Use of antiepileptic medications for seizures’ prevention during subarachnoid hemorrhage: A retrospective observational study

Abstract

The use of prophylactic antiepileptic drugs (AEDs) post-subarachnoid hemorrhage (SAH), particularly aneurysmal SAH, is controversial, with limited data available. This has led the new American Heart Association/American Stroke Association (AHA/ASA) guidelines to recommend against using AEDs. This study is aimed at determining whether the use of AEDs for primary prophylaxis is effective in reducing the incidence of seizures post-SAH. A retrospective observational study was conducted utilizing a reviewing chart for the period starting from June 2015 to the end of 2021. The reviews were conducted in the acute care areas of 2 tertiary hospitals primarily to assess the efficacy of AEDs against seizures in patients with SAH (particularly aneurysmal SAH). This was done by comparing the occurrence of early, late, and overall incidence of seizures between patients who received AEDs versus those who did not. Of the 62 patients, who mostly presented with aneurysmal SAH (71%), 42 received AEDs and 20 did not. Mostly, the baseline characteristics between the 2 groups were comparable. A few patients on AEDs developed early (n = 4/38), late (n = 3/29), and overall seizures (n = 6/33), whereas no early, late, or overall incidence of seizures was presented in the group who did not receive AEDs. However, this difference showed no significance (P > .05). The subjects who were given AEDs showed significantly longer hospital stays (42.11 ± 51.43 vs 14.10 ± 7.17; P = .002) and higher mortality rates (7/11 vs 0/11; P = .026). For all patients who received AEDs for prophylaxis, the overall incidence of seizures was negatively associated with the Glasgow coma scale (OR: 0.798; 95% CI 0.657–0.978; P = .022). Our findings support the 2023 AHA/ASA guideline recommendation to avoid using routine AEDs for prophylaxis for all SAH patients. Proper and careful stratification methods should be implemented in each given scenario.

1. Introduction

Subarachnoid hemorrhage (SAH) is a medical emergency that has a very high mortality and morbidity rate.^[1] SAH is defined as bleeding within the subarachnoid space that is located between the arachnoid and pia mater, which is normally filled with cerebrospinal fluid.^[2] SAH can occur due to different etiologies, including trauma, intracranial aneurysms, arteriovenous malformations, vasculitides, tumors, and coagulopathies.^[3] Although trauma is the primary cause of SAH, most cases of SAH are due to intracranial aneurysms.^[3,4]

The incidence of SAH is generally 1.24 times higher in men compared to women. Concerning the influence of gender and age on SAH incidence, a higher incidence was seen in younger males (aged 25–45 years), females aged between 55 and 85 years, and males aged > 85 years.^[1,5] SAH is associated with complications that include seizures, hydrocephalus, vasospasm, rebleeding, and imbalances in water metabolism. Surgical clipping or endovascular coiling are 2 approaches that are used to secure aneurysm. However, some patients with impaired neurological function or high comorbidity may be treated conservatively.^[1,6]

The overall risk of seizures following aneurysmal SAH ranges from 6% to 18%.^[1] In aneurysmal SAH, seizures have been classified into onset or ictal seizures, which are defined as episodes occurring within the first 12 to 24 hours of aneurysmal SAH, early seizures or postoperative seizures, which occur within 2 weeks postoperatively, and late seizures, which occur spontaneously at least twice after 2 weeks postoperatively.^[7,8] Some studies have defined early seizures as those that occur within a week, whereas late seizures are defined as those that occur spontaneously at least twice after the first week of SAH.^[9,10] Unfortunately, there is no uniform definition of postoperative epilepsy, combined with problems of different durations of follow-up.

The incidence of early seizures in aneurysmal SAH is 2.3%, and the incidence of late seizures is 5.5%.^[10] The previous American Heart Association/American Stroke Association (AHA/ASA) guidelines supported the use of prophylactic antiepileptic drugs (AEDs) in the immediate post-hemorrhagic period (3–7 days from ictus).^[1] However, the AHA/ASA guidelines did not specify either the drug of choice or the appropriate dose. Most importantly, there is no reliable data or evidence regarding the efficacy of AED continuing or stopping once an aneurysm is secured. Such questionable outcomes led the new AHA/ASA guidelines to advise against the use of AEDs due to the absence of clinical trials.^[11] Therefore, the use of prophylactic AEDs is controversial, with limited data available to develop recommendations. This study is aimed at determining whether the use of AEDs as primary prophylactics is effective in reducing the incidence of post-aneurysmal SAH seizures.

2. Materials and methods

This study was performed by retrieving and reviewing the files of all patients during the period from June 2015 to the end of 2021 in the acute care areas (emergency and intensive care units) of 2 tertiary hospitals: King Saud University Medical City (KSUMC) and King Abdulaziz Medical City (KAMC). The ethical approval letter was obtained from the ethical committees of the 2 hospitals, KSUMC (21/01077/IRB) and KAMC (NCR21R/306/06). Both ethical committees waived patients’ consent because the study is an observational, non-interventional study.

The primary outcome was to assess the efficacy of AEDs as seizure prophylaxis in patients with SAH (particularly aneurysmal SAH) by comparing the occurrence of early, late, and overall incidence of seizures between patients who received AEDs versus those who did not. The incidence of death and length of hospital stay between the 2 groups – those who received AEDs versus those who did not – and the association between patients’ factors or SAH characteristics on the seizure incidence among patients who received AEDs as prophylaxis were the secondary outcomes. Early seizure is defined as the 7 days post-patient diagnosis based on the duration of recommended agencies and literature, whereas late seizure is beyond 7 days.

Adult patients (≥18 years) with SAH diagnoses who either received AEDs for prophylaxis or not were included in our study. Patients were excluded if they: were younger than 18 years; presented with preexisting seizure disorder; and died within 72 hours of admission. The data were collected by reviewing the files of all patients who were diagnosed with SAH. Once the patient fits our inclusion and exclusion criteria, the demographic characteristics, clinical characteristics at presentation (Hunt and Hess grade, modified Rankin Scale, and location of aneurysm), treatment modality, AEDs’ prescription characteristics (type, dose, timing, and duration of use), onset of seizure, and adverse effects related to prophylactic AEDs were collected.

IBM SPSS statistics software was utilized for data analysis. Two types of variables were determined in our findings: quantitative (expressed as mean ± SD) and qualitative (expressed as frequency and percentage). An independent Student t test was used to compare continuous variables, while a Fisher exact test and a Chi-squared test were used to assess any association between the qualitative variables. Logistic regression was carried out to evaluate the association between anti-epileptic use and patients’ ages, body mass index (BMI), gender, Glasgow coma scale (GCS), size of aneurysm, type of aneurysm, intraventricular hemorrhage, and hydrocephalus. A statistical significance was determined with a P value < .05.

3. Results

A total of 68 patients met our inclusion criteria; however, 6 patients were excluded due to the lack of most of their baseline characteristics. Of the 62 included patients, 42 received AEDs and 20 did not. The mean age and BMI of our subjects were 49.95 ± 13.04 years and 29.95 ± 6.87 kg/m², respectively. Most patients presented with aneurysm SAH (n = 44; 71%), with an average size of 4.69 ± 3.72 mm. Mostly, the baseline characteristics between patients who received AED and those who did not receive AED were comparable. More details of the baseline characteristics are summarized in Table 1.

Table 1.

Demographics and clinical characteristics.

Characteristic	Total [n = 62]	No AED [n = 20]	AED prophylaxis [n = 42]	P value
Age (yrs; mean ± SD)	49.95 ± 13.04	49.30 ± 15.77	50.26 ± 11.73	.788
Weight (kg; mean ± SD)a	78.81 ± 16.62	79.395 ± 16.47	78.529 ± 16.89	.85
BMI (kg/m2; mean ± SD)a	29.95 ± 6.87	31.99 ± 8.82	28.92 ± 5.50	.104
Gender
Male	33 (53.2)	7 (21.2)	26 (78.8)	.06
Female	29 (46.8)	13 (44.8)	16 (55.2)
GCS on admission (mean ± SD)a	12.1 ± 4.05	14.30 ± 1.72	11.00 ± 4.44	.002
Hunt and Hessa
<3	10 (16.1)	2 (20)	8 (80)	.529
≥3	6 (9.7)	1 (16.7)	5 (83.3)
Comorbiditiesa
Cardiovascular	36 (58.1)	13 (36.1)	23 (63.9)	.584
Cerebrovascular	4 (6.5)	0	4 (100)	.295
Other	19 (30.6)	6 (31.6)	13 (68.4)	1
Type of SAH
Aneurysm	44 (71)	16 (36.4)	28 (63.6)	.557
Traumatic	9 (14.5)	2 (22.2)	7 (77.8)
Unknown	9 (14.5)	2 (22.2)	7 (77.8)
Intraventricular haemorrhage
No	54 (87.1)	20 (37)	34 (63)	.046
Yes	8 (12.9)	0	8 (100)
Location of aneurysma
MCA	8 (12.9)	3 (37.5)	5 (62.5)	1
ACA	10 (16.1)	2 (20)	8 (80)	.273
ACOM	11 (17.7)	5 (45.5)	6 (54.5)	.719
ICA	10 (16.1)	5 (50)	5 (50)	.460
Others	7 (11.3)	4 (57.1)	3 (42.9)	.394
Size of aneurysma
Average size (mean ± SD)	4.69 ± 3.72	4.15 ± 2.73	5.04 ± 4.29	.495
< 5mm	20 (32.2)	9 (45)	11 (55)
≥ 5mm	7 (11.3)	5 (33.3)	10 (66.7)	.262
NA	19 (30.6)	1 (12.5)	7 (87.5)
Aneurysm treatmenta
Clipping	2 (3.2)	0	2 (100)	.717
Coiling	37 (59.7)	14 (37.8)	23 (62.2)
None	3 (4.8)	1 (33.3)	2 (66.7)
NA	2 (3.2)	1 (50)	1 (50)
Seizure at presentation
No	52 (83.9)	20 (38.5)	32 (61.5)	.023
Yes	10 (16.1)	0	10 (100)
Hydrocephalus
No	50 (80.6)	19 (38)	31 (62)	.083
Yes	12 (19.4)	1 (8.3)	11 (91.7)

P value < 0.05 shows statistical significance.

ACA = anterior cerebral artery; ACOM = anterior communicating artery; AED = antiepileptic drug; BMI = body mass index; GCS = Glasgow coma scale; ICA = internal carotid artery; MCA = middle cerebral artery; SAH = subarachnoid haemorrhage.

^aSome missing information.

Few patients on AED prophylaxis developed early (n = 4/38), late (n = 3/29), and overall seizures (n = 6/33), whereas no early, late, or overall incidence of seizures was presented in the group that did not receive AEDs. However, this difference showed no significance (Table 2). Patients who used AEDs showed significantly longer hospital stays (42.11 ± 51.43 vs 14.10 ± 7.17; P = .002) and mortality rates (7/11 vs 0/11; P = .026). Furthermore, the choice of a particular AED agent (levetiracetam, phenytoin, or combination of AEDs) for prophylaxis showed a non-significant impact on the different incidence of seizures, length of hospital stay, or mortality (Table 3). The mean loading doses of levetiracetam and phenytoin were 1277.78 ± 754.62 mg and 1014.2 ± 520.44 mg, respectively. The mean daily maintenance dose of levetiracetam was 1082.86 ± 447.53 mg, whereas the mean daily maintenance dose of phenytoin was 882.14 ± 512.84 mg.

Table 2.

Impact of anticonvulsant therapy on inpatients’ health outcomes.

Outcome	All [n (%)]	No AED [n (%)]	AED prophylaxis [n (%)]	P value
Early seizure (≤7 d)a
No	54 (87.1)	20 (37)	34 (63)	.288
Yes	4 (6.5)	0	4 (100)
Late seizure (days 8–10)a
No	45 (72.6)	16 (35.6)	29 (64.4)	.541
Yes	3 (4.8)	0	3 (100)
Overall seizure during prophylaxis perioda
No	43 (69.4)	16 (37.2)	27 (62.8)	.159
Yes	6b (9.7)	0	6 (100)
Hospital length of staya	32.45 ± 43.75	14.10 ± 7.17	42.11 ± 51.43	.002
Deatha
No	22 (35.5)	11 (50)	11 (50)	.026
Yes	7 (11.3)	0	7 (100)

P value < 0.05 shows statistical significance.

AED = antiepileptic drug.

^aMissing patients.

^bOne patient of who developed early seizure died on day 4.

Table 3.

Impact of individual anticonvulsant therapy on inpatients’ health outcomes.

Outcome	Levetiracetam (n = 26)	Phenytoin (n = 6)	Combination (n = 6)	P value
Early seizure (≤7 d)
No	23 (67.6)	6 (17.6)	5 (14.7)	.614
Yes	3 (75)	0	1 (25)
Late seizure (days 8–10)a
No	19 (65.5)	5 (17.2)	5 (17.2)	.471
Yes	3 (100)	0	0
Overall seizure during prophylaxis perioda
No	18 (66.7)	5 (18.5)	4 (14.58)	.517
Yes	5# (83.3)	0	1 (16.7)
Hospital length of stay	48.65 ± 61.18	20.83 ± 11.82	37.29 ± 23.29	.489
Deatha
No	7 (63.6)	3 (27.3)	1 (9.1)	.535
Yes	6 (85.7)	1 (14.3)	–

^aMissing patients.

^#One patient of who developed early seizure died on day 4.

For all patients who received AEDs for prophylaxis, the overall incidence of seizures was negatively associated with GCS (OR: 0.798; 95% CI 0.657–0.978; P = .022). Also, the early incidence of seizures was negatively associated with GCS (OR: 0.765; 95% CI 0.604–0.968; P = .022), but not with the late incidence of seizures. In contrast, the early incidence of seizures was positively associated with intraventricular hemorrhage (OR: 9.8; 95% CI 1.124–85.42; P = .039). The association between AED usage as prophylaxis and patients’ factors on the overall incidence of seizures is presented in Table 4.

Table 4.

Factors associated with the incidence of seizures during all period of prophylaxis.

Patients’ factors	Odds ratio	95% CI	P value
Age	0.998	0.933 to 1.068	.957
BMI	0.689	0.503 to 0.944	.2
Gender	1.15	0.208 to 6.35	.873
GCSa	0.798	0.657 to 0.968	.022
Size of anuresym	1.013	0.748 to 1.372	.935
MCA	1.167	0.11 to 12.38	.898
ICA	2.57	0.358 to 18.5	.348
Other	1.45	0.133 to 15.79	.76
Intraventricular hemorrhage#	4.875	0.67t to 35.48	.118
Hydrocephalus	0.58	0.061 to 5.54	.638

P value < 0.05 shows statistical significance.

BMI = body mass index; GCS = Glasgow coma scale; ICA = internal carotid artery; MCA = middle cerebral artery.

^aGCS showed significant incidence of early seizure during AED prophylaxis (OR: 0.765; 95% CI 0.604–0.978; P = .026).

^#Intraventricular hemorrhage showed significant incidence of early seizure during AED prophylaxis (OR: 0.765; 95% CI 0.604–0.978; P = 0.026).

4. Discussion

The associated negative adverse effects of phenytoin and the lack of randomized clinical trials that support using antiepileptics, particularly those with a good safety profile, led to the most recent AHA/ASA guidelines for the management of aneurysmal SAH against using antiepileptic medications as seizure prophylaxis. However, the involvement of the middle cerebral artery (MCA), the high grade of Hunt and Hess (≥3) or modified Fisher scale (≥III), and the presence of hydrocephalus justify the benefits of antiepileptic utilization.^[11,12] With the absence of randomized comparisons between AEDs and placebo for the primary prevention of seizures in patients presenting with SAH, the recent guideline recommendations further excluded the retrospective results of studies/analyses. Some clinicians may be reluctant to obey the new recommendations, especially with the availability of AEDs with a few adverse effects.

In our study, the practice of administering prophylactic AEDs was more common compared to no administration of AEDs, which accounted for 67.7% and 32.3%, respectively. However, AEDs’ prophylaxis was not found to be significantly more effective than not using AEDs for preventing early, late, or overall seizures in SAH patients (Table 2). In contrast, the hospital length of stay was longer in patients using AEDs for prophylaxis compared to patients not on AEDs (14.1 vs 42.1 days; P = .002). The study results are consistent with multiple recent reports regarding the lack of effectiveness of adding AEDs for prophylaxis in patients with SAH.^[12–14] A retrospective study was conducted to evaluate the reduction of post-SAH seizure incidence with prophylactic administration of AEDs and showed that the use of prophylactic AEDs after SAH was not associated with a significant difference in the incidence of seizures.^[13] Moreover, Smith et al^[14] found no difference in the seizure incidence between patients who received prophylactic AEDs and those who did not. Additionally, a subsequent meta-analysis published in 2013 provided more evidence of the lack of effectiveness of AEDs for prophylaxis in patients with SAH.^[12]

Moreover, the mortality rate was higher in the AEDs’ prophylaxis group, as 7 patients died compared to those who received no AEDs (P = .026). However, 33 patients had missing data, which could be contributing to this significant difference. This finding may seem interesting, as previous data showed an increased risk of morbidity and mortality with the administration of phenytoin alone.^[15,16] Most of our patients were on levetiracetam, and this correlation brings a new insight into the impact of possibly all AEDs on the patients. Based on the provided data, it is probably recommended to reduce exposure to all AEDs following SAH. Additionally, a more robust study should be conducted to evaluate the use of AEDs in SAH patients who are at a high risk of developing seizures according to the 2023 AHA guidelines.^[16]

Overall, only 6 patients, accounting for 9.7%, experienced seizures during the prophylaxis period. The incidence of early seizures was 6.5%, whereas late seizures were less common, with an incidence of 4.8%. In patients using AED prophylaxis, there was no significant difference between the efficacy of levetiracetam and phenytoin. The use of levetiracetam was more common than that of phenytoin, which could be due to the necessity of frequent monitoring of phenytoin therapeutic levels. As summarized in Table 3, the number of patients with early seizures was 3 in the levetiracetam group (n = 26) versus none in the phenytoin group (n = 6). Moreover, 3 patients had a late seizure in the levetiracetam group (n = 22) compared to none in the phenytoin group (n = 5); however, the incidence of seizures between phenytoin, levetiracetam, or the combination groups did not reach a statistical difference. In accordance with previous data, the choice of AEDs for prophylaxis had no impact on the effectiveness of clinical outcomes.^{[10,11,13,16–19]} However, our data suggest that the use of levetiracetam is associated with a longer hospital stay compared to phenytoin, numerically, which was not reported previously. It is not clear whether this should be further investigated or explored, as it did not reach a statistical difference.

The 2023 AHA/ASA guidelines recommend the use of AEDs as prophylactic agents if a patient with SAH is presented with high-grade Hunt and Hess or Fisher scale, hydrocephalus complications, or MCA involvement.^[11] Among all included patients who were administered AEDs for seizure prophylaxis, a logic regression was conducted to assess the impact of patients’ factors on the incidence of overall seizure (Table 4). Neither the presence of hydrocephalus nor the involvement of MCA had an influence on the seizure incidence. Due to inadequate documentation, Hunt and Hess and Fisher scales were excluded from this analysis. The GCS is the only factor that should significantly impact the overall seizure incidence. As the GCS score increases, the likelihood of seizure occurrence decreases in patients who received AEDs for prophylaxis (OR: 0.765; 95% CI 0.604–0.968; P = .022). However, regardless of the administration of AEDs, the incidence of seizures is less with high GCS in patients with SAH.^[20] Additionally, a similar study comparing the incidence of seizures between the AED-administered group versus non-administered group in aneurysmal SAH found that the GCS score is not an independent predictor of seizure occurrence post-aneurysmal SAH (median, IQR: 7.15 vs 6.15; P = .45).^[21]

Despite our best efforts in this study, we recognize that the retrospective nature of our study may imply suboptimal data for clinical correlation. Additionally, given our baseline characteristics, the included patients may represent patients with mild to moderate seizures based on the risk stratification assessment that includes the surgical management of aneurysms, clinical grade (HH grade ≥ 3), MCA aneurysm location, and hydrocephalus. Moreover, missing data may have affected our interpretation of the given population. However, compared to previous data, our sample size seems to have included a larger number of patients, which further strengthens our results.

5. Conclusions

Despite the common practice of administering the new AEDs, particularly levetiracetam, as seizure prophylactics in patients with SAH, certainly aneurysmal SAH, the efficacy of such prophylaxis is not guaranteed, and the safety of possible death and a longer hospital stay could be a critical concern. In conclusion, our findings support the 2023 AHA/ASA guideline recommendations to avoid using routine AEDs for prophylaxis in all SAH cases. A proper and careful stratification method should be implemented in each given scenario.

Acknowledgments

The authors extend their appreciation for the support of the Research Supporting Project number (RSPD-2024R794), King Saud University, Riyadh, Saudi Arabia.

Author contributions

Conceptualization: Sultan Alghadeer, Reham M. Binhazza.

Funding acquisition: Sultan Alghadeer.

Methodology: Sultan Alghadeer, Abdulrahman Alwahibi.

Supervision: Sultan Alghadeer.

Validation: Sultan Alghadeer, Feda F. Alsaloom, Abdulrahman I. Alshaya.

Investigation: Reham M. Binhazza, Hayaa Alyahya, Abrar Al-Ghamdi.

Writing – original draft: Reham M. Binhazza, Abdulrahman A. Alghamdi.

Formal analysis: Abdulrahman Alwahibi.

Software: Abdulrahman Alwahibi.

Data curation: Feda F. Alsaloom.

Writing – review & editing: Abdulrahman I. Alshaya, Hayaa Alyahya, Abrar Al-Ghamdi.

Visualization: Abdulrahman A. Alghamdi.

Funding acquisition: Sultan Alghadeer.