Skip to content

Epileptiform Abnormalities Predict Long-term Outcome After Subacute Hemorrhage

Key findings

  • This prospective, observational cohort study evaluated whether EEG biomarkers linked to delayed cerebral ischemia (DCI) predict long-term functional outcome after spontaneous subarachnoid hemorrhage
  • Associations between new or worsening epileptiform abnormalities and poor modified Rankin score (mRS of 4–6) at discharge, three months later and six months later were similar to associations between mRS and DCI itself
  • Deterioration of background frequencies was not associated with longitudinal functional outcome
  • Neither were hospital complications (hospital-acquired infection and hydrocephalus) or changes in antiseizure medication associated with mRS at any timepoint studied
  • Patients with new or worsening epileptiform abnormalities may be candidates for enhanced therapies to improve functional outcome

Several findings on continuous electroencephalography (EEG) are known to predict delayed cerebral ischemia (DCI), a dangerous complication, after spontaneous subarachnoid hemorrhage (SAH). Chief among them are new or worsening epileptiform abnormalities (EAs) and background deterioration (BD). It's not known, however, whether these biomarkers are transient phenomena or have a sustained association with functional outcomes.

In a prospective observational study, India A. Lissak, MA, clinical research coordinator, W. Taylor Kimberly, MD, PhD, chief of the Division of Neurocritical Care, and Eric S. Rosenthal, MD, director of the Neurosciences Intensive Care Unit of the Department of Neurology at Massachusetts General Hospital, and colleagues found that EA burden, but not BD, predicts long-term functional outcomes. In Neurocritical Care, they suggest EAs could identify candidates for enhanced treatment.

Study Cohort

The study subjects were 59 adults who were admitted to an ICU with nontraumatic SAH, had EEG performed for ischemia monitoring for at least 72 hours and had a modified Rankin score (mRS) recorded at discharge, three months later and six months later. Poor mRS was defined as a score of 4–6.

Incidence of EEG Abnormalities

  • New or worsening EAs (increase in the duration of interictal epileptiform discharges, lateralized rhythmic delta activity, lateralized periodic discharges or generalized periodic discharges): 39% of patients
  • New or worsening BD (decreasing alpha–delta ratio, relative alpha variability or worsening focal slowing): 41%
  • Both EAs and BD: 20%

Primary Outcomes

After adjustment for age, clinical grade and radiologic grade, the association between new or worsening EAs and poor mRS was:

  • At discharge: OR, 4.99; P=0.006
  • At 3 months: OR, 3.28; P=0.03
  • At 6 months: OR, 2.71; P=0.06

The results were similar when further adjusted for the APACHE II physiologic subscore and disease severity at admission. Worsening BD was not associated with mRS at any of the three time points studied.

Secondary Outcomes

The adjusted association between DCI or mRS was:

  • At discharge: OR, 4.75; P=0.004
  • At 3 months: OR, 3.40; P=0.02
  • At 6 months: OR, 2.45; P=0.076

Hospital-acquired infection, hydrocephalus and changes in antiseizure medication were not associated with mRS at any time point studied. The incidence of aneurysmal rebleeding was only 3%, so its tie to mRS could not be studied definitively.

Clinical Relevance

New or worsening EAs represent poor cerebral metabolism and have been shown to occur before DCI. In addition, EAs have been linked to cortical spreading depolarizations, which are also associated with DCI and poor outcomes.

In light of that previous research, these new results indicate patients with new or worsening EAs may be candidates for enhanced strategies to improve functional outcomes. EAs should also be considered when developing criteria for clinical trial enrollment, so that targeted treatments go to the patients with the greatest opportunity to benefit.

5x
greater risk of poor outcome at discharge in patients who had new or worsening epileptiform abnormalities after nontraumatic subarachnoid hemorrhage

4.8x
greater risk of poor outcome at 3 months in patients with new or worsening epileptiform abnormalities after nontraumatic subarachnoid hemorrhage

Related topics

Related

Massachusetts General Hospital researchers have created a deep learning convolutional neural network capable of discriminating adult ADHD patients from healthy controls.

Related

Quantitative measures of functional connectivity on continuous EEG predicted success in weaning patients with refractory status epilepticus from intravenous anesthetics.