HCP Live
Contagion LiveCGT LiveNeurology LiveHCP LiveOncology LiveContemporary PediatricsContemporary OBGYNEndocrinology NetworkPractical CardiologyRheumatology Netowrk

Predicting Late Seizures After Intracerebral Hemorrhage: A New Tool

The knowledge that a patient is at high or low risk for late seizures has the potential to improve patient care.

A new clinical tool, the CAVE score, predicts the development of late seizures in patients with intracerebral hemorrhage (ICH).1 The CAVE score includes the following 4 risk factors: (1) cortical location (C), young age (A), large volume (V), and early (E) seizures.

In this study, the chances of developing late seizures were less than 1% in patients who had none of these clinical features; the chances were nearly 50% in those who had all 4 features. Seizure types included focal motor or autonomic signs only (29%), focal signs with impaired consciousness (29%), and convulsions (43%). Status epilepticus occurred in 16% of patients.


The authors retrospectively reviewed 993 consecutive patients with ICH from the observational Helsinki ICH study. Etiologies of ICH included amyloid angiopathy, anticoagulation, arteriovenous malformations, cavernomas, cirrhosis, hypertension, and undetermined causes. Patients with ischemic stroke, subarachnoid hemorrhage, trauma, or tumor were excluded. Seizures were divided into early (within 7 days) or late (after 7 days). Although an early seizure may be a transient event resulting from acute brain injury, late seizures are more likely to be recurrent and require ongoing antiepileptic drug treatment.


After 1 point was assigned to each of the 4 risk factors, the cumulative risk of developing late seizures was 0.6% (0 points), 3.6% (1 point), 9.8% (2 points), 34.8% (3 points), and 46.2% (4 points). For the whole population, the cumulative risk of late seizures was 7.1% after 1 year and 11.8% after 5 years. The c-statistic was 0.81 (0.76-0.86).

Validation Cohort

The authors applied these criteria to a separate patient cohort of 325 patients followed for 2.2 years (Lille Prognosis of InTra-Cerebral Hemorrhage study). The risk of late seizures predicted by the CAVE score in this cohort was 3.1% (0 points), 5.0% (1 point), 15.8% (2 points), 13.5% (3 points), and 37.5% (4 points). The c-statistic in this validation cohort was 0.69 (0.59-0.78).


The 4-point CAVE score is easy to apply. The inclusion of hemorrhage size and cortical location as 2 of the risk factors for late seizures is consistent with clinical experience. It is not surprising that the third risk factor, early seizures, is associated with late seizures. Finally, the, fourth risk factor, age younger than 65 years, also predicts late seizures, but the reason for this is unclear.


Late seizures develop in about 1 in 10 patients with ICH. As a prognostic indicator, the 4-point CAVE score is most useful for patients with extreme scores. For example, based on the Lille validation population, the risk of developing seizures is less than 4% in a patient with a score of "0"; there is a 37.5% chance in a patient with a score of "4."

These are only estimates, but the knowledge that a patient is at high or low risk for late seizures has the potential to improve patient care. For example, an episode of confusion in a patient with ICH might be more rapidly diagnosed as a partial complex seizure if the clinician knew that patient had a CAVE score of "4." Conversely, a CAVE score of "0" in that same patient would point toward a different etiology.

The predictive value of a CAVE score is not sufficient, however, to warrant prophylactic antiepileptic drug treatment in high-risk patients. Continued refinement of prognostic risk factors after ICH for the development of late seizures would be useful.


1. Happaniemi E, Strbian D, Rossi C, et al. The CAVE score for predicting late seizures after intracerebral hemorrhage. Stroke. 2014;45:1971-1976.