Assistant Professor of Clinical Neurology, Weill Cornell Medical College
SYNOPSIS: The reported incidence of post-stroke seizures varies and appears to have a higher incidence following hemorrhagic stroke. Post-stroke seizures have been associated with a greater length of hospitalization, increased mortality, and complications. This study compared four seizure prophylaxis strategies to determine the greatest net benefit for patients with spontaneous intracerebral hemorrhage. There was a better outcome with short-term antiepileptic medication use compared to long-term use.
SOURCE: Jones FJS, Sanches PR, Smith JR, et al. Seizure prophylaxis after spontaneous intracerebral hemorrhage. JAMA Neurol 2021; Jul 26;e212249. doi:10.1001/jamaneurol.2021.2249. [Online ahead of print].
The leading cause of seizures beyond the age of 60 years is cerebrovascular disease. The incidence of post-stroke seizures varies greatly and appears to have a higher incidence following hemorrhagic stroke. Following hemorrhage various causes can lead to seizures. Acutely, these include transient irritation of the cortex because of products of blood metabolism, and, chronically, hemosiderin depositions and gliotic scarring.
Post-stroke seizures have been associated with a greater length of hospitalization, increased mortality, and complications. Moreover, seizures may affect quality of life substantially with unpredictable reoccurrence, social and employment challenges, and additional associated costs. Thus, timely detection of seizures and optimal management are essential. Currently, there are no guidelines widely agreed upon for prophylactic use of antiseizure drugs following a brain hemorrhage, and there can be great variation in medication type, dosage, and duration.
Subsequent seizure risk potentially can determine if a patient will benefit from prophylactic/treatment strategies. Seizures can be classified as acute symptomatic seizures (early seizures), which occur seven days or less after a stroke injury, or seizures, which occur more than seven days after a stroke or injury. There is a suggested higher risk of recurrent seizures/epilepsy in patients with late seizures. Tools for predicting early (2HELPS2B score) and late (CAVE score) seizure risks can aid in prophylactic treatment decisions.1,2
The 2HELPS2B score is an electroencephalogram (EEG)-based seizure risk stratification tool based on the frequency of any periodic or rhythmic patterns > 2 Hz (excluding rhythmic delta activity), independent sporadic Epileptiform discharges, Lateralized periodic discharges, “Plus” features (superimposed fast, rhythmic, or sharp activity), prior Seizure/Suspicious event for clinical seizure, and Brief potentially ictal rhythmic discharges (BIRDs). Based on the score, the tool will recommend duration of EEG monitoring to detect 95% of seizures and the risk of seizures within 72 hours. The 2HELPS2B is helpful if there is accessibility to EEG monitoring.2 The CAVE score is a rapid method to calculate, with a high correlation, between risk factors and late seizures/epilepsy. This score is calculated based on Cortex involvement, Age < 65 years, hemorrhage Volume > 10 mL, and Early seizure (seven days or less after the intracerebral hemorrhage) for a maximal score of 4. In this model, a higher score correlates to higher risk of late seizures.1
This study by Jones et al evaluated and compared four seizure prophylaxis strategies for patients with spontaneous intracerebral hemorrhage. This was a decision analysis study and simulation model of four cases representing common clinical scenarios with a presentation of intracerebral hemorrhage without a history of epilepsy and receiving seizure prophylaxis:
• Case 1: A patient with a low risk of late seizure and average antiseizure drug adverse reaction and efficacy;
• Case 2: A patient with a low risk of late seizure, high risk of antiseizure drug adverse reaction, and low efficacy;
• Case 3: A patient with a high risk of late seizure and average antiseizure drug adverse reaction and efficacy;
• Case 4: A patient with high risk of late seizure, high risk of antiseizure drug adverse reaction, and low efficacy.
Within these four scenarios, four treatment strategies, based on primary vs. secondary prophylaxis, timing of seizures (early vs. late), and duration of therapy short (one week) vs. long (indefinite), were compared for a primary outcome of maximal quality-adjusted life years (QALYs):
1. Conservative: Patients with early seizures received short-term secondary early seizure prophylaxis.
2. Moderate: Patients monitored who developed early or late seizures received long-term secondary prophylaxis/therapy.
3. Aggressive: Patients received long-term primary prophylaxis on hospitalization.
4. Risk-guided: Patients underwent a screening EEG for early seizure risk stratification on admission using the 2HELPS2B score classified as low risk (0 points), medium risk (1 point), or high risk (> 2 points). Patients with low risk received conservative treatment, patients with medium and high risk received short-term primary prophylaxis, which was discontinued after one week if they remained seizure-free. If patients developed recurrent/late seizures, they received long-term secondary therapy.
The study suggested a better outcome with short-term antiseizure use over long-term medication use. In case 1, the risk-guided strategy (8.13 QALYs) outperformed conservative (8.08 QALYs), moderate (8.07 QALYs), and aggressive (7.88 QALYs) strategies. In case 2, the conservative strategy (2.18 QALYs) performed better than risk-guided (2.17 QALYs), moderate (2.09 QALYs), and aggressive (1.15 QALYs) strategies. In case 3, the aggressive (9.21 QALYs) strategy outperformed risk-guided (8.98 QALYs), moderate (8.93 QALYs), and conservative (8.77 QALYs) strategies. In case 4, the risk-guided strategy (11.53 QALYs) again per-formed better than conservative (11.23 QALYs), moderate (10.93 QALYs), and aggressive (8.08 QALYs) strategies.
These findings suggested that, overall, the risk-guided strategy and the conservative treatment strategy performed best for most scenarios in this simulation, with the exception of high-risk seizure patients (case 3, with a CAVE score ≥ 3) and patients with lower risk of long-term antiseizure medication adverse reactions (younger age with few comorbidities or comedications), for whom an aggressive strategy may be beneficial.
These findings are in keeping with clinical practice and published literature suggesting that the risk for post-stroke seizures is greater following hemorrhagic strokes. There is a greater risk of seizures with cortical involvement than with hemorrhages within deeper structures. Additionally, studies have found a higher risk of recurrent seizures/epilepsy in patients with late seizures than in patients with early seizures. There should be a low threshold for EEG monitoring in post-stroke patients with altered mental status that is out of proportion for the clinical picture. This study supports a risk-guided treatment strategy using seizure risk stratification tools, such as the 2HELPS2B and CAVE scores, to help guide treatment strategies and avoid prolonged and potentially unnecessary antiseizure medications. When treating with antiseizure medications, it is important to consider concurrent medication use and comorbidities to decrease potential adverse reactions.
The limitations of this study are related to using a simulated model with published data and assuming the lifetime late-seizure risk was equivalent to the CAVE derived risk. Further prospective studies would be beneficial to determine the optimal choice and duration of antiseizure drug use in the prevention and treatment of seizures in post-stroke patients.
- Haapaniemi E, Strbian D, Rossi C, et al. The CAVE score for predicting late seizures after intracerebral hemorrhage. Stroke 2014;45:1971-1976.
- Moffet EW, Subramaniam T, Hirsch LJ, et al. Validation of the 2HELPS2B seizure risk score in acute brain injury patients. Neurocrit Care 2020;33:701-707.