce an increase in neuronal excitability.6,9,34,35 However, there is controversy and no consensus due to previous research results and variations in research methods.36
Reperfusion therapy has been a debated factor in the development of PSE. Some studies initially suggested that those treated with (intravenous thrombolysis) IV-tPA reduced the risk of PSE.29,37 However, this is contradicted by recent studies suggesting a relationship between reperfusion treatments either pharmacological or mechanical with the risk of developing PSE as implied in our study.13,29,38 For example, a meta-analysis had an odds ratio of 1.91% between endovascular therapy, post-stroke seizures, and a cumulative incidence of 6% of PSE with a similar distribution of patients treated with mechanical thrombectomy, intra-arterial thrombolysis, and bridging therapy.39 Likewise, another study reported that hemorrhagic transformation following endovascular therapy for AIS had approximately 5 times higher rate of developing PSE within 2 years.40 Nevertheless, a recent multicenter study found no association in endovascular treatments, so this association remains controversial.14,35
In relation to seizures, most patients had focal onset to bilateral tonic–clonic component; this has been described in the literature in which up to 57.1% of patients with PSE. Focal seizures with awareness were the next, with 22.5% in the literature, being slightly higher in our study.41,42 Finally, a percentage of 11% presented status epilepticus in the group of ASS but none in the PSE. It has been described that 20% with an AIS develop PSE.43,44
Regarding the treatment of PSE in this study, given that most of the epilepsies were of focal onset, broad-spectrum drugs such as levetiracetam or specific drugs for focal seizures such as lacosamide, carbamazepine or phenytoin were used.12,29,35,36,45 Currently, there is no conclusive evidence on the ideal management of PSE seizures. However, recent studies have suggested that epilepsy has a high response rate to treatment, without finding significant differences between different antiseizure drugs.46 Additionally, new research has focused on studying the pharmacological interaction between the most common comorbidities and anticonvulsant drugs.33,35 More studies are needed to explore the appropriate pharmacological management according to seizure type, comorbidities, and drug interactions for the treatment of PSE.
LimitationsAmong the limitations of this study, first, as it is a cross-sectional study, it is not possible to follow the population over time, calculate incidence or establish causal associations.47,48 In relation to the population, there was only one hospital included, so it is susceptible to selection bias and the population size becomes smaller.49 Additionally, there was loss to follow-up in 122 patients, which may have overestimated the prevalence of PSE.48 Similarly, most of the seizures in this population are non-convulsive and are diagnosed with electroencephalogram; however, not all patients included had this exam, so some participants may not have been diagnosed.15,44,50
ConclusionsThe prevalence of epilepsy in this population was 5.25%. The patients who developed this complication were mostly men and had a higher severity, cortical involvement, cardioembolic etiology, hemorrhagic transformation, and reperfusion treatments than those who did not. Similarly, the most frequently observed comorbidities were DM, HTN, and AF. In our study, no patient with epilepsy following an AIS presented status epilepticus. The most common pharmacological treatment was levetiracetam. Finally, future studies are needed to evaluate these findings to estimate the risk of developing PSE to identify populations at risk, and evaluate preventive measures.
Consent of the patient (informed consent)This study was approved by the Institutional Review Board (IRB) of Fundación Santa Fe de Bogota.
Ethical considerationsThe authors declare that no experiments were carried out on humans or animals for this research.
FundingThis work did not receive any industry funding or was commissioned.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
The neurological research group for dedication and work in the investigation.
