Introduction: The Clinical Challenge of Poststroke Epilepsy
Ischemic stroke is not only a leading cause of long-term disability but also the most common cause of adult-onset epilepsy. Poststroke epilepsy (PSE)—defined as the occurrence of unprovoked seizures more than seven days after the index event—affects a significant subset of survivors, leading to increased morbidity, higher mortality rates, and a diminished quality of life. Despite its clinical importance, identifying which patients are at the highest risk for PSE has remained a challenge for clinicians. Previous models, such as the SeLECT and SeLECT2.0 scores, provided a foundation for risk assessment, but their performance has been variable across different populations and clinical settings. The need for a more robust, imaging-integrated, and internationally validated tool led to the development of the IsCHEMiA score.
The Development of the IsCHEMiA Score
Researchers at Massachusetts General Hospital, in collaboration with international institutions, sought to develop a more precise predictive model by incorporating specific neuroimaging features alongside clinical data. The IsCHEMiA score was derived from a cohort of 1,436 patients who experienced their first-ever acute ischemic stroke. Unlike earlier models that relied heavily on clinical presentation, the IsCHEMiA score leverages the richness of modern neuroimaging to capture the structural nuances of the stroke injury.
The mnemonic IsCHEMiA represents the six independent predictors included in the final multivariable model:
- Is: Infarct size (larger volumes correlate with higher risk)
- C: Cortical involvement (disruption of cortical networks is a known epileptogenic trigger)
- H: Hemorrhagic transformation (the presence of blood products can irritate cortical tissue)
- E: Early seizures (seizures occurring within the first seven days of stroke)
- Mi: MCA (Middle Cerebral Artery) involvement (the territory most associated with cortical injury)
- A: Age younger than 65 (younger patients often exhibit higher neuroplasticity but also a higher relative risk for PSE)
Methodological Rigor and International Validation
A key strength of the IsCHEMiA study is its rigorous statistical approach and extensive validation. The researchers used competing risk regression, with all-cause mortality treated as a competing event. This is a critical methodological choice in stroke research, as many elderly stroke patients may pass away before they have the opportunity to develop or be diagnosed with PSE. Ignoring this competing risk can lead to an overestimation of the true incidence of epilepsy.
To ensure the model’s generalizability, it was externally validated in three distinct international cohorts: Queen Mary Hospital and Ruttonjee Hospital in Hong Kong, and the National Cerebral and Cardiovascular Center in Japan. This validation across 2,534 patients from different healthcare systems and ethnicities provides high confidence in the score’s cross-population utility, particularly in the modern era of reperfusion therapies (intravenous thrombolysis and mechanical thrombectomy).
Key Findings: Precision in Prediction
In the overall study population, the incidence of PSE was 5.5%. The IsCHEMiA score demonstrated exceptional discrimination, with c-statistics ranging from 0.826 to 0.870 across the US and Asian cohorts. This level of accuracy significantly outperformed the existing SeLECT score (c-statistic 0.848 vs 0.782, p < 0.0001). Calibration was also excellent, meaning the predicted risks closely matched the observed outcomes at 1-year and 3-year intervals.
The practical implications of the score are best illustrated by its risk stratification capabilities:
- Low Risk: A patient with an IsCHEMiA score of 3 has a predicted PSE risk of only 2% at 1 year and 6% at 5 years.
- High Risk: Conversely, a patient with an IsCHEMiA score of 8 or higher faces a substantial risk, with a 67% probability of PSE at 1 year and a 78% probability at 5 years.
Expert Commentary: A New Standard for Clinical Practice
The IsCHEMiA score represents a significant leap forward in personalized stroke care. By identifying high-risk individuals early, clinicians can tailor follow-up protocols, provide better patient counseling, and remain vigilant for subclinical seizure activity. From a neurobiological perspective, the inclusion of cortical involvement and hemorrhagic transformation aligns with our understanding of post-ischemic excitotoxicity and iron-mediated epileptogenesis.
One notable aspect is the inclusion of MCA involvement and younger age. While stroke in the elderly is more common, younger brains may undergo different remodeling processes that, when combined with large cortical infarcts, increase the likelihood of developing chronic epilepsy. The researchers’ ability to validate these findings in the era of mechanical thrombectomy is particularly relevant, as the imaging characteristics of patients post-reperfusion can differ significantly from those who do not receive such interventions.
Clinical Utility and Translational Impact
Beyond individual patient care, the IsCHEMiA score has profound implications for clinical research. Historically, trials of prophylactic antiepileptic drugs (AEDs) after stroke have failed to show a definitive benefit, likely because they included broad stroke populations with a low baseline risk of epilepsy. By using the IsCHEMiA score to enrich trial populations with high-risk individuals (e.g., those with a score ≥7), researchers may finally be able to determine if early intervention can prevent the development of epilepsy—a concept known as antiepileptogenesis.
Furthermore, the score is “readily applicable.” Unlike some complex genomic or proteomic biomarkers, the components of the IsCHEMiA score are derived from standard clinical assessments and routine neuroimaging (CT or MRI) performed in almost every acute stroke center worldwide. This ensures that the tool can be implemented in both academic medical centers and community hospitals.
Conclusion: A Foundation for Personalized Management
The development and validation of the IsCHEMiA score mark a milestone in the long-term management of stroke survivors. By providing a reliable, evidence-based method to predict poststroke epilepsy, it moves the field away from a “one-size-fits-all” approach toward precision medicine. While further studies may explore its utility in purely hemorrhagic strokes or its integration with electroencephalography (EEG) data, the IsCHEMiA score currently stands as the most robust tool for predicting PSE in the ischemic stroke population.
References
Leung WCY, Tanaka T, Donahue RA, et al. Development and International Validation of a Novel Imaging-Based Risk Score (IsCHEMiA) for the Prediction of Poststroke Epilepsy. Neurology. 2026;106(2):e214486. doi:10.1212/WNL.0000000000214486.