Timing of Peri‐Ictal MRI Abnormalities in Status Epilepticus – One Size Does Not Fit All

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Objective: We aimed to investigate timing of occurrence of peri‐ictal MRI abnormalities – a potential risk biomarker of status epilepticus‐related cerebral injury (t2). Methods: This prospective study enrolled adult patients with status epilepticus and acute magnetic resonance imaging (MRI); patients with peri‐ictal MRI abnormalities underwent follow‐up MRI 4 weeks later. Predictive model, using logistic regression, integrated clinical factors (duration, semiology, and etiology of status epilepticus and the patients’ level of consciousness) to prognosticate occurrence of peri‐ictal MRI abnormalities. Cerebral injury due to status epilepticus was assessed by comparative volumetric analysis of acute and follow‐up MRIs. Results: Among 256 patients, 137 (53%) had peri‐ictal MRI abnormalities. The likelihood of their occurrence increased over time under the influence of semiology and etiology of status epilepticus as well as the patients’ level of consciousness: it was highest in non‐convulsive status epilepticus caused by acute primary or secondary etiologies at 10 minutes after onset in patients with stupor/coma (81–85%); it increased at 24 hours to 92%; and at 48 hours to 95%. Conversely, in alert/somnolent patients with prominent motor symptoms and acute triggering factors associated with epilepsy, the possibility of developing peri‐ictal MRI abnormalities at 24 hours was 4 to 5% and at 48 hours it was 6 to 11%. In 28 of 45 (62%) of follow‐up MRIs, structural long‐term consequences of status epilepticus were observed in the form of either cortical and hippocampal atrophy or global cerebral volume loss. Interpretation: In this novel multimodal approach based on MRI data, the patients’ level of consciousness, etiology, and duration of status epilepticus offers insights into the risk of possible brain injury in status epilepticus. ANN NEUROL 2026;99:523–534
Funding Information
  • Austrian Science Fund (KLI 969‐B)

This publication has 40 references indexed in Scilit: